Syntax

Syntax is the branch of linguistics that studies how words combine into phrases and sentences.

20 frameworks, from 750 to today · Linguistics · Humanities

still a working view ended · each leads to its entry below

Overview

Syntax is the branch of linguistics that studies how words combine into phrases and sentences1. Its central puzzle is that sentence structure is not just a matter of word order or meaning: Colorless green ideas sleep furiously is grammatical nonsense, while Ideas colorless sleep furiously green is word salad with the same words. Speakers of a language share tacit knowledge of which combinations are well-formed, how words group into larger units, and how those units move or change shape across sentence types. Syntax asks what that knowledge consists of, how it is acquired, and why human languages organize sentences the way they do. The history of the field is a history of competing answers to those questions, each one built on what the previous framework could not explain.

Grammar Before Linguistics: The Arabic and Universal Traditions

Systematic sentence analysis is older than linguistics as a university discipline. The Arabic grammatical tradition, emerging in Basra and Kufa from the eighth century onward, produced the first fully explicit theory of sentence structure. Grammarians such as Sibawayh, whose al-Kitāb (late eighth century) is the foundational text, analyzed Arabic in terms of iʿrāb—the case and mood endings that mark a word's grammatical role—and developed the distinction between the nominal sentence (jumla ismiyya), beginning with a noun, and the verbal sentence (jumla fiʿliyya), beginning with a verb. Their concept of ʿamal, "government," held that certain words (verbs, particles, prepositions) assign case to others, a notion that would reappear in European grammar centuries later. The tradition was pedagogical and normative as much as descriptive, but its categories—subject, object, predicate, governing element—shaped how sentences were understood across the Islamic world and, through contact, in medieval Europe.

In Europe, the tradition of universal grammar grew from the Port-Royal Grammaire générale et raisonnée (1660) of Antoine Arnauld and Claude Lancelot, which argued that beneath the surface differences among languages lies a shared logical structure of subject, copula, and predicate. This rationalist program treated grammar as an expression of thought, not as an arbitrary convention. It dominated European grammatical thinking into the eighteenth century and, in a different form, would be revived by Noam Chomsky in the twentieth.

Structuralism and the Constituent

Modern scientific syntax begins with the American structuralists of the early twentieth century. Leonard Bloomfield's Language (1933) set out to describe language without appeal to meaning or mental states, using distributional methods: a unit is defined by the environments in which it occurs. The key achievement was constituent analysis—the demonstration that sentences are not flat strings of words but hierarchical groupings. In Poor John ran away, the words group as Poor John and ran away, and each of those groups behaves as a unit (it can be replaced, moved, or coordinated). Structuralists such as Zellig Harris, Rulon Wells, and Charles Hockett developed substitution tests and immediate-constituent analysis to make these groupings explicit, often represented as tree diagrams. The method worked well for describing the phrase structure of a language, but it had no way to relate sentences that were obviously connected—active and passive, statement and question—except by listing them separately.

Transformational-Generative Grammar

Chomsky's Syntactic Structures (1957) broke with that limitation123. His proposal was that a grammar should generate all and only the grammatical sentences of a language through a system of rules, and that some sentences are derived from others by transformations—operations that rearrange, delete, or insert constituents13. The passive The ball was hit by John could be derived from the active John hit the ball by a rule that moves the object to subject position and inserts be and by34. This explained the systematic relationship between sentence types and, crucially, captured generalizations that phrase-structure rules alone could not.

The framework was revised repeatedly. Aspects of the Theory of Syntax (1965) introduced the distinction between deep structure (the underlying representation that determines meaning) and surface structure (the form that is pronounced), with transformations mapping one to the other2. The 1970s brought the "extended standard theory" and then Government and Binding Theory (GB), laid out in Lectures on Government and Binding (1981)152. GB replaced the specific rule list with a small set of general principles and parameters: universal principles hold across all languages, while a finite set of parameters (such as whether a language allows null subjects) accounts for cross-linguistic variation56. The theory introduced a modular architecture—X-bar theory for phrase structure, binding theory for pronouns and anaphors, case theory, theta theory for argument roles—and became the dominant framework of the 1980s. Its signature analyses, such as the trace left by a moved constituent (What did you see ___?), made movement a central object of study.

