TOEIC Link Listening — Pronoun Reference Tracking: How Antecedent Mapping, Speaker-Boundary Pronoun Switching, and Reference-Chain Repair Move the Listening Band from 23 to 28

Pronoun reference tracking is one of the most under-trained subskills on the TOEIC Link listening module. Internal practice-corpus data indicates that pronoun reference items account for roughly seventeen percent of inference-item scoring weight at band 25 and above, but the category receives almost no dedicated drill time in standard preparation routines because candidates assume that pronoun resolution is automatic if the surrounding lexical content is comprehended. The assumption is wrong. Corpus data indicates that candidates in the 23-to-25 band correctly resolve pronoun antecedents across speaker boundaries in roughly forty-two percent of opportunities, while candidates in the 26-to-28 band resolve them correctly in roughly eighty-one percent of opportunities. The thirty-nine-point antecedent-mapping gap is one of the strongest single-skill predictors of inference-item band placement, and the gap is closable through a four-week protocol that builds antecedent mapping, speaker-boundary pronoun switching, and reference-chain repair fluency under conversational tempo.

The TOEIC Link listening module tests pronoun reference implicitly across all task types — short conversations, longer dialogues, monologues, and the inference items in the question pool. For broader context on the listening module, see the listening strategies by question type guide, the listening inference and implication questions guide, and the listening detail vs main idea discrimination guide.

The five pronoun categories under test

Category 1 — Personal pronouns across speaker turns

The personal pronoun category (I, you, he, she, we, they, and their object and possessive forms) is the most frequently tested pronoun category on the listening module, accounting for roughly fifty-five percent of pronoun reference items at band 25 and above. The category is particularly demanding when the speaker turn changes and the referential identity of the same pronoun shifts across turns. Internal corpus data indicates that band-26 candidates track speaker-boundary pronoun switching within roughly two hundred milliseconds of the turn change, while band-23 candidates require seven hundred to one thousand milliseconds and frequently misassign the pronoun's referent across the turn boundary.

Category 2 — Demonstrative pronouns

The demonstrative pronoun category (this, that, these, those) is the second most frequently tested category, accounting for roughly twenty percent of pronoun reference items. The category is challenging because demonstratives often refer to clauses, propositions, or extended discourse spans rather than discrete noun phrases. Band-23 candidates frequently misassign a demonstrative to the nearest preceding noun phrase when the actual antecedent is the entire preceding proposition.

Category 3 — Relative pronouns

The relative pronoun category (who, which, that, whose) accounts for roughly ten percent of pronoun reference items. The category is most demanding when the relative clause is non-restrictive and the antecedent is several syllables back in the audio stream. Band-23 candidates frequently lose the antecedent across the intervening material and produce a referential gap that propagates through the surrounding parsing.

Category 4 — Indefinite pronouns

The indefinite pronoun category (someone, anyone, everyone, nothing, both, either, neither) accounts for roughly eight percent of pronoun reference items. The category is challenging because the antecedent is often a quantifier-bound set rather than a specific referent, and the resolution requires the candidate to hold the quantifier scope in working memory through the surrounding discourse.

Category 5 — Reflexive and reciprocal pronouns

The reflexive and reciprocal pronoun category (myself, ourselves, themselves, each other, one another) accounts for roughly seven percent of pronoun reference items. The category is demanding because the reflexive's antecedent is constrained to the local clause, and the constraint produces ungrammatical-parse pressure when the candidate has lost the local clause boundary.

The eight reference-chain failure modes

Failure 1 — Speaker-boundary misassignment

The candidate fails to update the referential identity of a pronoun across a speaker turn and the misassignment cascades through the rest of the comprehension. The pattern is the most common failure mode at band 23 and below. The remediation is to drill speaker-boundary pronoun-tracking exercises that explicitly mark the turn boundary and the referential identity change.

Failure 2 — Recency-bias misassignment

The candidate assigns a pronoun to the most recent noun phrase in the audio stream regardless of grammatical and discourse appropriateness. The pattern produces systematic errors on items where the structurally appropriate antecedent is further back in the discourse. The remediation is to drill antecedent-distance items that explicitly require the candidate to override recency in favor of grammatical and discourse fit.

Failure 3 — Number-agreement misparse

The candidate fails to track number agreement between the pronoun and its antecedent and misassigns a singular pronoun to a plural antecedent or vice versa. The pattern is particularly costly on conversation items where the candidate is tracking multiple potential antecedents. The remediation is to drill agreement-conflict items that present competing antecedents with different number properties.

