Notes on Constituency and Tests in English Syntax

Constituency and the Structure of Syntax

• Syntax adds a layer of structuring language beyond phonetics, phonology, and morphology. It governs how words group into phrases and how those phrases form sentences to convey meaning.
• Key claim: humans have an unbounded capacity to generate and understand syntax, which requires a rule-based account to explain how we apply general principles to infinite possibilities.
• Rule-based explanation in English example: declarative sentences can form yes/no questions by moving the first auxiliary after the subject. This illustrates how a simple rule can generate an infinite set of questions:
  • Basic statement → move auxiliary after subject → yes/no question.
• The goal of a theory of language is to capture this rule-based, generative capacity and explain how structure guides interpretation.

Constituency: The Basic Idea

• Constituency is about grouping words into units that function together as a unit within a sentence; these units are called constituents.
• Constituents are the basic units of sentence structure and are indispensable for understanding how sentences are built.
• In linguistics, words belong to parts of speech (e.g., nouns, verbs, adjectives, determiners, prepositions, complementizers). These categories label the chunks they head or constitute (e.g., noun phrases, verb phrases).
• The head of a phrase determines the syntactic type of the whole phrase and labels the entire constituent:
  • In a verb phrase (VP), the verb is the head.
  • In a noun phrase (NP), the noun is the head.
  • This head-labeling helps explain why the phrase is categorized as NP, VP, PP, etc.
• Trees and brackets: syntax trees (or bracketed structures) encode these hierarchical groupings. Brackets label constituents and their syntactic categories; nested brackets show constituents within constituents.
• Analogy for structure: left and right brackets in trees correspond to a hierarchical nesting (a box within a box, like a shoebox inside another box). Linear order on the page is not the same as hierarchical structure.
• Consequence: the same surface string can have different meaning because its internal constituency differs (e.g., The boy hit the ball vs. The ball hit the boy).

Constituents, Phrases, and the Labeling System

• Constituents can be identified as chunks that may contain other constituents inside them (hierarchical structure).
• Examples of phrase types and their heads:
  • Verb Phrase (VP): head is the verb; can contain other constituents (e.g., object NP, adverbs, prepositional phrases).
  • Noun Phrase (NP): head is the noun; may include determiners, adjectives, or other modifiers.
  • Prepositional Phrase (PP): head is the preposition; may contain an NP as its object.
• The central idea: constituents are units of internal syntactic structure that are built up from smaller constituents in a hierarchical fashion, not merely strings of adjacent words.
• The canonical positions in English for basic SVO order: subject, verb, object. Moving elements out of order can highlight their constituent status.

Evidence for Constituency: How Do We Detect Constituents?

• Because the internal structure is not visible on the surface, we rely on diagnostics/tests to identify constituents.
• Two related observations underpin the approach:
  • Some strings can move (front), and if they can, they are likely constituents (movement = unitary chunk).
  • If moving a string yields grammatical, acceptable sentences, that string behaves as a constituent.
• Logical idea: if a string can be fronted and the resulting sentence is grammatical, that string is a constituent. If it cannot, the test is uninformative (not disconfirming).

Movement as a Diagnostic: Fronting Constituents

• Movement (often called fronting) tests whether a string can be moved to the front of the clause.
• If a string can be moved and the resulting sentence is grammatical, that string is evidence of being a constituent.
• Example: Two sentences with the same words in different orders can differ in meaning due to internal structure, not just order:
  • The boy hit the ball.
  • The ball hit the boy.
    These differences arise because the internal constituencies differ, even though the surface material is the same.
• Fronting examples: moving a string like a noun phrase or a prepositional phrase to the front can produce acceptable sentences, indicating that the moved string is a constituent.
• In class, fronting is often demonstrated with sentences such as: The news, when it arrives, … (fronting the fronted element). The goal is to see if the fronted string can be interpreted as a constituent.
• Practical takeaway: use multiple fronting cases to collect evidence about constituent structure; alone, fronting is not foolproof for all constituents.

Clefting: A Clear, Portable Movement Test

• Clefting is a formal operation used to foreground a constituent by moving it to the front and exposing its role in the sentence.
• Core cleft pattern (template):
  • It is X that Y.
  • It was X who Y.
  • In general: It is/was X (that/who) Y.
• How to apply the cleft test:
  • Take a candidate string X from the sentence and test whether placing it in the X slot and the remaining material in Y yields a grammatical sentence.
  • If yes, X behaves like a constituent; hence, X is a constituent.
  • The test focuses on the X part (the moved string) and whether the resulting sentence is grammatical.
• Example: The student of linguistics bought a brand new car in the holidays.
  • Test whether The student of linguistics is a constituent: It was the student of linguistics that bought a brand new car in the holidays.
  • This yields a grammatical sentence, so “the student of linguistics” is a constituent.
  • Another test: It was in the holidays that the student of linguistics bought a brand new car.
  • This shows that “in the holidays” is also a constituent, as fronting it yields a grammatical sentence.
• Important rule about clefting:
  • When applying clefting, the string placed in the X slot is the one being tested for constituency; the material in the Y slot is the remainder.
  • The test is diagnostic only if the resulting sentence is grammatical; if not, the test is uninformative rather than conclusive that X is not a constituent.
• Practical guidance on clefting:
  • Clefiting is powerful for NP and PP testing; it works well for noun phrases and prepositional phrases.
  • It may not reliably detect all kinds of constituents; use it in combination with other tests.
  • The presence of felicity conditions matters: even grammatical clefts can sound odd in some contexts; the test’s usefulness depends on context and discourse (felicity).
• Summary: Clefing demonstrates a result where the fronted X is a constituent; a failed cleft does not prove non-constituency; use additional tests to corroborate.

