Questions: Syntax-Semantics Interface and Compositionality
5 questions to test your understanding
Score: 0 / 5
Question 1 Multiple Choice
The sentence 'Every student read a book' is ambiguous between two readings: one where each student read some (possibly different) book, and one where there is a single specific book that all students read. What is the standard syntactic-semantic account of this ambiguity?
AThe word 'a' is lexically ambiguous between 'one specific' and 'some arbitrary,' producing the two readings
BThe sentence contains a prosodic ambiguity — different stress patterns trigger different quantifier scope readings
CQuantifiers can undergo covert movement at the level of Logical Form (LF), taking different scope positions without changing surface word order, and generating both interpretations
DSurface subject-verb-object order always assigns universal quantifiers wide scope, so only the 'every student read some book' reading is grammatically licensed
The standard account invokes covert movement at LF. On the surface, 'every student' c-commands 'a book,' which would predict only the ∀ > ∃ reading. The ∃ > ∀ reading ('there is one book all students read') requires 'a book' to take scope over 'every student' — achieved by covert raising of 'a book' at LF, invisible in the phonology. This is the key insight: Logical Form is a syntactic level where interpretive operations occur that leave no surface trace. Option D is the common misconception — surface order does not uniquely determine scope.
Question 2 Multiple Choice
The idiom 'kick the bucket' (meaning 'to die') has no compositional semantic derivation from its parts. What does this show about the syntax-semantics interface?
AThe phrase is syntactically irregular — it lacks the VP structure of standard transitive constructions
BCompositionality fails completely for all complex expressions; meaning is stored holistically rather than computed
CIdioms are the limit case where compositional rules do not apply; the meaning is stored as a unit, marking the boundary of the compositional system rather than refuting it
DThe semantic module operates before the syntactic module, assigning meaning to full phrases before they are parsed
Idioms exist as non-compositional islands within an otherwise compositional system. Their existence does not refute compositionality — it maps its boundary. 'Kick the bucket' must be stored in the lexicon as a multi-word unit because no composition of 'kick' + 'the' + 'bucket' yields 'die.' The syntax is still regular (it can passivize: 'the bucket was kicked' — the idiom survives passivization in many cases), but the semantic composition simply does not apply. The key analytical point is what the exception tells us about the system's limits.
Question 3 True / False
Binding constraints — such as the requirement that reflexives like 'himself' be bound within their local syntactic domain — are stated over syntactic configurations (c-command) even though their effects are semantic (they determine reference).
TTrue
FFalse
Answer: True
This is precisely why binding theory illustrates the interface: the constraints are syntactic in form (stated over c-command relations, local domains) but semantic in consequence (they determine whether a pronoun can refer to a given antecedent). This means semantic interpretation cannot ignore syntactic structure — reference determination is partly a function of syntactic configuration, not just semantic context or world knowledge. Cross-linguistic variation in binding (e.g., long-distance reflexives in Chinese) shows that the boundary between syntactic and semantic resolution of binding is itself a research question.
Question 4 True / False
If the syntactic structure of a sentence uniquely determines its meaning, then most scope ambiguities should be resolvable by examining surface word order.
TTrue
FFalse
Answer: False
The existence of scope ambiguities like 'Every student read a book' directly contradicts this. The surface word order is fixed, yet two interpretations are available. The standard account (LF movement) posits that quantifier scope is determined at a level of representation (Logical Form) that is distinct from the surface order — meaning is not read off directly from the phonologically visible string. This is the central motivation for positing LF as an interface level: without it, the architecture cannot explain how identical surface forms can receive distinct semantic interpretations.
Question 5 Short Answer
What is Logical Form (LF), and why does the syntax-semantics interface require it rather than letting semantic interpretation operate directly on surface syntactic structure?
Think about your answer, then reveal below.
Model answer: Logical Form is a level of syntactic representation, derived from the surface structure via covert (phonologically invisible) operations, where scope, binding, and other interpretive phenomena are made explicit for semantic rules to apply to. The interface requires LF because surface syntax underdetermines meaning in systematic ways: the same surface string can be assigned multiple scope readings (quantifier scope ambiguities), anaphora resolution requires checking c-command relations that may not be apparent from word order alone, and languages exhibit binding patterns that only become regular at a post-surface level. LF is the level at which syntactic structure and semantic composition align consistently.
LF captures the insight that the syntax-semantics mapping is not a direct surface-to-meaning translation. Covert movement at LF (raising of quantifiers, wh-in-situ in some languages) explains why meaning can diverge from surface form without positing purely semantic scope-assignment rules disconnected from syntax. The architecture maintains that the same computational system (Merge, movement) drives both overt and covert operations, preserving the unity of the grammatical system while explaining surface/meaning mismatches.