Sociological systems theory offers a rich vocabulary for meaning, observation, and communication, but that vocabulary has proven difficult to translate into executable, testable, and revisable procedures. This gap has become more consequential as large language models are increasingly used to read, summarize, and simulate social-scientific discourse without any shared standard for distinguishing genuine semantic stabilization from fluent but ungrounded continuation. This paper introduces the Information–Meaning–Existence (IBE) framework, which reconstructs Luhmannian second-order observation as a validation architecture. A coherence index tracks whether actualized meaning-relations remain viable as their context changes, and a companion diagnostic identifies when that viability collapses—distinguishing collapse as failure from collapse as a precondition for new distinctions. The framework specifies how its central quantities could be computed from text using existing embedding-based and information-theoretic methods, disambiguates several senses in which a meaning-structure can be said to “exist,” and extends the coherence index to register cases in which apparent stability is sustained by structural power rather than genuine resonance. A concluding discussion confronts this coherence-centered vocabulary with dialectical accounts of development in which contradiction, not coherence, drives change, and uses that confrontation to recalibrate the framework's own claims. The paper specifies what an empirical pilot would require to calibrate and test the framework, and explicitly states that absent such calibration, its present status is proto-scientific.
更多
查看译文
关键词
active inference,AI-assisted text analysis,computational social science,in-silico sociology,Luhmann,second-order observation