Abstract
In many ancestral and spiritual traditions, the figure of the witch is understood as one who “sees” beyond the visible—through dreams, intuition, or trance. In these contexts, image-making, often accompanied by verbal invocation, constitutes a performative and ritualistic act believed to shape reality and mediate the sacred. By drawing parallels between the ritual use of images in ancestral practices and the algorithmic generation of images through linguistic prompts, this research proposes that AI image production may be interpreted as a contemporary form of symbolic invocation. The resulting image becomes a conjuration: a visual manifestation of linguistic intention that echoes in some ways the structure of ritual. AI-generated imagery, however, unlike traditional ritual practices that aim to liberate, often reinforces stereotypes due to its reliance on biased data. For example, prompting “witch” typically yields harmful, clichéd depictions. Feminist thinkers have shown that reclaiming such terms can challenge patriarchal narratives. This paper addresses the question of how critically engaging with prompt-based image generation may reclaim symbolic power and conjure counter-narratives to patriarchal representations of women, femininity, and the witch.
Details
Presentation Type
Paper Presentation in a Themed Session
Theme
KEYWORDS
Ritual, Woman, Witch, Femininity, Feminism, Image, Generative AI