Semantics Pragmatics

How to read this page: This article maps the topic from beginner to expert across six levels � Remembering, Understanding, Applying, Analyzing, Evaluating, and Creating. Scan the headings to see the full scope, then read from wherever your knowledge starts to feel uncertain. Learn more about how BloomWiki works ?

Semantics and Pragmatics are the branches of linguistics that deal with meaning. Semantics is the study of the literal, "dictionary" meaning of words and sentences, while Pragmatics is the study of how context, social norms, and speaker intentions influence meaning. In other words, semantics tells you *what* is said, and pragmatics tells you *what is meant*. Understanding this distinction is crucial for human communication, as we often say one thing to mean another (e.g., sarcasm, politeness, or indirect requests).

Remembering

• Semantics — The study of literal meaning in language.
• Pragmatics — The study of meaning in context.
• Lexical Semantics — The study of individual word meanings (e.g., synonyms, antonyms).
• Compositionality — The principle that the meaning of a sentence is determined by its parts and the way they are combined.
• Denotation — The literal, "point-to" meaning of a word (e.g., "snake" = a legless reptile).
• Connotation — The emotional or cultural associations of a word (e.g., "snake" = a treacherous person).
• Sense — The mental representation of a word's meaning.
• Reference — The actual object in the world that a word refers to.
• Speech Act — An utterance that performs an action (e.g., promising, ordering, greeting).
• Implicature — What is suggested in an utterance even though not expressed literally (H.P. Grice).
• Cooperative Principle — The assumption that participants in a conversation are trying to be helpful and relevant.
• Deixis — Words whose meaning depends entirely on context (e.g., "here", "now", "you").
• Entailment — A relationship between sentences where if sentence A is true, sentence B *must* be true.
• Presupposition — An underlying assumption that must be true for an utterance to make sense (e.g., "Have you stopped smoking?" presupposes you used to smoke).

Understanding

Meaning is a layered phenomenon.

Literal Meaning (Semantics): Semantics looks at relationships like:

• Synonymy (Big/Large)
• Antonymy (Hot/Cold)
• Hyponymy (Dog is a hyponym of Animal)
• Homonymy (Bank/Bank - river vs money)

It also explores "Truth Conditions"—what has to be true in the world for a sentence like "The cat is on the mat" to be true.

Contextual Meaning (Pragmatics): Pragmatics explains how we "read between the lines."

• Gricean Maxims: H.P. Grice argued that we follow four rules: Be truthful (Quality), Be informative (Quantity), Be relevant (Relation), and Be clear (Manner). When someone "flouts" a maxim (e.g., answering "What's for dinner?" with "Food"), they are creating an implicature (e.g., "I don't know yet" or "I'm annoyed").
• Speech Act Theory: J.L. Austin noted that we use language to *do* things. Saying "I do" at a wedding isn't just a description; it *is* the act of getting married. This is a "performative" utterance.

Applying

Modeling Gricean Implicature: <syntaxhighlight lang="python"> def decode_meaning(literal_text, context_relevance):

   """
   Demonstrates the difference between semantics and pragmatics.
   """
   # Literal Semantic Meaning
   semantic_meaning = literal_text
   
   # Pragmatic Interpretation
   if literal_text == "Can you pass the salt?" and context_relevance == "at_dinner":
       pragmatic_meaning = "Request: Give me the salt."
   elif literal_text == "It's cold in here" and context_relevance == "near_window":
       pragmatic_meaning = "Request: Please close the window."
   else:
       pragmatic_meaning = semantic_meaning
       
   return {
       "Literal (Semantics)": semantic_meaning,
       "Intended (Pragmatics)": pragmatic_meaning
   }

print(decode_meaning("Can you pass the salt?", "at_dinner"))

1. A computer without pragmatics would just answer 'Yes, I am capable.'

</syntaxhighlight>

Communication Failures

Semantic Ambiguity → "He saw the man with the binoculars" (Who has them?).

Pragmatic Failure → Failing to understand sarcasm or social cues (common in AI and some neurodivergent conditions).

Cross-Cultural Pragmatics → Different cultures have different rules for politeness, directness, and "small talk," leading to misunderstanding.

Presupposition Failure → "The King of France is bald" (but there is no King of France).

Analyzing

Semantics vs. Pragmatics

Feature	Semantics	Pragmatics
Focus	Linguistic meaning	Speaker meaning
Context	Independent of context	Highly context-dependent
Source	The Code (Mental Lexicon)	Social reasoning (Theory of Mind)
Question	What does this word mean?	What does the speaker mean by this word?

The Principle of Compositionality: If we know the meaning of "red" and the meaning of "apple," we know the meaning of "red apple." However, pragmatics often overrides this. A "white wine" isn't literally white (it's yellowish), and a "red potato" isn't literally red. Our knowledge of the world (encyclopedic knowledge) interacts with semantics to produce meaning.

Evaluating

Evaluating meaning theories: (1) Ambiguity Resolution: Can the theory explain how we choose the right meaning of "bank" in a sentence? (2) Universal Applicability: Are Grice's Maxims universal, or do they only describe Western communication styles? (3) Truth vs. Use: Is meaning primarily about truth conditions (Logic-based) or about social use (Usage-based)? (4) Computational Modeling: Can we build "Pragmatically-aware" AI that understands sarcasm and irony?

Creating

Future Directions: (1) Distributional Semantics: Using AI to map word meanings based on the "company they keep" in massive text datasets (e.g., Word2Vec). (2) Formal Pragmatics: Using game theory and Bayesian modeling to mathematically predict how people calculate implicatures. (3) The Semantics of Fiction: Exploring how we assign meaning and truth to statements about non-existent things (like dragons or unicorns). (4) Pragmatics in Translation: Developing machine translation that translates the "intent" and "tone" of a message, not just the literal words.