Qur'anic Agent Identification: A Corpus-Based Framework for Human, Angelic, Jinn, Satanic, and Unknown Agents

Counts differ because each database indexes a different set of publications. We treat OpenAlex as the canonical count; Google Scholar is not shown (no API, and crawling it violates its ToS).

The identification of an agent from observable behavior is a classical problem in artificial intelligence, pattern recognition, information theory, and philosophy of science. A related but substantially more difficult problem arises when the possible agent classes are not restricted to ordinary human actors, but include entities described in a religious corpus. This paper proposes Qur’anic Agent Identification (QAID), a corpus-based framework for representing and analyzing agent identity within the Qur’anic textual domain. The framework considers several candidate classes, including human, angelic, jinn, satanic, and unknown agents. Rather than assuming that these classes can be directly detected from appearance or isolated behavioral observations, QAID treats identification as an inference problem under uncertainty. The central methodological principle is that an observable pattern does not automatically determine an agent’s ontology. A human-like appearance, unusual behavior, extraordinary knowledge, or a particular reaction to revelation may provide evidence, but none of these properties alone constitutes a sufficient identification rule. The framework introduces an agent feature vector, an evidence-typing scheme, an explicit unknown class, and the concept of observational equivalence. It further defines a computational pipeline through which Qur’anic descriptions can be converted into a machine-readable dataset without prematurely converting interpretation into fact. QAID does not claim to experimentally detect angels, jinn, satanic agents, or extraterrestrial beings. Its purpose is narrower and more technical: to formalize the problem of agent identification as a structured inference task and to investigate the limits of such identification when different ontological classes may produce observationally similar evidence.