Questions: Binding Theory and Anaphora Resolution

Question 1

In the sentence 'John told Bill about himself,' the reflexive 'himself' can refer to John or Bill but not to some other person mentioned in prior discourse. Which principle of binding theory explains this?

• A. Principle B — pronouns must be free within the binding domain
• B. Principle C — R-expressions must be free everywhere
• C. Principle A — anaphors must be bound within the binding domain
• D. The binding domain parameter — reflexives in English are always clause-local

Answer: (C)

Principle A states that anaphors (including reflexives like 'himself') must be bound — must have a c-commanding antecedent — within their binding domain. In this sentence, the binding domain is the clause, and both 'John' and 'Bill' c-command 'himself' within it. A discourse-external person cannot serve as antecedent because Principle A requires the antecedent to be within the local clause. Option D correctly names the locality restriction but misidentifies its source as parameterization rather than Principle A itself.

Question 2

A student argues: 'The sentence "John likes him" sounds weird if "him" refers to John, so binding theory classifies it as merely pragmatically dispreferred.' What is the error in this reasoning?

• A. The student is correct — Principle B makes local co-reference merely dispreferred, not ungrammatical
• B. Binding constraints are structural grammaticality rules, not pragmatic preferences — Principle B makes that co-reference ungrammatical, not just awkward
• C. Principle B only blocks co-reference when the pronoun and antecedent are adjacent
• D. Principle B applies only to reciprocals, not to pronouns like 'him'

Answer: (B)

The Common Misconception in binding theory is treating the constraints as pragmatic preferences rather than grammaticality conditions. Binding theory makes categorical predictions: a sentence violating Principle B is ungrammatical with the local co-referential interpretation — not just awkward or dispreferred. The constraints are defined over structural relations (c-command, binding domains) that either hold or don't, producing categorical rather than gradient judgments.

Question 3

Binding constraints are structural grammaticality rules, not merely pragmatic preferences about which interpretations 'sound right.'

• T. True
• F. False

Answer: True

This is the core methodological point of binding theory: the constraints predict categorical grammaticality, not gradient acceptability. 'John hurt himself' is grammatical with local co-reference; 'Himself hurt John' is ungrammatical. The constraints are categorical because they are defined over structural relations (c-command, binding domains) that either hold or don't. Pragmatic factors may influence what interpretations are preferred in context, but they operate on top of the grammaticality constraint, not instead of it.

Question 4

The fact that Icelandic reflexives can refer to subjects in higher clauses across clause boundaries shows that binding theory's claim about binding domains is incorrect.

• T. True
• F. False

Answer: False

Icelandic long-distance reflexives don't refute binding theory — they motivate it to treat the binding domain as a parameterized variable rather than a fixed universal. The universal principle is that anaphors must be bound within their binding domain; the parameter is how each language fixes the boundaries of that domain. Icelandic defines its binding domain more broadly than English. Cross-linguistic variation is evidence for parameterization within the theory, not against the theory itself.

Question 5

Why must anaphors and pronouns have complementary distributions within a binding domain? What would break down if a pronoun could have a local c-commanding antecedent?

Model answer: If pronouns could co-refer with local c-commanding noun phrases, the distinction between pronouns and reflexives would collapse within that domain — both could take local antecedents. Languages maintain the contrast because reflexives signal bound, local co-reference (Principle A) while pronouns signal non-local or free reference (Principle B). Allowing pronoun-antecedent co-reference locally would create systematic ambiguity between 'John hurt him' (someone else) and 'John hurt himself' (John), undermining the expressive distinction these forms encode.

The complementarity of Principles A and B is not accidental — it creates a complementary distribution of anaphors and pronouns that allows speakers to reliably signal whether co-reference is local or non-local. This is a productive grammatical contrast that speakers exploit without conscious awareness.