Rival Formal Frameworks

Not everyone accepted transformations. Generalized Phrase Structure Grammar (GPSG), developed by Gerald Gazdar, Ewan Klein, Geoffrey Pullum, and Ivan Sag in the late 1970s and 1980s, argued that a sufficiently rich phrase-structure grammar could capture the same generalizations without movement rules7. GPSG used metarules and feature-passing conventions to relate constructions, and its treatment of unbounded dependencies (as in Who did you say that Mary saw?) showed that a constraint-based approach could handle long-distance relations. Head-Driven Phrase Structure Grammar (HPSG), developed by Carl Pollard and Ivan Sag in the 1980s and 1990s, extended this into a fully lexicalist, constraint-based theory in which grammatical information is encoded in feature structures and the head of a phrase determines its properties. Lexical-Functional Grammar (LFG), developed by Joan Bresnan and Ronald Kaplan in the late 1970s, separated constituent structure (the surface arrangement of words) from functional structure (grammatical relations such as subject and object), allowing languages with free word order to be described without movement. Tree-Adjoining Grammar (TAG), developed by Aravind Joshi in the 1970s, used tree-rewriting operations rather than string rewriting, and became influential in computational linguistics. Relational Grammar, associated with David Perlmutter and Paul Postal in the 1970s and 1980s, made grammatical relations themselves the primitive objects of syntax, analyzing passives and other constructions as changes in relational structure. Role and Reference Grammar, developed by Robert Van Valin from the 1980s, combined syntactic, semantic, and pragmatic information in a single representation and drew heavily on cross-linguistic data.

The Minimalist Turn

In the 1990s Chomsky's program shifted again. The Minimalist Program, launched in The Minimalist Program (1995), asked not what the rules of grammar are but why they should take the form they do, given that language is a biological capacity. It proposed that the language faculty is an optimal solution to the interface conditions imposed by the sound and meaning systems, and it reduced the machinery of GB to a few operations—most importantly Merge, which combines two syntactic objects into a larger one, and Move, which repositions a constituent (later reconceived as internal Merge). The program remains the dominant framework in generative syntax, though its specific proposals are debated and revised continuously.

Functional and Usage-Based Approaches

A parallel tradition has always insisted that syntax cannot be understood apart from meaning, communication, and use. Functional Grammar, associated with Simon Dik (1978 onward) and later with Michael Halliday's systemic functional linguistics, treats sentence structure as shaped by the functions it serves—topic, focus, given and new information—rather than by autonomous formal rules. Construction Grammar, developed by Charles Fillmore, Paul Kay, and others from the 1980s and extended by Adele Goldberg and others, holds that the basic units of grammar are constructions: form-meaning pairings that range from single morphemes to whole sentence patterns. The let alone construction (He can't read Latin, let alone Greek) or the what's X doing Y? construction (What's that fly doing in my soup?) are learned as units with their own meanings, not derived from general rules8. Construction grammar has become a major framework, especially in cognitive linguistics, and overlaps with usage-based approaches that emphasize frequency and entrenchment.

Typology and Universals

Running alongside these theoretical debates is the empirical tradition of linguistic typology, which compares languages to discover what is universal and what varies. Joseph Greenberg's 1963 paper on word-order universals established the method: survey a wide sample of languages and look for implicational generalizations, such as "if a language has verb-final order, it tends to have postpositions." Typology has repeatedly challenged formal theories to account for the full range of attested structures, and it has supplied the cross-linguistic evidence on which claims about universal grammar rest. Historical-comparative syntax, meanwhile, reconstructs the sentence patterns of unattested ancestor languages—for example, the word order of Proto-Indo-European—by comparing attested daughter languages, extending the comparative method from sounds and words to syntax.

The Present

Syntax today is a plural field. Generative grammar, in its minimalist form, remains the largest single research program, with active work on locality, agreement, case, and the syntax-semantics interface. Constraint-based frameworks—HPSG, LFG, and construction grammar—continue to develop, often with computational implementations. Typology and fieldwork supply the empirical base, and experimental methods (acceptability judgments, eye-tracking, neurolinguistics) increasingly test theoretical claims. The open questions are old ones in new form: how much of syntax is innate and how much is learned from usage; how syntactic structure interfaces with meaning and sound; and why human languages, despite their diversity, share so much deep structure.

