자동주의 기계로서의 카메라와 객체지향적 사물-이미지: AI가 작동시키는 카메라의 가능성

초록

This paper reconsiders the camera not as a passive technological medium that merely substitutes for human vision, but as a machine of automatism that produces images through its relations with things and environments. It proposes the concept of the “thing-image” to designate an object-oriented image that emerges through asymmetrical relations among the camera, lens, light, sensor, algorithm, data, environment, and filmed object without being reducible to those relations. The paper first reexamines theories and practices of camera automatism in the age of cinema as historical precedents for reconsidering the possibility of post-anthropocentric image production, and then introduces Object-Oriented Ontology as its principal theoretical framework. From this perspective, the thing-image is understood not simply as an image of things, but as an image that emerges through relations among heterogeneous objects while remaining irreducible to the sum of those relations. The paper analyzes Abbas Kiarostami’s Five and Ten, Pedro Costa’s In Vanda’s Room, Jean-Luc Godard’s Goodbye to Language, and Véréna Paravel and Lucien Castaing-Taylor’s Leviathan to examine how the duration, mobility, prolonged dwelling, miniaturization, and nonhuman movement of the digital camera expand the conditions under which the thing-image can emerge. Building on these analyses, the paper finally considers the AI-operated camera as a further extension of camera automatism, in which continuous human control may be diminished and things and environments may participate more actively in image formation.

키워드

카메라 자동주의, 객체지향 존재론, 사물-이미지, 탈인간중심적 시각, AI가 작동시키는 카메라

출처

박영석. (2026). 자동주의 기계로서의 카메라와 객체지향적 사물-이미지: AI가 작동시키는 카메라의 가능성. 문학과 영상, 27권 2호, 549–581.

https://doi.org/10.36114/JLF.2026.9.27.2.549