آفرینش متن نمایشی به شیوۀ تعاملی: مطالعه‌ای میان‌رشته‌ای در تلاقی درام‌شناسی و هوش مصنوعی مولد

Background: In the contemporary era, concepts and expectations such as "personalization" and "customization" have transcended the realms of industry and commerce. It appears that if the world of performing arts and drama does not move toward fulfilling this audience demand, it will gradually lose its dedicated patrons. Can such an expectation be met within the dramatic universe? In other words, can individual audience members, according to their personal preferences, influence any dramatic narrative, live performance, film, or series, and alter it into any form or condition dictated by their individual taste? Do the definitions and rules of drama permit such a level of "interactivity" between the "audience" and the "artwork/author" for the creation and recreation of the dramatic text and performance? To address this audience demand, efforts have been made since the second half of the twentieth century to present interactive cinematic works. However, given that interactive films have not achieved widespread adoption, they seem to have failed to satisfy the surging audience interest, despite the remarkable development of digital communications. In recent years, numerous studies and attempts have been conducted to understand how audiences participate in and "interact" with the "narrative," "fictional work," and "game/dramatic work." Consequently, various methods and models have been introduced. By reviewing the existing literature, current interaction modalities and research axes can be classified into four general categories: 1. "Branching Narrative": Audience participation, decision-making, and choice within the boundaries of "predetermined paths and narrative turning points"; 2. "Interactive Storytelling" based on text, or a combination of text and images; 3. "Human User Participation" to author a "fictional narrative" within the framework of a virtual system or digital medium, followed by its "rewriting" by expert writer(s); and 4. Human interaction as a "user/audience" with "digital technology" for "content production." Given that none of the aforementioned methods and studies have addressed the "creation or recreation of the dramatic text" through emerging technologies to generate a transformed narrative from the original work via "audience interaction" and their "free choice," the gap for such a study becomes increasingly prominent.
Objectives: The present study aims to provide the components and requirements for extracting a theory to produce an interactive dramatic narrative. By presenting these narratives through an interactive system, it seeks to facilitate the audience's free interaction with drama—an interaction and connection that is considered an inevitable necessity for the future of the performing arts world. The central question of this research is: Which existing elements and definitions within the structure of drama can facilitate the formation of the algorithm required to design and execute free and pervasive audience interaction with a dramatic work within the context of Generative Artificial Intelligence (Generative AI)? Furthermore, what rules and conditions must be fulfilled for Generative AI to operationalize such a level of interaction?
Methodology: This interdisciplinary, descriptive-analytical study employs a multi-layered, integrative review framework tailored for complex, cross-disciplinary research. Initially, through a qualitative literature review, the study examines the foundations of dramaturgy to identify the capacities and functions of audience interaction, as well as the structure and architecture of drama translatable into machine language. It then introduces and investigates the cognitive models and computational semiotics relevant to the research topic. Finally, by analyzing interactive storytelling models, it presents the characteristics of the mediating grammar between drama and machine, alongside the primary components necessary to enable the interactive creation of dramatic narratives. To manage the conceptual breadth and specify the boundaries of the study, an initial search system was established based on general concepts and selected keyword combinations, drawing inspiration from a scoping review protocol for step-by-step screening based on the primary research question and detailed keywords. Furthermore, the older, foundational texts in the fields of dramaturgy and literary criticism required a traditional synthesis and purposeful summarization, which were analyzed and consolidated through a narrative review approach. Lastly, the selection and evaluation of AI-based interactive storytelling models were guided by the concepts and criteria of the Design Science Research (DSR) methodology, incorporating aspects such as effectiveness, solution-oriented design, and an evaluable theoretical background.
Conclusion: By studying and reviewing the perspectives and findings of scholars in the field of dramaturgy, as well as interactive storytelling research and models—ranging from Aristotle and Propp to Dramatron—all core and constitutive elements of drama can be conceptualized as an engineerable algorithm based on its internal geometry. By data processing and annotating these elements and classifying their interconnected and coherent system, we will possess the necessary structured architectural principles. These principles not only supply the rules for producing interactive dramatic works but also enable the dynamic and free interaction of audiences with dramatic texts and works by leveraging Generative AI to present and execute these pieces. Crucially, this is achieved by managing the most significant challenge of free and pervasive interaction: the balance between the "chaos of user/audience creativity" and the "coherent structure of drama." Furthermore, it must be emphasized that to achieve the first prototype of a model that enables the interactive presentation of dramatic works, three components of interactive creation and recreation are mandatory: 1. A technical infrastructure and Generative AI tools possessing high computational processing power; 2. An intermediary grammar that, in addition to converting abstract dramatic elements and attributes into a machine-understandable language, plays a vital role in the continuous process of machine learning; and 3. A control and management framework that monitors dramatic indicators, user-issued commands, and adherence to regulations and ethics. In this manner, the author's creation and creativity enter into a two-way relationship with the audience's choice and modification, thereby shaping the future of the dramatic text and artwork.