A critical look at the views on authorship in story generation research

Submission Type

Paper

Start Date/Time (EDT)

20-7-2024 10:30 AM

End Date/Time (EDT)

20-7-2024 11:30 AM

Location

Algorithms & Imaginaries

Abstract

Computer science researchers have long been interested in generating creative texts with computers, and the development of story-generating programs has formed its own field of research, closely tied with research on computational creativity and artificial intelligence. In this presentation, I examine how authorship of computer-generated texts is approached in story generation research. I focus on the following questions: How is authorship discussed in relation to its conceptualisations in literary theory? How are the contributions and cooperation between the human author(s) and the program described in the research? The research material consists of story generation publications published during the last decade, selected based on recent reviews that discuss the state-of-the-art of story generation.

There are many, partly overlapping definitions of authorship in literary theory, such as as the romantic conceptualisation of the author as a creative genius, the institutional role of the author that is tied to print literature and copyright, as well as the author’s role as the party that is responsible for the aesthetic whole of the work. In my presentation I examine how these different aspects of authorship are present in the research material.

In electronic literature, the ambiguity of authorship is typical, as technology has made techniques such as combining and modifying existing texts considerably easier than before. Authors of electronic literature often elude disciplinary categorisations such as a poet or an engineer, and these different roles can merge. In story generation, ascribing authorship is also complicated by factors such as the design of the program, the data the program is given, as well as the program’s possible role as a co-author. My presentation is focused on how the interaction and collaboration between the program and its programmer – the machine and the human – is discussed in the research material in relation to authorship.

Bio

Tuuli Hongisto is a PhD student majoring in comparative literature at the University of Helsinki. She graduated from the University of Helsinki in 2020 with comparative literature as her major (thesis topic: “Essential narrative features in story generating algorithms”). Her PhD project focuses on the topic of reader- and authorship of computer-generated texts.

This document is currently not available here.

Share

COinS
 
Jul 20th, 10:30 AM Jul 20th, 11:30 AM

A critical look at the views on authorship in story generation research

Algorithms & Imaginaries

Computer science researchers have long been interested in generating creative texts with computers, and the development of story-generating programs has formed its own field of research, closely tied with research on computational creativity and artificial intelligence. In this presentation, I examine how authorship of computer-generated texts is approached in story generation research. I focus on the following questions: How is authorship discussed in relation to its conceptualisations in literary theory? How are the contributions and cooperation between the human author(s) and the program described in the research? The research material consists of story generation publications published during the last decade, selected based on recent reviews that discuss the state-of-the-art of story generation.

There are many, partly overlapping definitions of authorship in literary theory, such as as the romantic conceptualisation of the author as a creative genius, the institutional role of the author that is tied to print literature and copyright, as well as the author’s role as the party that is responsible for the aesthetic whole of the work. In my presentation I examine how these different aspects of authorship are present in the research material.

In electronic literature, the ambiguity of authorship is typical, as technology has made techniques such as combining and modifying existing texts considerably easier than before. Authors of electronic literature often elude disciplinary categorisations such as a poet or an engineer, and these different roles can merge. In story generation, ascribing authorship is also complicated by factors such as the design of the program, the data the program is given, as well as the program’s possible role as a co-author. My presentation is focused on how the interaction and collaboration between the program and its programmer – the machine and the human – is discussed in the research material in relation to authorship.