Questions: Labeling Algorithm and Syntactic Categories

Question 1

Classical Transformational Grammar used rules like VP → V NP to specify that verbs take noun-phrase complements. Why does Minimalist syntax not need such rules?

• A. Minimalist syntax assumes all verbs inherently require exactly one complement, making the rule trivially true and unnecessary
• B. The labeling algorithm computes that {V, NP} has category V because the verb is the most feature-prominent element, deriving the projection without stipulating it
• C. Minimalist syntax uses movement rules instead of phrase structure rules to create verb phrase structure
• D. The merge operation generates all possible combinations and the labeling algorithm filters out ungrammatical structures after the fact

Answer: (B)

Phrase structure rules like VP → V NP are descriptive stipulations — they list permitted combinations without explaining why. The labeling algorithm derives the same result from a principled property: V is the most prominent element in {V, NP} because its features (tense-assigning, case-assigning properties) are what other elements in the derivation are looking for. The category 'VP' is not stipulated; it is computed. This is Minimalism's explanatory goal: eliminate rules that describe patterns and replace them with mechanisms that derive patterns.

Question 2

When a subject DP merges with a VP to form a clausal structure {DP, VP}, how does the labeling algorithm assign a label to the result?

• A. The DP labels the structure because subjects are the most prominent elements in a sentence
• B. The VP labels the structure because it already has a computed label from an earlier merge
• C. The structure is labeled through feature sharing between the subject DP and the verbal head via φ-feature agreement
• D. The structure remains unlabeled until T (tense) merges above it, at which point T provides the label

Answer: (C)

In a specifier-head structure like {DP, VP}, neither element is the head of the other, so the straightforward 'head projects' rule does not apply. Chomsky's proposal is that labeling succeeds through feature sharing: if the DP and the head of the VP share φ-features through agreement, that shared feature serves as the label. This is why specifier-head agreement is not just a surface grammatical phenomenon — it is the mechanism by which specifier-phrase structures become properly labeled and the derivation can continue.

Question 3

The labeling algorithm applies after the merge operation, computing the syntactic category of the newly merged object from structural and feature-based properties.

• T. True
• F. False

Answer: True

Merge creates an unlabeled set {X, Y}. The labeling algorithm then determines what category this object has — whether it projects as a verb phrase, noun phrase, etc. — based on which element is most feature-prominent or (in specifier cases) on feature sharing. The sequence is: first merge, then label. This is why the algorithm is called an algorithm rather than a rule built into the merge operation itself.

Question 4

In head-complement structures, the complement provides the label of the merged object because the complement is typically a larger, more complex phrase.

• T. True
• F. False

Answer: False

It is the head, not the complement, that provides the label in head-complement structures. The head projects because it has features that the rest of the derivation is 'looking for' — tense, case assignment, agreement requirements. Size or complexity is irrelevant to labeling; prominence is defined by feature visibility to the derivation, not by structural size. This is why a single verb V labels the entire {V, NP} complex as a verb phrase, even when the NP is more complex.

Question 5

Why does Minimalism consider it an explanatory advance to derive phrase structure categories through the labeling algorithm rather than listing them in rules like VP → V NP?

Model answer: Phrase structure rules are descriptive stipulations: they list what combinations exist without explaining why. The labeling algorithm derives these patterns from a more primitive, independently motivated property — which element has features visible to the rest of the structure. Instead of being told that V takes NP complements to form VP, we derive this from the fact that V is feature-prominent. Minimalism's goal is to minimize stipulation: every property of grammar that can be derived from simpler mechanisms should be. The labeling algorithm reduces what must be listed in the grammar, bringing the grammar closer to a system explainable from general computational principles.

This connects directly to the broader Minimalist Program: the goal is not just a descriptively adequate grammar (one that correctly describes the data) but an explanatorily adequate one that derives grammatical properties from deeper principles. Eliminating phrase structure rules and replacing them with the labeling algorithm is a step toward that goal.