Scholarlink

<p>Recent advances in artificial intelligence and accumulating evidence from cognitive and clinical science have placed unprecedented pressure on traditional philosophical models of linguistic understanding. Classical approaches–truth-conditional, ability-based, and process-based–each isolate a single dimension of understanding and elevate it to explanatory primacy. As a result, they converge on the problematic implication that sufficiently advanced artificial systems qualify as genuine linguistic understanders, while simultaneously failing to account for systematic dissociations observed in human pathology. This paper argues that the source of this tension lies in a structural impoverishment of the inherited concept of understanding itself. A multi-layered and multi-axial model of linguistic understanding is proposed, according to which any instance of understanding is situated within a five-dimensional space defined by (i) the object of understanding, (ii) the mode of understanding (cognitive, emotional, practical), (iii) the depth or level of understanding, (iv) the social and normative frame of interpretation, and (v) biological and neurogenetic constraints. This framework dissolves persistent confusions in the AI understanding debate, explains selective dissociations in aphasia, autism spectrum conditions, and social-cognitive impairments, and re-maps the partial insights of classical theories as projections within a higher-dimensional structure. Linguistic understanding, on this view, is not a unitary epistemic or computational state, but a dynamically integrated form of situated participation in a meaningful world. Section&#xa0;5.6 specifies minimal and robust thresholds for when such participation counts as linguistic understanding.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Linguistic Understanding Beyond Knowledge, Ability, and Process: a Multi-Layered and Multi-Axial Model

  • 期刊论文

  • 发表日期：

    2026年3月6日

查看全文
  • Yongho Choi

摘要

Recent advances in artificial intelligence and accumulating evidence from cognitive and clinical science have placed unprecedented pressure on traditional philosophical models of linguistic understanding. Classical approaches–truth-conditional, ability-based, and process-based–each isolate a single dimension of understanding and elevate it to explanatory primacy. As a result, they converge on the problematic implication that sufficiently advanced artificial systems qualify as genuine linguistic understanders, while simultaneously failing to account for systematic dissociations observed in human pathology. This paper argues that the source of this tension lies in a structural impoverishment of the inherited concept of understanding itself. A multi-layered and multi-axial model of linguistic understanding is proposed, according to which any instance of understanding is situated within a five-dimensional space defined by (i) the object of understanding, (ii) the mode of understanding (cognitive, emotional, practical), (iii) the depth or level of understanding, (iv) the social and normative frame of interpretation, and (v) biological and neurogenetic constraints. This framework dissolves persistent confusions in the AI understanding debate, explains selective dissociations in aphasia, autism spectrum conditions, and social-cognitive impairments, and re-maps the partial insights of classical theories as projections within a higher-dimensional structure. Linguistic understanding, on this view, is not a unitary epistemic or computational state, but a dynamically integrated form of situated participation in a meaningful world. Section 5.6 specifies minimal and robust thresholds for when such participation counts as linguistic understanding.