TOEIC Link Reading Anaphoric and Cataphoric Reference Resolution Across Text Distance: The Reference-Tracking Discipline That Maintains Antecedent-and-Postcedent Links Across Multi-Paragraph Passages the Section's Reference-Anchored Items Extract
TOEIC Link Reading passages — particularly the multi-paragraph business-report, policy-document, expanded-news-article, and analytical-brief passages the section's extended-text band concentrates — deploy anaphoric references (pronouns, definite-noun-phrase descriptions, demonstratives, ellipted references that point backward to antecedents established earlier in the text) and cataphoric references (provisional references that point forward to postcedents introduced later in the text) at the cohesion density that business-genre prose carries. The candidates who allow reference-link decay across the text distance between reference and antecedent lose the antecedent-link the section's reference-anchored items target; the candidates who maintain explicit reference-tracking across multi-paragraph distance resolve the references to their antecedents and answer the items the reference-decay readers cannot.
The reference-decay failure pattern is the structural failure that the reference-anchored items extract. The items frequently require the candidate to identify what entity a later-passage pronoun refers to, what definite-noun-phrase description corresponds to an earlier-introduced entity, or what the demonstrative reference is grounded in — and the identification depends on the reference-link having been maintained across the text distance rather than the reader having processed each paragraph in isolation. The candidate who has allowed reference-decay cannot reconstruct the antecedent the item targets and is routed to the distractor that corresponds to a recency-confused antecedent or a salience-confused antecedent rather than the syntactically-and-semantically-licensed antecedent.
This article is the anaphoric-and-cataphoric-reference resolution discipline for TOEIC Link Reading. The guide identifies the reference categories the section's multi-paragraph passages deploy, the tracking protocols that maintain reference-links across text distance, the reference-resolution reasoning operations the items extract, and the deliberate-practice drills that build the reference-tracking competence multi-paragraph reading demands.
The reference categories
The multi-paragraph passages deploy referential expressions in five recurring categories, and each category encodes a specific resolution operation the section's items target. The candidate who has internalized the category repertoire can recognize each category at the reference-expression boundary and apply the category-appropriate resolution protocol; the candidate who has not applies undifferentiated reading that allows the reference-link to decay across the text distance.
Category 1 — pronominal anaphora. The passage deploys third-person pronouns ("it," "they," "this," "these," "he," "she") that point backward to antecedents the prior text has established. The pronominal-anaphora category requires antecedent-search across the prior text and matching the antecedent's number, gender (where marked), and semantic-type against the pronoun's constraints. The category is the highest-frequency reference category in extended business prose and is the most common source of reference-decay failures because pronouns lack the lexical content that would self-anchor the reference.
Category 2 — definite-noun-phrase anaphora. The passage deploys definite noun phrases ("the report," "the strategy," "the team," "the decision") that point backward to antecedents the prior text has established under different surface form. The definite-noun-phrase category requires the reader to recognize the definite article as anaphoric signal and to locate the antecedent that the definite-noun-phrase is re-referencing. The category supports the antecedent-identification items where the antecedent and the definite-noun-phrase reference are surface-different but referentially identical.
Category 3 — demonstrative reference. The passage deploys demonstratives ("this," "that," "these," "those") that point either backward to specific antecedents or to broader discourse-segments the prior text has established. The demonstrative category requires the reader to determine whether the demonstrative is pointing to a specific entity (entity-anaphora) or to a propositional content-segment (discourse-anaphora), and the determination depends on the syntactic position of the demonstrative and the semantic content of the surrounding context.
Category 4 — cataphoric reference. The passage deploys provisional references ("the following," "as follows," "the points outlined below," "the consideration we'll address") that point forward to content the later text will introduce. The cataphoric category requires the reader to recognize the forward-pointing signal, to hold the provisional reference in working memory across the intervening text, and to resolve the reference when the postcedent is introduced. The category is less frequent than anaphoric categories but is high-stakes because the reference-postcedent link spans text distance the reader must traverse.
Category 5 — bridging reference. The passage deploys references that are not co-referential with explicit antecedents but are licensed by inferential bridging from established entities — "the company" referring to an organization whose CEO was the explicit antecedent, "the team" referring to a unit whose lead was the explicit antecedent, "the deadline" referring to a constraint whose project was the explicit antecedent. The bridging category requires the reader to construct the inferential bridge between the reference and the licensing entity, and the reference-anchored items frequently extract the bridging inference.
The tracking protocols
The tracking protocols are the deliberate reading operations the reader executes against multi-paragraph passages to maintain reference-links across text distance. The protocols differ from short-passage reading operations in that the working-memory load of carrying antecedents across paragraphs exceeds what passive reading sustains, and the protocols deploy explicit tracking mechanisms to compensate for the working-memory limitation.
Protocol 1 — antecedent-introduction marking. The reader explicitly marks the introduction of each major entity the passage introduces — the proper-name entities, the indefinite-noun-phrase entities, the organizational-role entities — establishing the antecedent inventory that subsequent references will resolve to. The marking operation is the foundation of reference-tracking because the antecedents must be available for retrieval when later references are encountered, and the marking creates the working-memory anchors that resist decay across text distance.
Protocol 2 — reference-expression detection at expression boundaries. The reader explicitly detects reference expressions at expression boundaries — the pronoun occurrences, the definite-noun-phrase occurrences, the demonstrative occurrences, the cataphoric-signal occurrences. The detection operation produces the reference-event inventory the resolution operations will operate against, and is required because the comprehension items extract the resolved references rather than the reference-expressions themselves.
