- Atlas
- Humanities
- Syntax
Syntax
Syntax is the branch of linguistics that studies how words combine into phrases and sentences.
20 frameworks, from 750 to today · Linguistics · Humanities
- Arabic Grammatical Tradition 750-1300
- Universal Grammar (Port-Royal) 1660-1800
- Historical-Comparative Syntax 1900-Present
- Distributionalism 1933-1960
- Transformational grammar 1957-1980
- Systemic functional grammar 1961-Present
- Linguistic Typology 1963-Present
- Relational Grammar 1972-1990
- Tree-Adjoining Grammar 1975-2000
- Functional Grammar (Dik) 1978-2000
- Generalized Phrase Structure Grammar 1978-1990
- Lexical-Functional Grammar 1978-Present
- Government and Binding Theory 1981-1995
- Government and Binding Theory 1981-1993
- Role and Reference Grammar 1984-Present
- Construction Grammar 1985-Present
- Head-Driven Phrase Structure Grammar 1985-Present
- Role and Reference Grammar 1985-Present
- Minimalist Program 1993-Present
- Minimalist Program 1995-Present
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
- Syntax (introduction, Generative syntax), Wikipedia
- Lectures on Government and Binding (Background, introduction), Wikipedia
- Transformational grammar, Wikipedia
- Generative grammar (History), Wikipedia
- Government and binding theory, Wikipedia
- Principles and parameters, Wikipedia
- Generalized phrase structure grammar (Origins), Wikipedia
- 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.
- 750-1300
Arabic Grammatical Tradition
How do case endings mark grammatical roles?
- 1660-1800
Universal Grammar (Port-Royal)
What logical structure underlies all languages?
- 1900-Present
Historical-Comparative Syntax
How have sentence structures changed across language history?
- 1933-1960
- 1957-1980
- 1961-Present
Systemic functional grammar
How does communicative function shape sentence structure?
- 1963-Present
- 1972-1990
Relational Grammar
How do grammatical relations like subject and object work?
- 1975-2000
Tree-Adjoining Grammar
How can tree-rewriting capture long-distance dependencies?
- 1978-2000
Functional Grammar (Dik)
How do communicative functions determine sentence form?
- 1978-1990
Generalized Phrase Structure Grammar
Can phrase-structure rules handle all syntax without movement?
- 1978-Present
Lexical-Functional Grammar
How can surface order and grammatical function be separated?
- 1981-1995
Government and Binding Theory
What universal principles and parameters govern grammar?
- 1981-1993
Government and Binding Theory
What universal principles and parameters govern grammar?
- 1984-Present
Role and Reference Grammar
How do syntax, semantics, and pragmatics combine in one representation?
- 1985-Present
- 1985-Present
Head-Driven Phrase Structure Grammar
How do feature structures and heads determine phrases?
- 1985-Present
Role and Reference Grammar
How do syntax, semantics, and pragmatics combine in one representation?
- 1993-Present
- 1995-Present
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.
Transformational grammar grew out of Distributionalism
Chomsky built transformational grammar on the distributional and immediate-constituent methods of Harris and Bloomfield, then added transformations to relate sentence types that phrase-structure analysis could not.
Carried by Noam Chomsky; Syntactic Structures (1957); Zellig Harris.
Generalized Phrase Structure Grammar reacted against Transformational grammar
GPSG was developed by Gazdar and colleagues as a direct alternative to transformational grammar, arguing that a rich phrase-structure grammar could handle unbounded dependencies without movement rules.
Carried by Gerald Gazdar; Ewan Klein; Geoffrey Pullum.
Lexical-Functional Grammar reacted against Transformational grammar
Lexical-Functional Grammar was developed by Bresnan and Kaplan as an alternative to transformational grammar, separating constituent structure from functional structure to handle free word order without movement.
Carried by Joan Bresnan; Ronald Kaplan.
Government and Binding Theory grew out of Transformational grammar
Government and Binding Theory was Chomsky's revision of transformational grammar, replacing specific rules with principles and parameters while keeping movement and traces.
