Theories of Meaning: From Formal Logic to Cognitive Models

Chapter 1: What Is Semantics?

1.1 Definition and Scope

Semantics: The study of meaning as encoded in language. It is concerned with semantic knowledge – what speakers know about word and sentence meaning.

Linguistic Components:

  • Phonology: The sound system of a language.
  • Syntax: Sentence structure.
  • Semantics: Meaning.

1.2 Semantics vs. Semiotics

Semiotics: The general study of signs (developed by Charles Sanders Peirce).

  • Icon: A sign that resembles its object (e.g., a portrait).
  • Index: A sign with a causal link to its object (e.g., smoke indicates fire).
  • Symbol: An arbitrary, conventional link (e.g., language).

Ferdinand de Saussure distinguished between the signifier (form) and the signified (concept).

1.3 Core Challenges in Semantics

  1. Circularity of Definitions: A problem noted by lexicographers and semanticists alike. Dictionaries define words using other words, leading to infinite regress.
  2. Linguistic vs. Encyclopedic Knowledge: The challenge in distinguishing linguistic meaning from world knowledge. This was raised by Stephen Ullmann and Geoffrey Leech.
  3. Context-Dependence: Meaning varies depending on context. Highlighted by John Austin, Paul Grice, and Stephen C. Levinson.

1.4 Responses to the Challenges

  • Metalanguage: Advocated by Anna Wierzbicka, using a reduced core of universal terms. Problem: circularity can still persist.
  • Mental Lexicon: Explored in cognitive linguistics (e.g., Ray Jackendoff). Focuses on the psychological reality of stored word meanings.
  • Literal vs. Contextual Meaning: John Searle made distinctions between sentence meaning and speaker meaning, leading to the study of pragmatics.

1.5 Semantics in Grammatical Models

  • Noam Chomsky: Introduced the idea of modular grammar, where syntax, semantics, and phonology are separate but interacting components.
  • Ray Jackendoff: Meaning arises from conceptual structures integrated with syntax.
  • Compositionality (Frege’s Principle): The meaning of the whole is derived from the meaning of the parts plus the syntactic structure. This principle enables recursion and infinite productivity.

1.6 Key Conceptual Distinctions

Reference and Sense (Frege)

  • Reference: The actual object or entity.
  • Sense: The internal content or mode of presentation.

Example: “Morning star” vs. “Evening star” both refer to Venus but differ in sense.

Sentence, Utterance, Proposition (Searle, Frege)

  • Utterance: Context-bound speech.
  • Sentence: Abstract grammatical structure.
  • Proposition: Underlying, truth-evaluable content.

Literal vs. Figurative Language (Lakoff & Johnson, 1980)

  • Figurative use is often grounded in conceptual metaphor (e.g., Time is money: “spend time,” “waste time”).
  • Dead metaphors: Once figurative, now conventional (e.g., “table leg”).

Semantics vs. Pragmatics

  • Semantics: Literal, sentence-based meaning.
  • Pragmatics: Context-sensitive, speaker-intended meaning (Paul Grice introduced the Cooperative Principle and implicature).

Chapter 2: Meaning, Thought, and Reality

2.1 Introduction

John Lyons raised the question of how language connects with reality. Example: “That dog looks vicious” denotes a class and refers to a real entity.

2.2 Referential vs. Representational Views

Referential (Denotational) Semantics

  • Meaning equals correspondence to reality.
  • Rooted in Aristotelian and logical positivist tradition.
  • Sentences are truth-evaluable (Tarski’s semantic theory of truth).

Representational Semantics

  • Language is the expression of mental representations.
  • Developed by Ray Jackendoff (Conceptual Semantics) and George Lakoff (Cognitive Semantics).
  • Useful for understanding: fiction (e.g., Russell’s example: “The King of France is bald”) and counterfactuals.

2.2.3 Noun Phrases and Reference

Bertrand Russell focused on the logical form of definite descriptions, while Peter Strawson emphasized presupposition and speaker intent.