Protocol 3 — antecedent-search execution on each detected reference. The reader executes antecedent-search on each detected reference, searching backward through the antecedent inventory for the candidate antecedents that satisfy the reference's grammatical and semantic constraints. The search operation respects the constraints — pronouns match number and gender where marked, definite-noun-phrase references match semantic type and discourse-salience, demonstratives match the entity-vs-discourse pointing type the syntax licenses. The search produces the resolved antecedent that the resolution-anchored items extract.
Protocol 4 — reference-chain maintenance across the passage. The reader maintains a running reference-chain across the passage that records the entities each reference has resolved to, building the integrated entity-tracking representation the comprehension items extract against. The chain-maintenance is required because the items frequently extract reference-relations across multiple paragraphs and depend on the integrated entity-trajectory rather than on isolated paragraph-local reference resolution.
The reference-resolution reasoning operations
The reader who has executed the tracking protocols holds the reference-resolved content in working memory; the reader has not yet executed the reference-resolution reasoning operations the items extract. The operations are the analytical operations that convert the resolved-reference content into the comprehension responses the reference-anchored items target.
Operation 1 — antecedent-identification. The operation identifies the antecedent of a specific reference expression the item targets — the entity the pronoun refers to, the prior introduction the definite-noun-phrase re-references, the entity the demonstrative points to. The operation produces the antecedent-identification response the items extract and depends on the Protocol-3 antecedent-search having correctly executed against the constraint structure of the reference.
Operation 2 — referential-distance computation. The operation computes the text distance between a reference and its antecedent — the number of intervening sentences, paragraphs, or topic-shifts. The referential-distance is content-bearing because long-distance references frequently encode topic-return or thematic-recurrence that the items extract, and short-distance references encode local-coherence that the items extract differently.
Operation 3 — bridging-inference reconstruction. The operation reconstructs the bridging inferences that license bridging references — the part-whole relations, the membership relations, the temporal-precedence relations, the causal-precedence relations that the bridging reference assumes. The operation produces the bridging-inference response the bridging-anchored items extract and depends on the Protocol-3 antecedent-search having extended to licensing entities rather than only co-referential entities.
Operation 4 — cataphoric-resolution timing. The operation resolves cataphoric references at the postcedent-introduction point in the passage — recognizing that the postcedent the later text introduces is the resolution of the earlier cataphoric signal. The operation produces the cataphoric-resolution response the items extract and depends on the working-memory maintenance having carried the cataphoric provisional reference across the intervening text distance.
The deliberate-practice drills
The reader who has internalized the categories, protocols, and operations has solved the knowledge problem; the reader has not yet solved the execution-automaticity problem at multi-paragraph passage scale. The execution-automaticity problem is the problem of running the reference-tracking across the four-to-six-paragraph passages the section deploys, without the working-memory load collapsing the reference-chain mid-passage.
Drill 1 — antecedent-marking practice on marked passages. The reader processes multi-paragraph passages whose entity-introductions are pre-marked, marking antecedents during reading and verifying the markings against the reference. The drill develops the Protocol-1 antecedent-introduction pathway and surfaces the antecedent-detection failures the reader must remediate.
Drill 2 — reference-resolution practice on annotated passages. The reader processes passages whose references are annotated with their resolved antecedents in an answer key, resolves references during reading, and verifies the resolutions. The drill develops the Protocol-3 antecedent-search pathway and prevents the antecedent-confusion failures (recency-confused antecedents, salience-confused antecedents, gender-or-number-mismatched antecedents).
Drill 3 — reference-chain construction practice on extended passages. The reader processes extended multi-paragraph passages while constructing an explicit reference-chain, building the working-memory capacity the Protocol-4 chain-maintenance depends on. The drill develops the chain-sustainability across the longer passages and prevents the chain-collapse that under-developed maintenance produces at the five-and-six-paragraph passages the section's harder items concentrate in.
Drill 4 — bridging-inference construction practice on inference-rich passages. The reader processes passages constructed with bridging-reference density and explicitly reconstructs the bridging inferences each bridging reference assumes. The drill develops the Operation-3 bridging-inference pathway and prevents the bridging-failure that defaults to literal-co-reference search and finds no antecedent for bridging references.
Candidates who run this four-drill sequence systematically — antecedent-marking practice daily, reference-resolution drill three times weekly, reference-chain construction twice weekly, bridging-inference construction twice weekly, across a six-to-eight-week window — typically observe a measurable improvement on the reference-anchored comprehension items where the prior reference-decay approach had been losing the antecedent-link points the items extract. The improvement is realized through reference-tracking discipline development rather than through general reading-comprehension improvement.
The related discipline of TOEIC Link Reading rhetorical flow mapping across paragraph boundaries addresses the rhetorical-organization dimension that the reference-tracking discipline operates against when multi-paragraph passages organize content across rhetorical-flow boundaries that intersect with reference-chain boundaries, and the related discipline of TOEIC Link Reading tone shift and author stance pivot detection addresses the stance-tracking dimension that complements the entity-tracking dimension the reference-tracking discipline addresses. The further related discipline of TOEIC Link Reading textual cohesion and lexical chains tracking addresses the lexical-cohesion dimension that the reference-cohesion dimension operates alongside in extended business prose. The four disciplines combine to build the full cohesion-aware reading competence the section's extended-text items demand.