character

How can a convincing interactive character, with apparent psychological depth, be modelled in a playable narrative that adapts to a reader’s choice? This is the central question of my practice-based research that I address through the authoring (in both natural language and computer code) of an interactive text-based psychological thriller Stitched Up.

Narratives “by their nature are riddled with gaps” and characters are “some of narrative’s most challenging gaps” (Abbott 2008), yet filling in these gaps can be an enthralling source of readerly pleasure. On the other hand, flat characters “seem to exist on the surface of the story, along with objects and machines. There are no mysterious gaps to fill since what you see is what you get” (Abbott 2008). The majority of simulated characters in video games and interactive adventures tend to be more flat than round probably because, as Montfort (2007) has argued, a flat character can still be compelling and meaningful due to the nature of simulation, especially when combined with narration. Nevertheless, I aim to create round simulated characters in Stitched Up. These individuals in the storyworld will be compelling precisely because they are complex and undergo development as a result of reader–player interaction. In my playable psychological thriller, the readerly process of filling in the characters’ “mysterious gaps” is the core gameplay loop.

Stitched Up is based around the idea of a character as a “black box”. An observer or external entity can only infer what is inside a black box from its inputs and outputs. Interaction between two human beings could be viewed similarly. One person can only infer what the other one is thinking and feeling from their outputs, from their behaviour or what they say.

Since an interactive character must be constructed in code, I am researching how the properties and processes of programming in JavaScript can be used as functional metaphors to represent the psychological make-up of fictional characters and their dynamic interpersonal relationships. In JavaScript, as in other programming languages, encapsulation (the technical term for the black box) is an important strategy for organising complex code into modules (and/or functions), whereby internal code is hidden from external objects so that they can interact with each other safely and effectively via an Application Programming Interface (API). In this paper, I will discuss how I am repurposing the modularity of such JavaScript design patterns to dynamically model the internal mental states of my interactive fictional characters – their emotions, memories, moral values, opinions, etc. – and how this affects the process of creatively writing characters in natural language. Overall, this entails developing a modular form of character design where these attributes are discrete elements that can be amalgamated and delivered in multiple combinations yet still offer an individuated, meaningful encounter with a person in a storyworld.

(source: ELO 2015 Conference Catalog)

ABSTRACT: We consider a specific character, Princess Charlotte, in the 1999 interactive fiction work Varicella by Adam Cadre. To appreciate and solve this work, the interactor must both interpret the texts that result (as a literary reader does) and also operate the cybertextual machine of the program, acting as a game player and trying to understand the system of Varicella’s simulated world. We offer a close reading focusing on Charlotte, examining the functions she performs in the potential narratives and in the game. Through this example, we find that in interactive fiction — and we believe in other new media forms with similar goals — works must succeed as literature and as game at once to be effective. We argue that a fruitful critical perspective must consider both of these aspects in a way that goes beyond simple dichotomies or hierarchies.

Blue Lacuna is an ambitious new long form interactive fiction comprising nearly 400,000 words of prose and natural language source code. The longest work yet produced in the Inform 7 language, it is also among the most substantial text-based story games in existence, an interactive novel with an average play time of fifteen to twenty hours. In development between 2006 and 2009, Blue Lacuna features several experiments of interest to creators of long-form interactive stories. This paper describes these experiments and performs an anecdotal post-mortem on what succeeded and failed in the project's realization. I focus on how successful I was at achieving my three principal goals: 1) simplifying the IF interface so those unfamiliar with the medium can easily participate, 2) telling a story which revolves around the player's ability to make choices with real dramatic repercussions, and 3) creating a character able to form a complex relationship with the player across the span of a novel-length story. Among the mechanics discussed are Blue Lacuna's streamlined keyword-based system for entering commands, its attempts to match a much broader range of input styles than traditional IF, and various techniques to adapt the story to include the player's narrative goals, such as tracking which character the player thinks the story is about. I'll also discuss the design of the story's central character, a mentally unstable castaway named Progue, who evolves into one of twelve archetypes (such as friend, mentor, lover, or sycophant) based on the way the player treats him during 70 potential scenes across ten chapters of story.

While much of the attention towards ergodic fiction has been focusing on plot (either dynamic or multiple-path), its characters still lack complexity and expressiveness. In this paper we will explore two different techniques to face this problem.

One major issue in videogames is the lack of personality in user-controlled characters. In other words, the author of a videogame cannot give a deep personality to her character, because the user will be the one who will control it. For example, you cannot design a melancholic, non-violent character, if there is a knife available in the environment. Many users would just take the knife and start a Doom-like game, turning the originally pacifist character into a serial killer. The designer can obviously prevent the user from manipulating harmful objects. However, through this arbitrary rule, users will see their freedom limited. This would also diminish the environment's coherence: why some objects can be manipulated and other cannot?

So, how can we restrain the users' freedom of choice without disappointing she? Actually, the game structure does exactly that. Games set rules that limit the user's actions and usually players accept them as natural. Thus, the simplest workaround to achieve our goal is to transform the character impersonation into a game. Instead of "being" the character, the user would "pretend to be" the character. If the user does not behave according to her role, the elements of the environment (including other characters) will react in a way that may difficult the user's performance in the game.

All these elements do not necessarily imply major technical difficulties for its realization. The idea, a simple change of point of view in the traditional perspective of ergodic characters, is almost banal. However, current videogames have not explored this path, which, as we will show, can open interesting possibilities.

In videogames, the usual metaphor is "the user IS the character". Some time ago, we proposed a system ("A Blueprint for Laura") where the user was not the character. Instead, he was just able to modify the environment in order to affect the independent, computer-controlled, main character. We propose to upgrade this system, adding user-controlled secondary characters, while keeping the main one controlled by the computer.

A clear way to explain this is using the movie "The Truman Show" as an example. In this movie, Truman is not aware that his life is an illusion, taking place on a TV set, where everybody he knows are actors. Now imagine "The Truman Show" as a videogame, where Truman is an independent character, and the user just controls the environment and secondary characters. The goal is to keep Truman thinking that his world is real, while trying to influence on his actions.

In such a system, we not only break with the convention "the user is the main character", but we are also able to have a main character (controlled by AI) with a richer and deeper personality. Now the author has much more control over his character than in traditional videogames. The "pretender" technique can be used in this system to control the secondary characters, who have to act without doing any suspicious action that let Truman suspect about his condition.

(Source: DAC 1999 Author's abstract)

Harger's interview with Mateas and Stern focuses on the development of theiir conception of interactive drama and the Façade project.