Elamur Research⌕
☰
home / papers / 1998 / Eliminating deceptions and mistaken belief to infer conversational implicature

Eliminating deceptions and mistaken belief to infer conversational implicature

Paper · 1998 · Mark Lee and Yorick Wilks

Summary

Caveat: the supplied extraction contains only a page marker ("с. 1") and no body text. This summary relies on the paper's title and publicly known abstract, so treat it as provisional until the full text is checked. Lee and Wilks address a gap in Gricean accounts of conversational implicature. Standard treatments say a hearer infers an implicature when a cooperative speaker visibly flouts a conversational maxim. But the same surface violation could also come from a speaker who is lying or who simply holds a false belief. The authors argue that these three situations can be told apart by the nested beliefs (speaker about world, speaker about hearer, and so on) that the hearer must attribute to each participant. Implicature is inferred only after deception and mistaken belief have been ruled out. The work fits Wilks's belief-ascription tradition (ViewGen-style nested belief environments). It matters because a dialogue system that reads every maxim violation as an implicature will misinterpret liars and misinformed speakers.

Key points

  • Authors: Mark Lee and Yorick Wilks; arXiv:cmp-lg/9806005 (1998).
  • Problem: earlier implicature work does not separate flouting from deception and from mistaken belief, which all look alike on the surface.
  • Approach: each case is characterised by a distinct pattern of beliefs ascribed to speaker and hearer, including beliefs about each other's beliefs.
  • Inference strategy: the implicature reading is reached by elimination, once the deceptive and mistaken readings fail.
  • No corpus, dataset or quantitative evaluation can be reported from the provided text. The full paper must be consulted for any examples or formal details.

Interesting ideas

  • I1. Treat "the speaker broke a maxim" as an ambiguous observation with competing explanations (flouting, lying, being wrong), and only commit to implicature once the rivals are excluded. This avoids over-reading pragmatic meaning.
  • I2. Separate pragmatic cases by the nested belief configurations they require, not by surface cues. This gives a principled, checkable criterion instead of heuristics.
  • I3. Inference by elimination is a natural format for transparent reasoning, because each rejected alternative leaves a stated reason.
  • I4. Deception and honest error are distinct states that share the same false utterance, so a system must track whether the speaker knows the truth.

For Mova

How these ideas could serve Mova, our small open language model — its files, stages and tests:

  • I1 (frames): crate prag could label every maxim-violation candidate with explicit hypotheses {implicature, deception, mistaken belief} and log why each one was kept or dropped, keeping the step explainable.
  • I2 (world): the world of the text could keep per-agent belief layers (what X believes, what X believes Y believes) so prag can query them when classifying a violation.
  • I3 (answer): readers answering fable or tale questions such as "did the fox lie or was it mistaken?" could expose the eliminated alternatives as part of the answer's justification.
  • I4 (learn): evaluation sets could include negative controls with lying versus mistaken speakers, checking that the implicature detector does not fire on them.

Our summary of the paper, in our own words — not its text. The work belongs to its authors; read it at the links above.

 Institutions

🎓University of SheffieldSheffield · United Kingdom113 works · 21,542 citations

 Fields, topics and concepts

📚 Shelves 4The field of science this work belongs to (OpenAlex fields).Computer Science · 16,667 worksArtificial Intelligence · 16,194 worksArts and Humanities · 16,039 worksLanguage and Linguistics · 15,891 works

🧭 Topics 1Research subjects and languages the work is about.Implicature · 138 works

Cite

Cite as Mark Lee and Yorick Wilks. 1998. Eliminating deceptions and mistaken belief to infer conversational implicature. arXiv:cmp-lg/9806005.

The paper itself, not this page. BibTeX from the library's record of the work. DOI 10.48550/arxiv.cmp-lg/9806005 · arXiv cmp-lg/9806005

@misc{lee1998eliminating,
  title = {{Eliminating deceptions and mistaken belief to infer conversational implicature}},
  author = {Mark Lee and Yorick Wilks},
  year = {1998},
  doi = {10.48550/arxiv.cmp-lg/9806005},
  eprint = {cmp-lg/9806005},
  archivePrefix = {arXiv},
  url = {https://doi.org/10.48550/arxiv.cmp-lg/9806005}
}

References

14 references, 3 of them with a card of ours. ↗ leads to the work, «find» searches for one we have no id for yet.

  1. Allen, J. (1983). Recognising intentions from natural language utterances. In Brady, M. and Berwick, R., editors, Computational Models of Discourse. MIT Press, Cambridge, MA. find
  2. Ballim, A. and Wilks, Y. (1991). Artificial Believers. Lawrence Erlbaum Associates, Hillsdale, New Jersey. find
  3. Carberry, S. (1990). Plan recognition in natural language dialogue. MIT Press, Massachusetts. find
  4. Grice, H. (1975). Logic and conversation. In Cole, P. and Morgan, J., editors, Syntax and Semantics, volume 3: Speech Acts. Academic Press, New York. ↗
  5. Lee, M. and Wilks, Y. (1996a). An ascription-based approach to speech acts. In Proceedings of the 16th Conference on Computational Linguistics (COL-ING-96), Copenhagen.
  6. Lee, M. and Wilks, Y. (1996b). Speech acts on demand. In Proceedings of the 5th International Pragmatics Conference, Mexico City. find
  7. Leech, G. (1983). Principles of Pragmatics. Longman, London. find
  8. McAllester, D. and Rosenblatt, D. (1991). Systematic nonlinear planning. In Proceedings of AAAI90, pages 634–639. find
  9. Perrault, C. (1990). An application of default logic to speech act theory. In P.R. Cohen, J. Morgan, M. P., editor, Intentions in Communication. MIT press. ↗
  10. Pollack, M. (1990). Plans as complex mental attitudes. In Cohen, P., Morgan, J., and Pollack, M., editors, Intentions in Communication. Bradford/MIT press, Cambridge, Mass. ↗
  11. Searle, J. (1969). Speech acts. Cambridge University Press, Cambridge.
  12. Sperber, D. and Wilson, D. (1986). Relevance: Communication and Cognition. Basil Blackwell Ltd., Oxford. find
  13. Utsumi, A. (1996). A unified theory of irony and its computational formalization. In Proceedings of the 16th Conference on Computational Linguistics (COL-ING-96), Copenhagen.
  14. Wilks, Y., Barnden, J., and Ballim, A. (1991). Belief ascription, metaphor and intensional identity. Cognitive Science, 15(1). context bel(System, bel(Simon, deleted(_,sub_directory) -> not(happy(David)))) bel(System, bel(Simon, bel(System, deleted(System, sub_directory)))) bel(System, bel(Simon, deleted(System, sub_directory))) i. bel(System, bel(Simon, not(happy(David)))) ascribe(Simon, System, deleted(_,sub_directory) -> not(happy(David)))))) ii. bel(System, bel(Simon, bel(System, deleted(_,sub_directory) -> not(happy(David))))) modus_ponens(System, deleted(_,sub_directory) -> not(happy(David), deleted(System, sub_directory), not(happy(David))) iii. bel(System, bel(Simon, bel(System, not(happy(David))))) goal accomplished bel(System, goal(Simon, bel(Sys, not(happy(David))) find

Updated

↑ back to top

This page in Markdown → https://md.research.elamur.ai/free/en/papers/lee-1998-eliminating-deceptions-mistaken-belief-infer/