Noun phrases can refer to:

  • Individuals: a man, the dog.
  • Non-existents: the unicorn.
  • Groups: the committee (collective vs. distributive).
  • Abstracts: freedom, hunger.

2.3 Proper Names and Theories of Reference

Description Theory

Proponents: Gottlob Frege, Bertrand Russell, John Searle. Names are shorthand for definite descriptions (e.g., “Aristotle” = “the teacher of Alexander”).

Causal Theory

Proponents: Saul Kripke, Keith Donnellan, Michael Devitt & Kim Sterelny. Names refer via causal-historical chains, not sense or description.

2.4 Other Types of Reference

  • Generics: Whales are mammals — refer to kinds, not individuals.
  • Deixis: Context-dependent expressions (studied by Stephen Levinson).
  • Quantifiers: some, all — relate to classes, not individuals.

2.5 Reference and Logic

  • Alfred Tarski established formal semantics and model-theoretic truth.
  • Russell’s Theory of Descriptions distinguishes between surface grammar and logical form.

Concepts:

  • Extension: Actual referents.
  • Intension: Conceptual criteria.

2.6 Theories of Meaning and Reality

  • Linguistic Solipsism: Focuses on internal semantic structure (seen in structuralist approaches like Saussure).
  • Denotational Expansion: Uses tools like possible worlds (David Lewis, Richard Montague) to extend reference, employed in formal semantics.
  • Conceptual Structure Theory: Meaning is mental representation. Key proponents: Jackendoff, Lakoff, Barsalou.

Chapter 3: Word Meaning (Lexical Semantics)

3.1 Introduction

Lexical Semantics studies the meaning of individual words and how words relate to each other. Theories reflect influence from:

  • Saussure: Syntagmatic vs. paradigmatic relations.
  • Structuralist linguists: Emphasize word meaning via relationships in a lexical network.

3.2 Words and Grammatical Categories

Early definitions of words (Bloomfield, Sapir) were based on form/function. Defining “word” cross-linguistically is difficult.

Lexeme: The basic dictionary unit (e.g., “run” includes runs, ran, running).

3.3 Ambiguity, Polysemy, and Vagueness

The Zeugma test is used to reveal ambiguity (e.g., He broke her heart and the window).

Cruse (1986) was foundational in distinguishing:

  • Homonymy: Unrelated meanings (e.g., bat – animal vs. sports equipment).
  • Polysemy: Related senses (e.g., mouth of a river and mouth of a body).
  • Vagueness: Under-specified meaning (e.g., heap).

3.4 Lexical Relations

RelationExampleTheorists/Frameworks
Synonymybegin/start, buy/purchaseLyons (1968), Cruse (1986)
Antonymyhot/cold (gradable), dead/alive (complementary)Lyons (1977)
Hyponymysparrow < bird < animalCognitive hierarchy research (e.g., Rosch)
Meronymywheel is part of carCruse (1986)

Also includes: Reverses (enter/exit), Converses (buy/sell), and Taxonomic Sisters (cat/dog).

3.5 Derivational and Morphological Relations

  • Causatives: die – kill, open – open (something).
  • Agentive nouns: baker, singer derived from verbs.

See Levin & Rappaport Hovav, Aronoff & Fudeman (2005) for derivational morphology.

3.6 Lexical Typology (Cross-Linguistic)

  • Berlin & Kay (1969): Established the color term hierarchy.
  • Anna Wierzbicka & Cliff Goddard: Developed the Natural Semantic Metalanguage (NSM) using universal semantic primes (e.g., I, YOU, GOOD, BAD).
  • Swadesh List: A 100-word basic vocabulary used for language comparison.

3.7 Semantic Shift and Lexical Change

  • Grammaticalization: Lexical items become grammatical markers (e.g., have (possession) ⟶ have done (perfective)). Theorists: Kuteva & Heine (2008), Lehmann (2015).
  • Metaphor and Metonymy as mechanisms of change: grasp an idea (metaphoric extension); plastic = credit card (metonymy).
  • Conceptual Metaphor Theory: Lakoff & Johnson (1980).
  • Frame Semantics: Fillmore.

