Inquiries into Truth and Interpretation

Citation

Davidson, D. (2001). Inquiries into Truth and Interpretation (2nd ed.). Oxford University Press.

Chapter Summary

The book is a collection of eighteen essays, about 280 pages of essays and a short appendix, preceded by a preface to the second edition and a ten-page introduction. The publisher notes that this edition adds one previously uncollected essay to the 1984 volume. The essays fall into five thematic groups, which the publisher's description sets out: the constraints on a theory of meaning (Essays 1–5), quotation, indirect speech and mood (6–8), interpretation and its evidence (9–12), metaphysical consequences (13–16), and the hard cases of metaphor and convention (17–18). They refer to one another by essay number, so the order matters.

Essay 1: Theories of Meaning and Learnable Languages

Asks what a language must be like to be learnable, rejects the "building-block" picture of language learning, and proposes that a language is learnable by a finite creature only if it has finitely many semantic primitives. By this test, accounts that treat quotations as names, analyse belief attributions through unstructured entities, or posit intensional entities are criticised.

Essay 2: Truth and Meaning

Turns the learnability requirement into a constraint on a theory of meaning: it must show how finitely many primitives yield the meaning of every sentence of an unlimited language. Davidson argues that a Tarski-style truth theory meets this requirement and gives a sentence's meaning through its place among entailments, and he considers objections such as indexicals and non-extensional contexts.

Essay 3: True to the Facts

Asks whether truth defined recursively through satisfaction can be developed into a correspondence theory. Davidson argues that both the redundancy theory and the usual correspondence theories fail, the latter because coextensive redescriptions collapse all facts into one "Great Fact."

Essays 4–5: Semantics for Natural Languages; In Defence of Convention T

Essay 4 says a truth theory gives a testable answer to the demands on a theory of meaning, with a metalanguage that goes minimally beyond the object language, and considers non-extensional constructions. Essay 5 defends Tarski's Convention T and first-order quantification, and argues that modal, possible-worlds and substitutional approaches rely on relativised notions of truth.

Essay 6: Quotation

Applies the truth-theoretic approach to quotation, arguing that a quotation must be analysed by finite means and that the quoted words are used, not merely mentioned.

Essay 7: On Saying That

Asks for the logical form of attributions such as "Galileo said that p." Davidson develops "parataxis," on which the attribution splits into two sentences linked by a demonstrative "that," and the link holds if the utterances "say the same."

Essay 8: Moods and Performances

Asks whether a truth theory can handle mood and force. After examining ways of individuating mood, Davidson sets constraints and extends the paratactic analysis to account for it.

Essay 9: Radical Interpretation

Asks what knowledge suffices to turn an uninterpreted utterance into an interpreted one. The essay sets out the evidence on which interpretation can proceed without presupposing facts about the speaker's attitudes, contrasts the approach with Quine's radical translation, and presents the truth-theoretic account.

Essay 10: Belief and the Basis of Meaning

Continues Essay 9: meanings and attitudes must be assigned together, which requires the Principle of Charity, on which interpreters assume the speaker's beliefs are mostly true and maximise agreement. The essay discusses how much indeterminacy remains.

Essay 11: Thought and Talk

Asks whether thought can be explained independently of speech. Beliefs come in webs that fix each other's content, and only a creature that can interpret others and be interpreted can have thoughts.

Essay 12: Reply to Foster

Responds to John Foster's criticisms that the axioms of a truth theory must be law-like and that a theory's evidence may not make its use as an interpretation theory apparent, and answers them by recourse to the paratactic method.

Essay 13: On the Very Idea of a Conceptual Scheme

Attacks the intelligibility of conceptual relativism. Scheme-content dualism, the "third dogma of empiricism," is rejected, and the possibility of radically different conceptual schemes is argued to be untenable given what interpretation requires.

Essay 14: The Method of Truth in Metaphysics

Argues that successful communication shows a shared and largely true view of the world, so error is local. Truth is relative to languages that are largely correct about the world, and a truth theory can reveal the ontology a language requires.

Essays 15–16: Reality Without Reference; The Inscrutability of Reference

Essay 15 replies to Field and Harman, who held that truth theories cannot do the work of meaning theories without elucidating reference; Davidson argues reference is not the central semantic concept. Essay 16 treats Quine's ontological relativity and the indeterminacy of reference, and objects to the causal theory of reference.

