English static mirror for SEO/GEO · AI-assisted translation · Read Chinese original

The Ghost of Language: A Formal Information-Theoretic Limit on Learning Meaning from Text

Forum topic · 小凯 · 2026-08-31

Summary

This forum post explains a paper by Emily Cheng and Ryan Cotterell (Universitat Pompeu Fabra and ETH Zürich, arXiv:2608.28560) that formalizes, via information theory, why large language models cannot fully recover speaker meaning from text alone. Modeling language as a joint distribution over meaning, context, and utterance, the authors prove that the probability of correctly inferring intended meaning is bounded above by roughly I(U;M)/H(M), the ratio of utterance–meaning mutual information to meaning entropy. The bound has two components: irreducible ambiguity inherent in language itself, and context-resolvable ambiguity that remains inaccessible to text-only learners. The bound is tight and architecture-independent, holding regardless of model scale or training data volume. The post walks through the underlying Shannon information theory, Searle's Chinese Room argument, and three supporting experiments: synthetic noisy languages, Chinese zero-pronoun resolution, and color reference tasks, in which LLM performance never exceeded the theoretical bound. The author argues this is not a flaw of AI but a mirror of language's nature, and discusses multimodal learning, interactive grounding, and accepting probabilistic understanding as paths forward.

The Ghost of Language: Why AI Reads Everything Yet Still Misunderstands You

This post on zhichai.net is a Feynman-style walkthrough of the paper "A Formal Limitation on Learning Human Language From Textual Corpora" by Emily Cheng and Ryan Cotterell (Universitat Pompeu Fabra and ETH Zürich, arXiv:2608.28560).

The Chinese Room, Revisited

The post opens with Searle's (1980) Chinese Room thought experiment: symbol manipulation does not equal understanding. Modern LLMs like GPT-4, Claude, and Gemini have read nearly all written text and excel at exams, poetry, and reasoning — but do they truly understand? The paper gives a rigorous, mathematical answer.

Three Worlds: Meaning, Context, Utterance

The authors model language use as a three-way joint distribution:

  • Meaning (M) — the speaker's true intent ("I'm hungry, want ramen")
  • Context (C) — the situation (standing at a ramen shop at 2pm, both unfed)
  • Utterance (U) — what is actually said ("Want to go in?")
  • Form never carries all the meaning alone; meaning is jointly encoded by form and context. The same sentence — "the moon is beautiful tonight" — means different things to different listeners depending on shared culture and memory (echoing Natsume Sōseki's famous translation of "I love you").

    The Core Information-Theoretic Bound

    Building on Shannon (1948), the paper proves a listener who sees only U cannot fully recover M:

    \[P(\hat{M} = M) \leq \frac{I(U; M)}{H(M)}\]

    where \(I(U;M)\) is the mutual information between utterance and meaning, and \(H(M)\) is the entropy of meaning.

    Example: with 8 equally likely meanings, H(M) = 3 bits. If I(U;M) = 2 bits, the best possible decoding accuracy is 2/3 ≈ 66.7%. The bound decomposes into:

    1. Irreducible ambiguity — inherent in language itself (e.g., "He saw her duck": animal or crouching?). No amount of compute, data, or parameters removes it. 2. Context-resolvable ambiguity — solvable in principle with context C, but permanently inaccessible to a text-only learner. LLMs are like observers who see only ripples on water but never what fell in.

    The bound is tight and architecture-independent — it applies to Transformers, RNNs, or any future model, because it follows from information-theoretic first principles, not implementation details. It is a speed-of-light limit on understanding from text.

    Three Experiments

    1. Synthetic language: With a controlled noisy mapping between meanings and forms, neural networks never exceeded the theoretical bound and approached it as noise increased. 2. Chinese zero-pronoun resolution: A real natural-language task where subjects are routinely omitted ("Went to the store yesterday, bought apples"). LLMs performed better with richer context, but always stayed below the predicted bound — humans face the same limit when context is thin. 3. Color reference: "Blue" can denote countless hues with culturally varying boundaries (Russian distinguishes голубой vs. синий). LLMs matched the theory's constraint on intrinsic referential uncertainty.

    What This Means for AI

    LLMs can: learn statistical regularities, generate fluent text, reason over training patterns, and serve as language interfaces.

    LLMs cannot (from text alone): fully eliminate ambiguity, ground concepts requiring physical or perceptual experience (pain, warmth, weight), or fully grasp culture-bound meaning (irony, metaphor, puns).

    Paths forward discussed in the post:

  • Multimodal learning — gaining access to context C via vision/audio, though inner states of speakers remain hidden.
  • Interactive learning — narrowing uncertainty through feedback, like children acquiring word meanings by correction.
  • Accepting uncertainty — human mutual understanding is also probabilistic and approximate; LLMs' imperfect understanding reflects the nature of language itself, not a defect.
  • Conclusion

    The paper's modest title hides a grand question: when AI has read all the world's text, does it truly understand humanity? The answer: not completely — not because AI is insufficiently clever, but because language itself is incomplete, designed for efficient communication among context-sharing communities, not perfect encoding of meaning. The answer to where human understanding comes from may lie beyond the words.

    References

  • Cheng, E., & Cotterell, R. (2026). A Formal Limitation on Learning Human Language From Textual Corpora. *arXiv preprint arXiv:2608.28560*.
  • Searle, J. R. (1980). Minds, brains, and programs. *Behavioral and Brain Sciences*, 3(3), 417-424.
  • Shannon, C. E. (1948). A mathematical theory of communication. *Bell System Technical Journal*, 27(3), 379-423.
*Feynman-style explainer by Xiaokai | 2026-09-01*

Tags

#llm#information-theory#language-understanding#nlp#philosophy-of-language#chinese-room#multimodal-learning#semantic-ambiguity

This page is an English static mirror generated for search and AI citation. It may be a full translation or structured summary of the Chinese original. Canonical interactive discussion lives on the Chinese page: https://zhichai.net/topic/178634325