Chapter 4: Sentence Meaning and Logical Relations

4.1 Entailment, Presupposition, Contradiction

ConceptDefinitionKey Theorists
EntailmentIf A entails B, then B must be true if A is true.Strawson, Lyons, Frege
PresuppositionA speaker assumes background knowledge is true.Stalnaker (1974), Heim (1983)
ContradictionSentences where both cannot be true simultaneously.Standard in formal logic

John Austin and Paul Grice were foundational for speech act and contextual meaning. Sperber & Wilson (1995) developed Relevance Theory and pragmatic presupposition.

4.2 Sentence Relations and Truth

  • Truth-conditional semantics: Central to formal semantics (Fregean tradition). A.C. Grayling and Alfred Tarski focused on the correspondence theory of truth, where sentences are evaluated against world facts.
  • Counterfactuals: “If he had left, she would have cried.” David Lewis (1973) used possible worlds semantics. Kripke (1980) studied necessity and a posteriori truth.

4.3 Semantic vs. Pragmatic Account

  • Frege: Semantic presupposition is tied to logical form.
  • Heim (1983): Dynamic context-update model (Discourse Representation Theory).
  • Sperber & Wilson (1995): Pragmatics is driven by context and speaker assumptions.

4.4 Logical Tools for Sentence Meaning

Propositional logic and predicate logic are used to describe:

  • Tautologies (always true): It is what it is.
  • Contradictions (always false): It is raining and it is not raining.
  • Analytic truths (true by virtue of meaning): All bachelors are unmarried.

Analysis uses logical equivalence, truth tables, and inference rules.

4.5 Formal Semantics and Meaning

  • Montague Grammar: Combines logic and natural language syntax.
  • Kripke, Carnap, and Quine pioneered modal and intensional logic.

Semanticists distinguish between:

  • Necessary truths: True in all possible worlds (Leibniz).
  • Contingent truths: True in some, but not all, possible worlds.

Chapter 5: Sentence-Level Meaning: Time and Aspect

5.1 Overview: Sentence-Level Meaning

Focuses on how sentences characterize situations and events. Introduces three core semantic systems:

  1. Situation type (Aktionsart)
  2. Tense: Grammatical time
  3. Aspect: Internal temporal structure

Key Concept: Construal – The speaker’s semantic choice in framing an event (from Langacker, 1987 – Cognitive Grammar).

5.2 Situation Types (Aktionsart)

A. Static vs. Dynamic Situations

  • Stative: No internal phases or change (e.g., love, know, be).
  • Dynamic: Events with subparts or change (e.g., run, drive, grow).

Stative Verb Properties:

  • Do not occur in the progressive: *I am knowing Swahili.
  • Do not occur as imperatives: *Know Swahili!

Vlach (1981) noted that the progressive implies dynamism, incompatible with statives.

ILP vs. SLP Distinction:

  • Individual-Level Predicates (ILPs): Permanent traits (e.g., intelligent).
  • Stage-Level Predicates (SLPs): Temporary states (e.g., tired).

Theorists: Carlson (1977), Diesing (1992), Kratzer (1994).

B. Durative vs. Punctual

  • Durative: Extended in time (sleep, walk).
  • Punctual: Instantaneous (cough, blink).

Semelfactives are punctual and atelic (from Latin semel “once”). Theorists: C.S. Smith (1991), Verkuyl (1993).

C. Telic vs. Atelic

  • Telic: Has a natural endpoint (build a house).
  • Atelic: Has no endpoint (gaze at sea).

Contextual modification: Run (atelic) vs. Run the marathon (telic). Comrie (1976) compared German essen vs. aufessen. Verkuyl (1972) showed that the Object NP influences telicity. Rothstein (2004) studied compositional aspect from verb + argument structure.

