6.1 Constituents and phrase structure rules

Sentence Structure and Constituents

Sentences aren't just strings of words. Words group together into meaningful units called constituents, and understanding how those groupings work is the key to understanding sentence structure. This section covers how to identify constituents, the rules that describe how phrases are built, and the roles those phrases play in a sentence.

Constituents in Sentences

A constituent is a word or group of words that functions as a single unit within a sentence. Noun phrases, verb phrases, and prepositional phrases are all examples of constituents. The sentence The big red ball rolled under the fence contains several constituents nested inside each other, not just a flat sequence of eight words.

How do you prove that a group of words is actually a constituent? There are four standard constituency tests:

1. Replacement — Substitute the group with a single word (usually a pronoun). If the sentence still works, it's a constituent. The big red ball → It rolled under the fence.
2. Movement — Move the group to a different position in the sentence. On the table, I placed the book shows that on the table is a constituent because it can be fronted.
3. Cleft — Place the group into an "It is/was ... that" frame. It was the red car that I bought confirms the red car is a constituent.
4. Coordination — Join the group with another similar group using a conjunction. I like [apples] and [oranges] shows both are constituents of the same type.

The most common types of constituents you'll encounter:

• Noun phrases (NP) — express entities or concepts (the tall building, she)
• Verb phrases (VP) — describe actions or states (is running quickly, gave her the letter)
• Prepositional phrases (PP) — indicate spatial, temporal, or other relationships (under the bridge)
• Adjective phrases (AdjP) — modify nouns (extremely happy)
• Adverb phrases (AdvP) — modify verbs, adjectives, or other adverbs (very quickly)

Phrase Structure Rules

Phrase structure rules are formal notation for how sentences and phrases are built. They follow the format $\text{Category} \rightarrow \text{Constituent(s)}$, where the arrow means "consists of."

Here are the basic rules for English:

• $S \rightarrow NP \; VP$ — A sentence consists of a noun phrase and a verb phrase.
• $NP \rightarrow (Det) \; (AdjP) \; N \; (PP)$ — A noun phrase has a noun as its core, with an optional determiner, adjective phrase, and prepositional phrase. Parentheses mean "optional."
• $VP \rightarrow V \; (NP) \; (PP)$ — A verb phrase has a verb, with an optional noun phrase (object) and prepositional phrase.
• $PP \rightarrow P \; NP$ — A prepositional phrase contains a preposition followed by a noun phrase.

One powerful feature of these rules is recursion: a rule can refer back to the same category it's building. For example, $NP \rightarrow NP \; PP$ lets you stack prepositional phrases to create complex noun phrases like the book [on the shelf] [in the library]. Each PP attaches to the NP, and each PP itself contains another NP.

Tree diagrams are the visual version of these rules. They show the hierarchical relationships between constituents, branching downward from S at the top to individual words at the bottom. If you can draw a correct tree, you understand the phrase structure.

Heads vs. Dependents

Every phrase is built around a head, the word that determines what type of phrase it is. The noun is the head of a noun phrase; the verb is the head of a verb phrase. Everything else in the phrase is a dependent of that head.

Dependents come in two types:

• Complements are required by the head to complete its meaning. In give [the book] [to John], both the book and to John are complements of give. Without them, the sentence feels incomplete.
• Adjuncts are optional modifiers that add extra information. In eat [quickly], quickly is an adjunct. You can remove it and the sentence still works fine.

X-bar theory provides a more detailed framework for organizing phrase structure. It proposes that every phrase has the same internal architecture:

• The head (X) combines with its complement to form an intermediate projection (X').
• X' then combines with a specifier and any adjuncts to form the maximal projection (XP).
• In a sentence, the specifier position typically holds the subject, the complement position holds the object, and adjuncts attach at the X' level.

For an intro course, the main takeaway is this: heads are obligatory and define the phrase, complements are required by the head, and adjuncts are optional extras.

Constituents and Sentence Roles

Once you can identify constituents, you need to know what jobs they do in a sentence. There are two ways to describe those jobs:

Grammatical functions describe a constituent's structural role: subject, predicate, object, complement, or modifier. These are about where the constituent sits in the sentence.

Thematic roles describe the semantic relationship between a constituent and the verb. The agent performs the action (The dog chased the cat), and the patient is affected by it (The dog chased the cat). The same grammatical subject can have different thematic roles depending on the verb.

English follows an SVO (Subject-Verb-Object) word order as its default pattern. This is why The dog bit the man means something very different from The man bit the dog.

Verbs also differ in what constituents they require, a property called subcategorization. Transitive verbs need a direct object (She kicked the ball), while intransitive verbs don't (He slept). Some verbs are ditransitive, requiring two objects (She gave him the book).

Finally, watch out for structural ambiguity, where the same string of words can be grouped into constituents in more than one way. Take The man saw the girl with the telescope. If with the telescope attaches to the VP, the man used the telescope to see. If it attaches to the NP the girl, then the girl had the telescope. The words are identical, but the two tree diagrams are different, and so are the meanings.