Carried by Noam Chomsky; Lectures on Government and Binding (1981).
Head-Driven Phrase Structure Grammar grew out of Generalized Phrase Structure Grammar
HPSG grew out of GPSG, extending its constraint-based, feature-passing approach into a fully lexicalist theory with feature structures and head-driven principles.
Carried by Carl Pollard; Ivan Sag.
Minimalist Program grew out of Government and Binding Theory
The Minimalist Program was Chomsky's reworking of Government and Binding Theory, reducing its machinery to operations like Merge and Move and asking why the language faculty takes that form.
Carried by Noam Chomsky; The Minimalist Program (1995).
Transformational grammar revived Universal Grammar (Port-Royal)
Chomsky's transformational grammar revived the Port-Royal idea of a universal grammar grounded in the mind, treating grammar as an expression of thought rather than convention.
Carried by Noam Chomsky.
Relational Grammar reacted against Transformational grammar
Relational Grammar, associated with Perlmutter and Postal, rejected the transformational emphasis on constituent structure and made grammatical relations the primitive objects of syntax.
Carried by David Perlmutter; Paul Postal.
Functional Grammar (Dik) reacted against Transformational grammar
Dik's Functional Grammar was developed as a functional alternative to transformational-generative syntax, treating sentence structure as determined by communicative functions.
Carried by Simon Dik.
Construction Grammar reacted against Transformational grammar
Construction Grammar arose in opposition to the generative view that general rules derive all sentence patterns, arguing instead that constructions are learned form-meaning pairings.
Carried by Charles Fillmore; Paul Kay; Adele Goldberg.
Role and Reference Grammar influenced by Linguistic Typology
Role and Reference Grammar drew heavily on cross-linguistic typological data, incorporating typological findings into its syntactic representations.
Carried by Robert Van Valin.
Lexical-Functional Grammar rival of Generalized Phrase Structure Grammar
GPSG and LFG were contemporary constraint-based alternatives to transformational grammar, competing to explain unbounded dependencies and free word order without movement rules.
Carried by Gerald Gazdar; Joan Bresnan.
Linguistic Typology influenced by Transformational grammar
Linguistic Typology, especially Greenberg's work, engaged with generative grammar's search for universals, though it emphasized cross-linguistic variation and empirical classification.
Carried by Joseph Greenberg.
Construction Grammar influenced by Linguistic Typology
Construction Grammar was influenced by typological work showing that constructions vary across languages and must be learned as form-meaning pairings.
Carried by Adele Goldberg.
Tree-Adjoining Grammar rival of Generalized Phrase Structure Grammar
GPSG and Tree-Adjoining Grammar were contemporary formal frameworks competing to handle long-distance dependencies without transformational movement.
Carried by Gerald Gazdar; Aravind Joshi.
Head-Driven Phrase Structure Grammar rival of Lexical-Functional Grammar
LFG and HPSG were contemporary lexicalist, constraint-based frameworks competing to describe syntax without transformations.
Carried by Joan Bresnan; Carl Pollard.
Beyond the field
Where Syntax's ideas came from in other fields, and what they led to there: the strongest 12 of 81 links.
Arabic Grammatical Tradition influenced Khalilian Arud in Poetics And Prosody
The Arabic grammatical tradition, with its analysis of roots, patterns and syllable structure, supplied al-Khalil ibn Ahmad with the morphological and phonological categories he reorganized into the arud system of meters. The prosody was built directly on the grammarians' descriptive apparatus.
Carried by Al-Khalil ibn Ahmad.
Historical-Comparative Syntax grew out of Comparative Method in Linguistics
Historical-comparative syntax is the application of the comparative method established by Jones and Bopp to syntactic rather than phonological and morphological data.
Carried by William Jones; Franz Bopp.
Historical-Comparative Syntax grew out of Comparative Method in Historical Linguistics
Historical-comparative syntax extended the comparative method of nineteenth-century Indo-European linguistics, applying its techniques of correspondence and reconstruction from phonology and morphology to sentence structure.