Essay 17: What Metaphors Mean

Argues that metaphors have no special metaphorical meaning beyond their literal meaning; their effect is to make the hearer notice things, which explains why they resist paraphrase.

Essay 18: Communication and Convention

Examines conventionalist accounts of language by Dummett, Chomsky and Lewis, and concludes that convention is not a precondition for language.

Appendix

A short appendix (pp. 281–286) responds to replies by Lewis and Quine to Essay 10, on interpretive principles, the reducibility of the mental, and how much indeterminacy remains.

Key Concepts

Essays 1–5: Truth-theoretic semantics

• Learnability constraint: a language must have finitely many semantic primitives to be learnable.
• Truth theory as a theory of meaning: a Tarskian theory of truth that yields a truth condition for each sentence, treated as an empirical theory.
• Convention T: Tarski's adequacy condition on definitions of truth.

Essays 6–8: Logical form

• Parataxis: the analysis of "says that" as two sentences, one a demonstrative reference to the other.
• Logical form: the structure revealed by a truth theory that preserves entailments.

Essays 9–12: Interpretation

• Radical interpretation: interpreting a speaker without prior knowledge of their meanings or attitudes.
• Principle of Charity: assume the speaker's beliefs are largely true and consistent.

Essays 13–18: Metaphysics and the limits of semantics

• Scheme-content dualism: the picture that conceptual schemes organise an uninterpreted empirical content, which Davidson rejects.
• Indeterminacy of reference: Quine's claim that reference assignments are underdetermined, treated here from within truth theory.

Critical Analysis

Strengths

• A thematic structure. The five groups move from the constraints on a theory of meaning to extensions of it, to its evidence in interpretation, to its metaphysical consequences and finally to metaphor and convention, so the book works as a connected case and not only as a pile of papers.
• A programme under challenge. Essay 12 answers Foster, Essay 15 answers Field and Harman, and the appendix answers Lewis and Quine, which shows where the programme is contested.
• Short, self-contained essays. Essays run about ten to twenty pages, so individual ones can be assigned.
• Cross-references. The essays point to each other by number, so a reader can follow a thread such as finite learnability or charity across the volume.

Limitations

• Not a systematic exposition. The essays were written separately, so some material repeats and some arguments rely on essays elsewhere in the volume.
• A companion volume is assumed. The logical-form work in Essays 6 to 10 continues in Davidson's companion volume, Essays on Actions and Events, which is not part of this book.
• Heavy prerequisites. Tarski's truth definition, Frege's sense and reference and Quine on translation and ontology are taken for granted.
• Technical in places. Essays 5 to 8 and 15 to 16 depend on formal notions such as satisfaction and logical form.

Where It Fits in the Course

This is a primary-source volume for readers who have already worked through an introduction to the philosophy of language. Morris's An Introduction to the Philosophy of Language builds its chapters around classic texts, McGinn's Philosophy of Language: The Classics Explained explains ten of them step by step, Lycan's Philosophy of Language: A Contemporary Introduction organises the field around reference, theories of meaning, pragmatics and figurative language, and Miller's Philosophy of Language reconstructs the analytic tradition from Frege onward in detail. Miller and Lycan are good preparation for the Fregean and Tarskian background.

Martinich and Sosa's anthology The Philosophy of Language gathers primary papers by topic and is a good way to meet the authors Davidson answers, such as Quine, Dummett and Lewis, before or alongside Davidson.

How to Use This Book

The book is for graduate students and advanced undergraduates who know Frege, Russell, Tarski and Quine on translation.

A practical reading plan:

• Core: Essays 1 and 2 (the learnability constraint and the truth-theoretic proposal), then 9 and 10 (radical interpretation and charity), 11, and 13.
• Applications of the core: Essays 3 to 8 for correspondence, quotation, indirect discourse and mood; Essay 14 for the metaphysical consequences; Essay 17 on metaphor, which can be read independently.
• Replies and responses: Essays 12, 15 and 16, and the appendix, after the essays they respond to.
• Optional: Essay 18 on convention, once the main argument is clear.

Read the essay each response answers first, track the cross-references, and consult Essays on Actions and Events where logical form is at issue.