D. Vendler’s (1967) Four-Way Classification

C.S. Smith (1991) added semelfactive as a fifth class.

TypeExamplesProperties
Statesknow, believe, love[+stative, +durative]
Activitiesrun, walk, push a cart[–stative, +durative, –telic]
Accomplishmentspaint a picture, walk to school[–stative, +durative, +telic]
Achievementsrecognize, spot, win[–stative, –durative, +telic]
Semelfactivesblink, cough, knock[–stative, –durative, –telic]

E. Diagnostic Tests for Situation Types

From Dowty (1979), Smith (1991):

Test TypeDistinguishesExample
Progressive formStates vs. events*I am knowing French
Imperative formStates vs. events*Know French!
Simple PresentHabitual vs. stativeShe knows Cannes vs. She visits Cannes
‘What happened was…’Eventivity*What happened was that she was tired
‘In/for X time’ adverbsTelic vs. atelicShe reached school in 1hr vs. *played cards in 1hr
‘Almost’ testAccomplishmentsJohn almost wrote a novel (ambiguous)

5.3 Tense (Deictic Time Anchoring)

Tense locates the event (E) relative to the time of speaking (S). It is a grammatical category across languages.

English Tense Examples:

  • Past: She spoke
  • Present: She speaks
  • Future: She will speak

Reichenbach (1947) Model: E = Event time, R = Reference time, S = Speech time.

TenseE, R, S Relation
Simple PastE = R < S
PresentE = R = S
FutureS < R = E
Present PerfectE < R = S
Past PerfectE < R < S
Future PerfectS < E < R

Givón (1972) noted rich tense systems (e.g., Chibemba with four pasts and four futures).

5.4 Aspect (Temporal Contour of Events)

“Aspect has to do, not with when something happens, but how it unfolds in time.” — Hockett (1958)

English Aspectual Forms:

  • Present Perfect: I have listened (Result/relevance to now)
  • Past Perfect: I had listened (Event before reference time)
  • Progressive: I am listening (Ongoing action)
  • Simple Aspect: I listen (Habitual/general truth)

Comrie (1976): Aspect is viewpoint.

  • Perfective: External view (complete).
  • Imperfective: Internal view (in progress).

Cross-linguistic Comparison: Russian čital = imperfective (in progress); pročital = perfective (complete). Smith (1991) defined Perfective as [begin+end] and Imperfective as [middle phase].

5.5 Modality and Mood

Modality

  • Epistemic: Degree of certainty (e.g., She must be home).
  • Deontic: Obligation/permission (e.g., You must leave).

Modal verbs (must, might, should, can, could) also signal abilitive, bouletic, and teleological modality.

Possible Worlds Semantics: David Lewis (1973, 1986) used modality to model hypothetical vs. actual worlds.

Mood

Grammatical mood includes Indicative, Subjunctive, Optative, and Imperative. Languages with morphologically marked mood include Somali (indicative, potential, conditional) and Ngiyambaa (realis vs. irrealis).

5.6 Evidentiality

Encodes the source of information: visual, auditory, reported, inferred, etc. Jakobson (1957) introduced evidentiality. Aikhenvald (2004) studied cross-linguistic evidential systems (e.g., Tariana).

5.7 Negation

Logical negation (¬p) interacts with natural language systems:

  • Syntax: She didn’t come, did she?
  • Morphology: French ne…pas.
  • Scope and focus: Only vs. not only.

Miestamo (2017) studied cross-linguistic negation systems.

Chapter 6: Sentence Semantics 2: Participants

6.1 Introduction: Classifying Participants

Focuses on how entities in events (participants) are characterized using thematic roles (also semantic roles or theta-roles).

Key Theorists:

  • Fillmore (1968): Case Grammar—first formalized semantic roles.
  • Jackendoff (1972, 1990): Introduced two-tier models of roles.
  • Gruber (1976): Proposed spatial metaphors for thematic roles.
  • Dowty (1991): Introduced Proto-Roles to resolve definitional fuzziness.

