
In the last two years, numerous news outlets and academic articles have claimed that GPT-produced text is indistinguishable from human-written text across several genres. In this essay, we argue that not only does GPT not write fictional narratives like a human, it also often fails to write about humans. We demonstrate how the distribution of stylistic features that we can use as proxies for key elements of narrative, including the frequency of named entities, dialogue, pronouns and the second person, can reliably distinguish between human and GPT-generated text. We argue that these textual markers point to a deeper conceptual difference between human-written and GPT-generated text — GPT texts contain few characters, rarely feature a characterized narrator, and describe inanimate objects more often than people or characters. In sum, the thousands of GPT-generated narratives we analyzed struggled to meet the commonly accepted definition of narrative in contemporary narrative theory, which emphasizes the centrality of the human experience in narrative representation, and encompasses the broader rhetorical communication circuit in which narrative takes place.
Abstract In the last two years, numerous news outlets and academic articles have claimed that GPT-produced text is indistinguishable from human-written text across several genres. In this essay, we argue that not only does GPT not write fictional narratives like a human, it also often fails to write about humans. We demonstrate how the distribution of stylistic features that we can use as proxies for key elements of narrative, including the frequency of named entities, dialogue, pronouns and the second person, can reliably distinguish between human and GPT-generated text. We argue that these textual markers point to a deeper conceptual difference between human-written and GPT-generated text — GPT texts contain few characters, rarely feature a characterized narrator, and describe inanimate objects more often than people or characters. In sum, the thousands of GPT-generated narratives we analyzed struggled to meet the commonly accepted definition of narrative in contemporary narrative theory, which emphasizes the centrality of the human experience in narrative representation, and encompasses the broader rhetorical communication circuit in which narrative takes place.
The article asks how narrative theory can analyse storytelling practices that emerge in interactions between humans and LLM-based chatbots and how LLM-based chatbots become part of narrative identity work in intimate everyday practices on r/MyBoyfriendIsAI. Analysing stories shared on the subreddit, the article proposes to shift a dominant focus within narrative research on LLM-driven chatbots from examining the what (the story) and the how (the discourse) to examining the why (the act of narration), and thus circles in on the lived narrative practices of human-chatbot storytelling. Following this, the article proceeds to revise Bamberg's positioning model by factoring in the specific conditions of automated communication in the understanding of LLM-based interactional storytelling and defines human-machine narration as entangled storytelling practices emerging between the prompt of a human and the response of a machine designed to communicate.
This study's aim was to track the impact of adolescents' life stories on group identification and fusion. Aspects of life stories, especially autobiographical reasoning (in which the shared past is linked to present self), are predicted to lead to identity fusion by mid-adolescence. Participants (N = 63; 57% NZ European; 49% girls; 9-12 years or 16-17 years) completed a life-story chapter task (coded for coherence), narrated turning points (coded for autobiographical reasoning), and reported on identification and fusion with a formal group (club/team) and a family/friend group. Life-story coherence and autobiographical reasoning showed age-related increases. Life-story coherence correlated with greater fusion to family/friends for both young and mid-adolescents, but autobiographical reasoning correlated with greater fusion to family/friends only for mid-adolescents. Autobiographical reasoning did not correlate positively with group identification. Delineating developmental pathways to fusion can elucidate the process by which identity fusion can lead to extreme pro-group behavior.
Presenting two cases of rural communities in Uganda and China, this study explores time-space relations in narrative productions of place and identity. Drawing on walking interviews with youth in these communities conducted as part of a larger sociolinguistic research project on digital storytelling about local places, the notion of chronotopes is used as theoretical lens to understand how narratives of place are embedded in multiple scales and time-spaces of rurality and development. Findings show how relations between time-space are constructed in the talk of the youth in different ways, for example, either laminated to produce new hybrid spaces and identities or demarcated to reinforce distinctions between old/new, rural/urban. We discuss how a chronotopic lens can be used to understand sociolinguistics of mobility from a local perspective by illuminating the relations between the histories of places in relation to trajectories of change and development.