📄 Book Reviews: Foundations of Illocutionary Logic

Summary

Leonard Bolc reviews Foundations of Illocutionary Logic by John R. Searle and Daniel Vanderveken (Cambridge University Press, 1985). The book sets out to give speech acts a formal logic. It breaks the force of an utterance into separate components and builds the full space of forces from a small base. It then states axioms and theorems that govern success, commitment and consistency among acts.

The early chapters introduce the concepts and a set-theoretic notation. Later chapters develop the logic itself. The final chapter catalogues English verbs together with their semantic analyses.

Bolc praises the many examples. He also values that some theorems show where classical logic does not carry over. His criticisms are three:

• the book covers only English;
• it never addresses use in NLP systems;
• it has no index.

He still recommends it as a stimulating contribution to logical accounts of natural language.

Key points

• Illocutionary force is split into seven components:
  • point;
  • strength of point;
  • mode of achievement;
  • propositional content conditions;
  • preparatory conditions;
  • sincerity conditions;
  • strength of sincerity conditions.
• There are five illocutionary points: assertives, directives, commissives, declarations and expressives. The point is the core component, and the other components only modulate it.
• Five elementary forces are given, and all other forces are built recursively from them through operations on force.
• The book defines successful performance, non-defective performance and failure. It also defines strong commitment, illocutionary consistency and congruence. Chapters 6–8 collect the axioms and theorems.
• The last chapter is a lexicon of English speech-act verbs with semantic descriptions tied to the logic. The reviewer flags the English-only scope and the absence of an applications chapter.

Interesting ideas

• I1. Represent a speech act as a bundle of components rather than as one atomic label, so that "request" and "implore" share a point and differ only in strength.
• I2. Generate the inventory of forces compositionally from five primitives plus operations, instead of listing them flat.
• I3. Preparatory conditions act like presuppositions: asking someone to stop shouting presumes that shouting is going on.
• I4. Propositional content conditions are tense- and time-sensitive, for example one cannot apologise for a future event. Sincerity clashes such as expressing sorrow while denying it reveal defective acts.
• I5. Pair a lexicon of English verbs with formal force descriptions.

For Mova

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

• I1 (frames): prag could store speech acts as feature records (point, strength, mode, conditions), making each act label explainable and the differences between acts comparable.
• I2 (frames): a small generative speech-act inventory keeps the frame layer compact and lets an unseen performative verb be described as a primitive plus modifications.
• I3 (world): when prag detects a directive or apology, it could add the implied preparatory state to the world of the text, so readers can answer questions about it.
• I4 (learn): such constraints make cheap negative controls, flagging frame outputs that violate them and serving as hints for irony detection.
• I5 (frames): the verb catalogue idea suggests attaching force features to VerbNet performative verbs, linking lemma to speech act directly.