6.2 Basic Thematic Roles

RoleDescription & Example
AgentInitiator of action (David cooked the rashers)
PatientEntity affected/changed (The sun melted the ice)
ThemeMoved or located entity (Roberto passed the ball)
ExperiencerFeels or perceives something (Kevin felt ill)
InstrumentMeans used by agent (He opened it with a key)
GoalEndpoint of movement (She went to Dublin)
SourceOrigin of movement (He came from London)
LocationPlace of event (He sat in the room)
RecipientTarget of transfer (He gave her a gift)
BeneficiaryReceives benefit (He baked her a cake)

6.3 Grammatical Relations and Thematic Roles

Typical mapping:

  • Agent → Subject
  • Theme → Object
  • Instrument → Oblique (PP)

Languages may reorder these roles for emphasis, information structure, or grammar.

6.4 Thematic Role Grids (Theta-Grids)

Every verb has a set of required thematic roles, known as its lexical specification. Haegeman (1994) provided grammatical tests to determine argument status.

Example: The verb put requires <AGENT, THEME, LOCATION> → John put the book on the shelf.

6.5 Problems with Thematic Roles

  • Definitions are fuzzy (e.g., Theme vs. Patient).
  • Roles may overlap or be fulfilled simultaneously.

Jackendoff (1990) Two-Tier Model:

  • Action tier: AGENT, EXPERIENCER, PATIENT
  • Thematic tier: THEME, GOAL, LOCATION, etc.

Dowty (1991) Proto-Roles:

  • Proto-Agent: Volitional, causes change, sentient.
  • Proto-Patient: Undergoes change, affected by another.

6.6 The Purpose of Roles

Roles help explain:

  • Syntactic structure (e.g., passive/middle voice).
  • Verb alternations (e.g., causative-inchoative, dative shift).
  • Morphological agreement (e.g., noun class systems).
  • Language acquisition and computational semantics (e.g., role labeling in NLP systems: Palmer et al. 2005).

6.7 Causation and Role Alternation

Causative-Inchoative Alternation:

  • The water boiled (no agent)
  • Helen boiled the water (AGENT added)

Haspelmath (1993) studied the cross-linguistic prevalence of this alternation. Levin & Rappaport Hovav (1995, 2012) developed verb class theory. Some languages mark causatives morphologically (e.g., Somali: fur = open; fur-m-ay = opened (decausative)).

6.8 Voice Systems: Passive and Middle

Passive Voice

Changes the mapping between thematic roles and syntax, often promoting a non-agent to the subject position. Example: The book was read (by Mary) → THEME becomes subject. Levin (1993) and Kuno (1987) explored discourse motivation (e.g., empathy effects).

Middle Voice

The subject is the affected participant, with no explicit agent. Example: This book sells well.

6.9 Classifier and Noun Class Systems

Classifiers

Common in Asian languages (e.g., Chinese, Thai). They mark semantic properties (shape, animacy) and are lexical, not fully grammaticalized.

Noun Classes

Fully grammatical categories, widespread in Bantu languages. Aikhenvald (2000) noted Proto-Bantu classes:

  • Class 1/2: Humans
  • Class 5/6: Fruits, paired body parts
  • Class 14: Abstract qualities
FeatureNoun ClassesClassifiers
SizeSmall, closed setLarge, open set
RealizationGrammaticalLexical
ScopeWhole noun phraseLocal to noun

Chapter 7: Meaning and Context

7.1 Introduction: The Role of Context in Meaning

Sentence meaning is incomplete without context. Semantics is encoded meaning; Pragmatics is meaning-in-use, context-sensitive. Context affects reference, inference, deixis, implicature, and discourse structure.

Key Theorists:

  • Paul Grice (1975): Cooperative Principle, conversational implicature.
  • John Austin (1962): Speech Act Theory.
  • Stephen Levinson (1983, 2000): Deixis, discourse modeling.
  • Charles Fillmore (1982, 1985): Frame semantics.
  • Dan Sperber & Deirdre Wilson (1986): Relevance Theory.