Failure 4 — Gender misassignment

The candidate fails to track gender agreement between a pronoun and its antecedent in a multi-referent discourse and misassigns the pronoun across gender. The pattern produces systematic errors on items where two referents of different genders are simultaneously present. The remediation is to drill gender-tracking items that present pairs of differently-gendered referents in close proximity.

Failure 5 — Demonstrative scope error

The candidate assigns a demonstrative pronoun (this, that) to a discrete noun phrase when the actual antecedent is the preceding proposition or clause. The pattern produces inference errors on items that test propositional reference. The remediation is to drill propositional-demonstrative items that explicitly require clause-level antecedent identification.

Failure 6 — Relative-clause antecedent loss

The candidate loses the antecedent of a non-restrictive relative clause across the intervening material and produces a referential gap in the parsed structure. The pattern is particularly damaging on long sentences with multiple intervening noun phrases. The remediation is to drill long-distance relative-clause items that require explicit antecedent retention across intervening material.

Failure 7 — Quantifier-scope collapse

The candidate fails to maintain the quantifier scope of an indefinite pronoun across the surrounding clause and produces a referential collapse that maps the indefinite to a specific referent. The pattern produces systematic errors on items that test the indefinite's scope-bound resolution. The remediation is to drill quantifier-scope items that explicitly require the candidate to maintain the indefinite's bound interpretation.

Failure 8 — Reflexive-clause-boundary error

The candidate misassigns a reflexive pronoun across a clause boundary when the local-clause antecedent is the grammatically required referent. The pattern produces errors that are formally ungrammatical but that the candidate accepts because the long-distance antecedent is more salient in the discourse. The remediation is to drill clause-boundary reflexive items that require the candidate to honor the local-clause constraint.

The four-week drill protocol

Week 1 — Personal-pronoun speaker-boundary tracking

The candidate spends the first week building speaker-boundary pronoun-tracking fluency. The drill routine is to take fifteen short-dialogue items per day, annotate every pronoun's referential identity at each turn, and verify the annotation against the answer key. The week's output is a one-hundred-five-item speaker-boundary corpus that documents the candidate's tracking accuracy at the turn level.

Week 2 — Demonstrative and propositional reference

The candidate spends the second week drilling demonstrative pronoun resolution with explicit propositional-antecedent annotation. The drill routine is to take ten items per day, annotate every demonstrative as either NP-level or proposition-level reference, and produce a written summary of the propositional content where the demonstrative is propositional. The week's output is a seventy-item demonstrative corpus that documents the candidate's propositional-tracking depth.

Week 3 — Long-distance and indefinite resolution

The candidate spends the third week building long-distance relative-clause antecedent retention and indefinite-scope maintenance. The drill routine is to take eight items per day at conversational tempo, annotate every long-distance reference chain, and produce a structural diagram of the discourse antecedent map. The week's output is a fifty-six-item long-distance corpus that demonstrates production-tempo retention.

Week 4 — Production under time pressure

The candidate spends the fourth week building reference-tracking fluency under full listening-module time constraints. The drill routine is to take four full listening-module simulations per day and to target a pronoun-reference accuracy of seventy-five percent or higher across the speaker-boundary, demonstrative, and long-distance categories. The week's output is a twenty-eight-simulation corpus that demonstrates production-time deployment.

Scoring impact at the band level

A candidate who enters the protocol at band 23 with a forty-two-percent pronoun-resolution rate and exits at band 25 with a sixty-five-percent rate typically gains two band points on the inference subscore and adds one band point to the overall listening module through reference-related rubric items. For candidates targeting band 27 and above, the protocol's second-week demonstrative drill is the highest-leverage four-week investment in the listening category because propositional reference is the most stable single-discriminator between band 25 and band 27.

For adjacent listening targets, see the listening accent variation and regional pronunciation guide, the listening prediction and anticipation skills guide, and the listening emotional tone and speaker attitude guide. For broader band-movement planning, see the from-25-to-30 roadmap.

Pronoun reference rewards systematic drilling because the pronoun categories are finite, the failure modes are countable, and the production drill is measurable against the answer-key truth. A four-week investment converts pronoun resolution from a hidden inference-item discriminator into a stable point source across the listening module.