An analysis of syntax as direct links between words — each word depends on one head.
In precise terms
A representation in which syntactic structure is a set of asymmetric head-dependent relations between words, with no intermediate phrasal nodes. Widely used in computational parsing and in cross-linguistic annotation projects.
An example
The child read the book.
What is going on here
'Read' is the root; 'child' and 'book' depend on it; 'the' depends on each noun. No phrasal nodes are posited — the structure is entirely relations between words. This is the format most modern parsers use.
Why this matters when you are learning a language
Dependency annotation is what makes large multilingual treebanks comparable, and it is the format behind most of the syntactic analysis tools a learner or researcher will actually encounter.
Across languages
Dependency representations handle free-word-order languages more comfortably than phrase structure, because they do not require contiguous constituents — an advantage for languages like Czech, Latin and Warlpiri.
Related terms
- Constituent structureAdvanced — The hierarchical grouping of words into nested phrases.
- Phrase and constituentIntermediate — A group of words that behaves as a single unit inside a sentence.
- Word orderIntermediate — The order in which subject, verb and object normally appear.
- Linguistic typologyAdvanced — Classifying languages by their structural features rather than their family history.
Where to go next
- Language guides — See grammatical and lexical categories used in context.
- Terminology quiz — Test the terms on this page.