7.2 Deixis

Deixis is context-dependent reference (“pointing language”).

TypeDescriptionExample
Person DeixisReference to speaker/addresseeI, you, he
Spatial DeixisReference to physical locationhere, there, above
Temporal DeixisReference to timenow, then, tomorrow
Discourse DeixisReference within the textthe following, as said
Social DeixisEncodes status/formalitytu/vous, usted

Cross-linguistic Examples: Spanish has a 3-way spatial system (aquí, ahí, allí). Hausa adds speaker/addressee anchoring (Jaggar 2001). Daga includes vertical deixis (Anderson & Keenan 1985). Japanese has complex social deixis (Kuno 1973).

7.3 Reference and Inference

Reference is the speaker’s use of language to point to entities, requiring contextual interpretation and background knowledge.

Phenomena:

  • Metonymy: The top floor rejected it. (“top floor” = management)
  • Synecdoche: He’s got two reds left. (“reds” = red snooker balls)
  • Ellipsis: Seen Mary today? (= Have you seen Mary today?)

Key Insight: Reference is often reconstructed via pragmatic inference (Clark 1978, Gibbs 1987).

7.4 Background Knowledge and Frames

Three layers of contextual knowledge:

  1. Physical Context: Speaker, hearer, place, time.
  2. Discourse Context: What has been said before.
  3. Background Knowledge: Cultural, encyclopedic, mutual assumptions.
  • Frame Semantics (Fillmore 1982): The “buy” frame includes buyer, seller, and goods.
  • Script Theory (Schank & Abelson 1977): Default event sequences (e.g., restaurant script).

Experimental Evidence: Anderson et al. (1977) showed interpretation changes with framing (“prison escape” vs. “wrestling match”).

7.5 Information Structure

Speakers package utterances based on what is shared versus new information.

ConceptFunction
TopicWhat the sentence is about.
FocusNew or contrastive information.
Given InformationAssumed known to the listener.
New InformationPresented as new or asserted.

Cross-linguistic Systems: English uses definite/indefinite articles and word order. Somali uses focus particles (baa). Japanese and Chinese use topic-comment structures. Key Theorists: Halliday (1967) (Theme-Rheme), Lambrecht (1994), Gundel et al. (1993) (Givenness Hierarchy).

7.6 Inference

Listeners use inference to connect meanings not made explicit.

  • Bridging Inference: The chandelier was impressive → must be in a hall. (Clark 1977)
  • Anaphora Resolution: “It really disappointed her” = the holiday.
  • Structural Ambiguity: “John chased the dog with the stick.” (Kess & Hoppe 1985)

7.7 Speech Acts

Speech acts are utterances that perform actions (Austin 1962).

Austin’s Three Levels (1962):

  • Locutionary Act: Producing a meaningful utterance.
  • Illocutionary Act: Speaker’s intention (e.g., request, assert).
  • Perlocutionary Act: Effect on the hearer (e.g., persuade).

Searle’s Classification (1969):

Act TypeExample
AssertiveI believe it’s raining
DirectiveClose the door
CommissiveI promise I’ll come
ExpressiveI’m sorry
DeclarativeI now pronounce you married

Felicity Conditions (Austin): Appropriate participants/context, accepted conventions, and speaker intention/sincerity are required for a speech act to succeed. Implicit performatives are common (e.g., “Can you tell me the time?” is a request, not a question).

Chapter 8: Meaning Components

8.1 Introduction to Semantic Components

Lexical items are analyzed as built from smaller semantic units (components or primitives). Componential Analysis (CA) posits elements like [ADULT], [HUMAN], [MALE], [UNMARRIED].

Example: woman = [FEMALE] [ADULT] [HUMAN]; bachelor = [MALE] [ADULT] [HUMAN] [UNMARRIED].