Distributionalism influenced Distributional Semantics in Semantics
Distributional semantics revives and formalises the distributional hypothesis of Harris and Firth, that words occurring in similar contexts have similar meanings. Harris's distributional method in structural linguistics supplied the core assumption that co-occurrence patterns carry meaning, which computational distributional models then operationalised with vector spaces.
Carried by Zellig Harris; John Rupert Firth.
Distributionalism grew out of Structural linguistics in Linguistics
Distributionalism is the American branch of structural linguistics, deriving from Saussure's insistence on studying relations among units and from Bloomfield's behaviourist version of that programme.
Carried by Ferdinand de Saussure; Leonard Bloomfield.
Transformational grammar led to Universal grammar in First Language Acquisition
Transformational grammar was Chomsky's earlier formulation, with deep and surface structure and transformations, out of which universal grammar as an acquisition theory grew. The principles-and-parameters turn of the late 1970s refined this into the learnability framework.
Carried by Noam Chomsky; Aspects of the Theory of Syntax (1965).
Transformational grammar led to Principles and parameters in Linguistic Typology
Principles and parameters developed out of transformational grammar, retaining transformations while recasting them as the interaction of general principles and parameters. The government and binding framework was the direct successor to the transformational model.
Carried by Noam Chomsky.
Transformational grammar led to Generative phonology in Phonology
Generative phonology was built as the phonological component of Chomsky's transformational grammar, taking over its rule formalism, underlying-to-surface derivations and the competence/performance distinction. The Sound Pattern of English applies transformational rule notation to sound structure.
Carried by Noam Chomsky; Morris Halle; The Sound Pattern of English (1968).
Systemic functional grammar reacted against Generative Grammar in Linguistics
Halliday's systemic functional grammar was framed against Chomskyan generative grammar, rejecting its autonomy of syntax and its idealised competence in favour of meaning and social function.
Carried by M. A. K. Halliday.
Linguistic Typology led to Quantitative and Phylogenetic Typology in Linguistic Typology
Quantitative and phylogenetic typology is a methodological outgrowth of linguistic typology: it keeps the typological project of comparing languages worldwide but replaces impressionistic sampling with databases and statistical inference. The World Atlas of Language Structures, published in 2005, supplied the coded cross-linguistic data that made this quantitative turn possible.
Carried by World Atlas of Language Structures (2005).
Linguistic Typology led to Sign Language Typology in Sign Language Linguistics
Sign language typology grew directly out of general linguistic typology, applying its comparative method and cross-linguistic categories to sign languages. Typologists extended the Greenbergian programme to a new modality, testing whether typological universals hold for signed languages.
Carried by Joseph Greenberg.
Linguistic Typology led to Typology-Informed Reconstruction in Historical Linguistics
Typology-informed reconstruction grew out of linguistic typology: Greenberg's implicational universals and word-order correlations were taken up by historical linguists as constraints on what a reconstructed proto-language could plausibly have been.
Carried by Joseph Greenberg.

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 SyntaxRelated fields
Fields its ideas are linked to
- Linguistics 26 links
- First Language Acquisition 8 links
- Cognitive Science 6 links
- Historical Linguistics 4 links
- Language Documentation And Revitalization 4 links
- Linguistic Typology 4 links
- Morphology 4 links
- Corpus Linguistics 3 links
- Automata 3 links
- Psycholinguistics 2 links
- Semantics 2 links
- Poetics And Prosody 2 links
More in Linguistics
- Corpus Linguistics
- Dialectology
- Discourse Analysis
- First Language Acquisition
- Historical Linguistics
- Language Documentation And Revitalization
- Lexicography
- Linguistic Typology
- Linguistics
- Morphology
- Neurolinguistics
- Phonetics
- Phonology
- Pragmatics
- Psycholinguistics
- Second Language Acquisition
- Semantics
- Sign Language Linguistics
- Sociolinguistics
- Writing Systems