References

  1. Syntax (introduction, Generative syntax), Wikipedia
  2. Lectures on Government and Binding (Background, introduction), Wikipedia
  3. Transformational grammar, Wikipedia
  4. Generative grammar (History), Wikipedia
  5. Government and binding theory, Wikipedia
  6. Principles and parameters, Wikipedia
  7. Generalized phrase structure grammar (Origins), Wikipedia
  8. Construction grammar (History), Wikipedia

The frameworks of Syntax

Each framework is a way of working the field has used: when it held, and the question it answers.

  1. 750-1300

    Arabic Grammatical Tradition

    How do case endings mark grammatical roles?

    Fly to its tower in Nousfarer

  2. 1660-1800

    Universal Grammar (Port-Royal)

    What logical structure underlies all languages?

    Fly to its tower in Nousfarer

  3. 1900-Present

    Historical-Comparative Syntax

    How have sentence structures changed across language history?

    Fly to its tower in Nousfarer

  4. 1933-1960

    Distributionalism

    How do words group into hierarchical constituents?

    Fly to its tower in Nousfarer

  5. 1957-1980

    Transformational grammar

    What rules generate all grammatical sentences?

    Fly to its tower in Nousfarer

  6. 1961-Present

    Systemic functional grammar

    How does communicative function shape sentence structure?

    Fly to its tower in Nousfarer

  7. 1963-Present

    Linguistic Typology

    What sentence patterns are shared across languages?

    Fly to its tower in Nousfarer

  8. 1972-1990

    Relational Grammar

    How do grammatical relations like subject and object work?

    Fly to its tower in Nousfarer

  9. 1975-2000

    Tree-Adjoining Grammar

    How can tree-rewriting capture long-distance dependencies?

    Fly to its tower in Nousfarer

  10. 1978-2000

    Functional Grammar (Dik)

    How do communicative functions determine sentence form?

    Fly to its tower in Nousfarer

  11. 1978-1990

    Generalized Phrase Structure Grammar

    Can phrase-structure rules handle all syntax without movement?

    Fly to its tower in Nousfarer

  12. 1978-Present

    Lexical-Functional Grammar

    How can surface order and grammatical function be separated?

    Fly to its tower in Nousfarer

  13. 1981-1995

    Government and Binding Theory

    What universal principles and parameters govern grammar?

    Fly to its tower in Nousfarer

  14. 1981-1993

    Government and Binding Theory

    What universal principles and parameters govern grammar?

    Fly to its tower in Nousfarer

  15. 1984-Present

    Role and Reference Grammar

    How do syntax, semantics, and pragmatics combine in one representation?

    Fly to its tower in Nousfarer

  16. 1985-Present

    Construction Grammar

    What form-meaning pairings make up grammar?

    Fly to its tower in Nousfarer

  17. 1985-Present

    Head-Driven Phrase Structure Grammar

    How do feature structures and heads determine phrases?

    Fly to its tower in Nousfarer

  18. 1985-Present

    Role and Reference Grammar

    How do syntax, semantics, and pragmatics combine in one representation?

    Fly to its tower in Nousfarer

  19. 1993-Present

    Minimalist Program

    Why does the language faculty take its optimal form?

    Fly to its tower in Nousfarer

  20. 1995-Present

    Minimalist Program

    Why does the language faculty take its optimal form?

    Fly to its tower in Nousfarer

How the ideas connect

Each arrow runs from an earlier idea to the later one it shaped. The list below says what passed between them.

Arabic Grammatical TraditionUniversal Grammar (Port-Royal)Historical-Comparative SyntaxDistributionalismTransformational grammarSystemic functional grammarLinguistic TypologyRelational GrammarTree-Adjoining GrammarFunctional Grammar (Dik)Generalized Phrase Structure GrammarLexical-Functional GrammarGovernment and Binding TheoryGovernment and Binding TheoryRole and Reference GrammarConstruction GrammarHead-Driven Phrase Structure GrammarRole and Reference GrammarMinimalist ProgramMinimalist Program

Beyond the field

Where Syntax's ideas came from in other fields, and what they led to there: the strongest 12 of 81 links.

A city in Nousfarer: a red-carpeted stair climbing terraces to a crowned tower on an acropolis, more towers round it, forest and snowy mountains beyond.
A city in Nousfarer, the game played in the browser.

Syntax is a planet in Nousfarer

Its city has a tower for each of these 20 frameworks. Land, walk in, and climb one concept at a time.

Land on Syntax

Related fields

Fields its ideas are linked to

More in Linguistics

See where Syntax sits on the Map of Human Knowledge