CA helps define lexical relations, syntactic behavior, and psychological structure.

Key Theorists:

  • Jerrold Katz (1963, 1972): Semantic metalanguage, dictionary model.
  • Ray Jackendoff (1972–2007): Conceptual structure, mentalist semantics.
  • James Pustejovsky (1995): Generative Lexicon theory.
  • Beth Levin & Malka Rappaport Hovav: Verb semantics and event decomposition.

8.2 Lexical Relations in Componential Analysis

Definitions using components:

  • Hyponymy: P is a hyponym of Q if all the features of Q are contained in P.
  • Incompatibility: Items share features but differ by one or more (e.g., gender, marital status).

Redundancy Rules: HUMAN → ANIMATE → CONCRETE. These rules allow for more economical dictionary entries by avoiding the repetition of entailed features.

8.3 Katz’s Semantic Theory

8.3.1 Basic Principles

Developed within generative grammar (Katz & Fodor 1963; Katz & Postal 1964). Meaning is compositional, combining word meaning and syntax via projection rules.

Components:

  • Dictionary with semantic markers (shared features).
  • Projection rules for compositional meaning.
  • Metalanguage of components.

8.3.2 Dictionary Example: Bachelor

The lexeme bachelor {N} has several readings:

  • (human)(male)[never married]
  • (human)(male)[young knight]
  • (human)[holds first degree]
  • (animal)(male)[young fur seal without mate]

Two semantic types: Semantic markers (general features) and Distinguishers (idiosyncratic meanings).

8.3.3 Projection Rules

These rules operate on syntactic trees to combine word meanings into phrase/sentence meaning. They enforce selectional restrictions (e.g., adjectives require appropriate noun types, like colorful ball).

8.4 Grammatical Rules and Semantic Components

Beth Levin’s Verb Class Analysis (1993): Verbs are grouped by behavior in constructions (middle, conative, body part ascension).

Verb class examples:

  • break verbs: change of state
  • cut verbs: cause + change + contact + motion
  • hit verbs: contact + motion
  • touch verbs: contact only

Key Components: CAUSE, CHANGE, MOTION, CONTACT.

Steven Pinker (1989): Only some components are grammatically relevant. Lexical rules operate only on the Grammatically Relevant Subsystem (e.g., butter = [CAUSE + BUTTER]).

8.5 Talmy’s Typology of Motion Events

Leonard Talmy (1985, 2000) identified components of motion events: FIGURE, GROUND, MOTION, PATH, MANNER. Typology is based on how these are conflated in verbs.

Typological Patterns:

  • Motion + Path (e.g., Romance languages)
  • Motion + Manner/Cause (e.g., English, Chinese)
  • Motion + Figure (e.g., Atsugewi)

Satellite-framed vs. Verb-framed languages:

  • Satellite-framed (e.g., English): Verb encodes manner, satellites encode path (She ran out of the house).
  • Verb-framed (e.g., Spanish): Verb encodes path, manner is outside (Salió de la casa corriendo).

Impact: Slobin (1994, 2004) showed that cross-linguistic narrative styles reflect motion encoding patterns.

8.6 Jackendoff’s Conceptual Semantics

Ray Jackendoff (1983, 1990, 1992) proposed that meaning is mental representation built from conceptual primitives: GO, BE, CAUSE, INCH, etc. Conceptual structure is hierarchical and compositional.

Example: Bill went into the house → [Event GO ([Thing BILL], [Path TO (IN HOUSE)])]

Localist Semantics: Spatial metaphors are extended to time, property, and possession (BELOC, BETEMP, BEIDENT, BEPOSS).

Cross-category generalization: The feature [±BOUNDED] applies to both nouns (mass nouns = [–b]) and events (telic verbs = [+b]).

8.7 Pustejovsky’s Generative Lexicon

James Pustejovsky (1995) proposed four levels of lexical representation:

  1. Argument structure: Mapping to syntax.
  2. Event structure: Internal event composition.
  3. Qualia structure: Four roles (Constitutive, Formal, Telic, Agentive).
  4. Lexical inheritance: Network of relations between words.

