摘要
生成式AI图像的兴起,对图像阐释学带来了巨大挑战。唯有批判性地反思图像阐释学知识脉络中的三大基础命题——“再现”“象征”和“互文”,才能重构人工智能时代图像阐释学的知识框架。作为一种符号模式,“再现”揭示了世界对符号的依赖,亦揭示了图像阐释的语言基础。如果说传统图像的再现之本质是及物性、摹仿性、镜像性,生成式AI图像则否定了图像存在的再现基础,也否定了图像生产的再现语言,其本质上是不及物的、反再现的、生成性的图像。Sora作为世界模拟器的意义,并不仅仅体现为对世界的视觉理解,更为深刻的装置内涵在于形成了一套关于世界的“通用形象”,从而实现了文字之语象和图像之形象之间的通约可能。相较于语言文本而言,建立在互文性基础上的互文语境,为图像阐释提供了不可或缺的阐释规则和锚定体系。然而,生成式AI图像的出场,则宣告了伴随文本的消亡,亦宣告了互文世界的坍塌,图像不得不陷入一个没有符号“踪迹”的孤独世界。
The emergence of generative AI images poses significant challenges to the field of image hermeneutics.To reconstruct the knowledge framework of image hermeneutics in the era of artificial intelligence,it is imperative that we engage in critical reflection on three fundamental propositions within its knowledge context:"representation","symbolicity",and"intertextuality.As a semiotic mode,"representation"highlights the world's reliance on signs and exposes the linguistic foundations of image interpretation.Traditional image representation is characterized by transitivity,mimicry,and mirror-like qualities.However,generative AI images reject this representational basis of image existence and the language of image production.They are inherently intransitive,anti-representational,and generative in nature.The significance of generative AI,such as Sora as a world simulator,extends beyond its mere visual comprehension of the world.Its deeper implication lies in the construction of a"general image"of the world,thereby enabling the commensurability between the verbal image and the visual image.While language texts have traditionally provided a rich intertextual context grounded in intertextuality,offering indispensable interpretive rules and anchoring systems for image hermeneutics,the advent of generative Al images signals a paradigm turn.It announces the disappearance of co-texts and the collapse of the intertextual world,ultimately confining images to a desolate realm devoid of symbolic"traces".
出处
《南京社会科学》
北大核心
2024年第8期86-100,共15页
Nanjing Journal of Social Sciences
基金
国家社科基金重大项目“中华文化经典符号谱系整理与数字人文传播研究”(23&ZD211)的阶段性成果。