Felicity Conditions and Test Limitations

• Felicity conditions: tests have to fit the communicative context; moving something can sound odd if misapplied, even if X is a constituent.
• Not all constituents cleft cleanly; some strings do not cleft well, and the cleft test can fail even for constituents.
• So: do not rely on a single test; use a battery of tests and consider the context (the ideal scenario is to gather positive evidence from multiple tests).

Other Constituency Tests: A Practical Toolkit

• The textbook prescribes a range of constituency tests beyond clefting, including but not limited to:
  • Stand-alone test (fragment/short answer test): a single fragment can answer a question about the constituent, indicating it is a constituent.
  • Pro-form test: replace the candidate string with a pro-form (e.g., it, that, do so) to test for constituenthood.
  • Movement as a unit test: move the candidate as a unit to another position; if grammatical, supports constituency.
  • Coordination test: test whether the candidate can be coordinated with another string (e.g., NP1 and NP2) using conjunctions like and/or; if so, the candidate is a constituent.
• The lecturer emphasizes that the textbook should be used for a comprehensive repertoire of tests; not all tests work for all strings, and some tests yield confusing results in certain environments.
• The coordination test is highlighted as particularly powerful; the stand-alone (fragment) test is also very useful and popular for quick checks.
• The general advice: build a repertoire of constituency tests and practice applying them in context to detect constituents reliably.

Putting It All Together: Building and Reading Structures

• Trees/Brackets: once you identify constituents, you can bracket them in the sentence to reveal the hierarchical structure.
• The goal is to infer the internal syntactic structure that underlies the surface string, which is otherwise invisible in linear order.
• In constructions like The elephant snorted up all of the peanuts, determiner usage (the) interacts with the noun to form NP, and the rest of the sentence interacts with the VP/PP structure.
• The general principle: constituents can be nested, and there is no intrinsic upper bound to the depth of nesting in English; you can always form a larger constituent by combining existing constituents.

Practical Takeaways for Exam Preparation

• Remember the key concepts:
  • Constituents are units of internal syntactic structure; they can contain other constituents.
  • The head of a phrase labels the entire constituent (e.g., head of VP is the verb).
  • Hierarchical structure, not surface order, drives meaning and interpretation.
• Be fluent with the main constituency tests and their caveats:
  • Movement/fronting: indicates constituency if grammatical.
  • Clefting: foregrounds X; grammaticality of It is X that Y or It was X who Y supports X as a constituent.
  • Felicity conditions: context matters; tests can produce odd sentences even for valid constituents.
  • Stand-alone, pro-form, coordination tests: provide additional evidence and help cover cases where clefting fails.
• Practice strategy:
  • Use multiple tests in combination rather than relying on a single diagnostic.
  • Practice with NP, VP, and PP examples to build familiarity with different phrase types.
  • Build and read trees/bracket structures to internalize hierarchical organization (boxes within boxes).
• Real-world relevance and connections:
  • Understanding constituency underpins parsing in natural language processing and linguistic theory.
  • The idea of a rule-based, hierarchical structure connects to broader theories of language acquisition and cognition.
• Ethical/philosophical note: recognizing the limits of tests and the importance of context highlights the complexity of language and the careful interpretation required in linguistic analysis.

Quick Reference: Key Notation and Concepts

• Constituents: chunks of words functioning as a unit; can nest other constituents.
• Phrase types and heads:
  • NP (head: noun)
  • VP (head: verb)
  • PP (head: preposition)
• Canonical English order: $S o ext{Subject} o ext{Verb} o ext{Object} (SVO)$
• Cleft construction templates:
  • $ext{It is } X ext{ that } Y$
  • $ext{It was } X ext{ who } Y$
• Movement/Fronting: if a string can be moved to the front producing a grammatical sentence, it is evidence of constituency.
• Felicity conditions: context-dependent acceptability of constructions; testing requires appropriate discourse context.
• Key examples to study:
  • The boy hit the ball vs. The ball hit the boy (structure drives meaning).
  • The elephant snorted up all of the peanuts (illustrates NP formation with determiners).
• Strategy: combine multiple constituency tests (clefting, stand-alone, pro-form, movement, coordination) and read textbook for a broad toolkit.