This model supports polysemy, coercion, and flexible word senses.

Chapter 10: Cognitive Semantics

10.1 Introduction: Cognitive Semantics and Functionalism

Cognitive semantics views linguistic knowledge as part of general cognition. It rejects the modularity of mind (Fodor, Chomsky) and opposes formalist models.

Functionalist view: Grammar is shaped by cognitive processes like attention, memory, and perception.

Key Theorists:

  • George Lakoff (1987, 1988), Mark Johnson (1987, 1999): Cognitive semantics.
  • Ronald Langacker (1987–2009): Cognitive Grammar.
  • Gilles Fauconnier & Mark Turner (2002): Blending theory.

10.2 Categorization and Embodiment

10.2.1 Classical Theory of Categories (Rejected)

The classical model (Aristotle) held that categories have necessary and sufficient features (e.g., bachelor = [+HUMAN] [+ADULT] [–MARRIED]). This model fails to explain fuzzy boundaries or prototype effects.

10.2.2 Prototype Theory & Embodiment

Categories have central members (prototypes) and peripheral ones, influenced by experience and embodied interaction.

  • George Lakoff (1987): Radial categories.
  • Eleanor Rosch (1978): Prototype effects.
  • Mark Johnson (1987): Image schemas (e.g., CONTAINER, PATH).

10.3 Polysemy

Words have multiple related meanings, not homonymy. Example: run (person, nose, computer, engine) — the central meaning is extended metaphorically or metonymically.

ICMs (Idealized Cognitive Models): Fillmore (1982) argued that interpretation depends on background cultural models (e.g., the bachelor example excludes priests, even if the definition fits).

Evans (2009) LCCM Theory: Lexical concepts link to a network of cognitive models (e.g., “France” activates political, geographic, or cultural models).

10.4 Metaphor

The Cognitive View holds that metaphor is central to thought, not just rhetoric.

Conceptual Metaphor Theory (Lakoff & Johnson 1980): Understanding one domain (Target Domain) in terms of another (Source Domain).

Examples:

  • LIFE IS A JOURNEY: birth = departure, death = arrival.
  • ARGUMENT IS WAR: “He attacked my claims.”

10.5 Meaning Construction and Context

Lexical meaning is underspecified and constructed in use. Example: book can refer to the physical object, the content, or the reading event. Evans (2009) argues that meaning is usage-based and context-sensitive.

10.6 Mental Spaces and Blending

Fauconnier (1994, 1997) Mental Space Theory: Mental spaces are temporary conceptual domains set up during discourse.

Blending Theory (Fauconnier & Turner 2002): Input spaces combine into a blended space with emergent structure through processes like composition, completion, and elaboration.

Examples:

  • France votes no → blends political and geographic models.
  • Kant debates contemporary philosopher → cross-time blend.

10.7 Langacker’s Cognitive Grammar

Core Assumptions:

  • Grammar is a symbolic system (meaning + form).
  • All structures are meaningful, including morphology and syntax.
  • Rejects the traditional split between lexicon and syntax.

Main Concepts:

  • Construal: Speaker’s perspective affects meaning.
  • Figure-Ground: Focus vs. background (cf. Talmy).
  • Profiling: Focusing on a part of a scene.
  • Scanning: Viewing events as sequential (verb) or summary (noun) (e.g., Keegan entered the room vs. Keegan’s entrance).

10.8 Construction Grammar

Developed by Charles Fillmore, Paul Kay, George Lakoff, Adele Goldberg, and Langacker.

Key Ideas:

  • Constructions: Pairings of form and meaning (not derived solely from syntax).
  • Apply to all levels (morphology, syntax, idioms, discourse).

Examples:

  • The X-er the Y-er: The bigger they are, the harder they fall.
  • Let alone construction: I barely got up in time to eat lunch, let alone cook breakfast.