TOEIC Link Reading — Discourse Coherence And Bridging Inference Recognition: The Inter-Sentence Coherence-Bridging Discipline That Converts Implicit-Connection, Causal-Bridge, And Elaboration-Bridge Inferences From Surface-Sentential Reading Into Rubric-Scored Discourse-Integration Comprehension

The TOEIC Link reading section deploys discourse-coherence relations and bridging inferences — implicit-connection bridging that the reader must construct between adjacent sentences where the surface text omits the explicit connector, causal-bridge inference that the reader must construct between an event-stating sentence and a consequence-stating sentence where the cause-effect relation is implicit, elaboration-bridge inference that the reader must construct between a topic-introducing sentence and a topic-specifying sentence where the elaboration relation is implicit, and contrast-bridge inference that the reader must construct between two propositions where the contrastive relation is implicit — at a density that band-22 candidates routinely process as a disconnected sentence sequence and that band-25 candidates routinely process as the rule-governed discourse-integration content. The band-22 candidate encounters the bridging-inference passage, parses each sentence as an independent assertion against the candidate's sentence-level comprehension model without constructing the inter-sentence coherence relation that the discourse-bridging requires, and produces the surface-sentential representation that drives the candidate to select the rubric-incorrect answer option matching the sentence-level interpretation rather than the rubric-correct answer option matching the discourse-integrated interpretation. The band-25 candidate encounters the bridging-inference passage, applies the inter-sentence coherence-recovery procedure that constructs each bridging inference between adjacent or proximal sentences, parses the integrated discourse representation against the standard comprehension model, and produces the rubric-scored discourse-integration comprehension that the bridging-inference-bearing items reward.

The structural difference between the two reading patterns is the inter-sentence coherence-bridging availability that the band-25 candidate deploys and the band-22 candidate omits. The inter-sentence coherence-bridging procedure is the operational adaptation that the reading-section bridging-inference density requires and is the prerequisite for the rubric-scored discourse-integration comprehension on the items that constitute approximately twenty-six percent of the reading-section item set. The inter-sentence coherence-bridging procedure is also the structural complement to the rhetorical-structure mapping strategies that the reading rhetorical structure and argument mapping guide formalizes; the two strategies share the operational premise that the TOEIC Link reading items reward decoding-against-discourse-structure rather than decoding-against-surface-sentence-content alone, and the two strategies share the inter-sentence processing sequence that anchors the comprehension to the discourse-integrated representation rather than to the sentence-aggregated representation.

This guide formalizes the four-category bridging-inference taxonomy that the reading section deploys, the inter-sentence coherence-recovery procedure that constructs each bridging relation, the discourse-context integration that the recovery procedure depends on, and the four-week installation drill that builds the decoding discipline to automatic execution under reading-section pacing. For adjacent reading-strategy context, see the reading textual cohesion and lexical chains tracking guide and the reading anaphora and cataphora resolution strategy guide.

Why the surface-sentential reading caps at band 22

The TOEIC Link reading items that contain bridging-inference content evaluate the candidate's comprehension on the discourse-integrated representation rather than on the aggregated sentence-level content alone, because the rubric-scored comprehension content is the inter-sentence coherence relation that the bridging-inference passage encodes against the discourse-coherence rules. The surface-sentential-reading strategy parses each sentence as an independent assertion against the candidate's sentence-level comprehension model without the inter-sentence coherence-construction, attempts to extract the rubric-scored comprehension as the sentence-aggregated representation, fails to recover the discourse-integrated coherence relation that the rubric scoring requires, and produces the disconnected-sentence representation that the bridging-inference-bearing items penalize.

The surface-sentential representation drives the candidate to the rubric-incorrect answer option through a systematic mechanism. The reading-section items' answer options are calibrated to distinguish the candidates who construct the inter-sentence bridging inference from the candidates who parse the sentence-level content; the answer-option set includes the sentence-level-interpretation distractor as the option-trap that the surface-sentential-reading candidate consistently selects, and the answer-option set includes the discourse-integrated option as the rubric-correct option that the bridging-inference candidate consistently selects. The distractor-calibration structure is the operational mechanism by which the surface-sentential reading strategy caps at band 22 on the bridging-inference-bearing items, because the surface-sentential-reading candidate's option selection is systematically deflected to the distractor option that the calibration is constructed against.

The surface-sentential reading also produces a secondary penalty on the global-comprehension dimension because the candidate's bridging-failure cascades into the candidate's downstream passage-level comprehension representation, which then incorrectly models the passage's macro-organizational structure as a disconnected sentence sequence rather than as a discourse-integrated argumentative or expository structure. The cascading misclassification produces the multi-item global-comprehension-dimension penalty that compounds the per-item bridging-inference-bearing-item penalty, and the combined cascading penalty is the structural mechanism by which the surface-sentential reading strategy cannot reach the band-25 reading-section subscore.

The four-category bridging-inference taxonomy

The TOEIC Link items deploy four categories of bridging inference that the inter-sentence coherence-recovery procedure must recognize and construct against. The four-category taxonomy is the operational template that the candidate maps each encountered bridging-inference passage against, and the within-category recovery rule specifies the coherence relation that each category requires.

Category 1 — Implicit-connection bridging

The implicit-connection bridging category includes the inter-sentence coherence relations where the surface text omits the explicit connector that the discourse-coherence framework would supply in the explicit-connector version — "The procurement deadline shifted. The vendor selection accelerated by two weeks" where the implicit consequence relation is the bridging content, "The audit identified three control gaps. The remediation roadmap allocated forty engineering hours" where the implicit response relation is the bridging content, "The pilot deployment exceeded the throughput target. The general-availability rollout proceeded without revision" where the implicit warrant relation is the bridging content, "The training program completed in March. The certification rate climbed to ninety-two percent by April" where the implicit temporal-consequence relation is the bridging content — and the surface-sentential reading that processes these passages without the implicit-connection construction would either misparse the sentence pair as a disconnected sequence or misparse the second sentence as topic-shift content. The recovery procedure for this category constructs the implicit connector by the discourse-coherence framework's default-connector inference, anchors the constructed connector as the inter-sentence coherence relation, and produces the implicit-connection-constructed representation that the rubric-scored item-answer selection operates against.

Category 2 — Causal-bridge inference

The causal-bridge inference category includes the inter-sentence coherence relations where the cause-effect relation between two propositions is implicit and the bridging inference is the cause-effect construction — "The supply chain disruption persisted through the second quarter. The inventory turnover ratio fell to a six-year low" where the cause-effect bridging is the cause-effect construction, "The pricing optimization launched in January. The gross margin expanded by three hundred basis points by March" where the cause-effect bridging is the cause-effect construction, "The legacy authentication service was deprecated. The third-quarter security incident count fell by sixty-eight percent" where the cause-effect bridging is the cause-effect construction, "The release cadence shifted to weekly deployments. The feature-to-feedback cycle compressed from twenty-eight days to nine" where the cause-effect bridging is the cause-effect construction — and the surface-sentential reading that processes these passages without the causal-bridge construction would either misparse the consequence as a coincidental observation or misparse the cause as a contextual statement. The recovery procedure for this category constructs the causal-bridge inference by the cause-effect-default discourse rule, anchors the constructed causal relation as the inter-sentence coherence relation, and produces the causal-bridge-constructed representation that the causal-bridge-bearing item answer requires.

Category 3 — Elaboration-bridge inference

The elaboration-bridge inference category includes the inter-sentence coherence relations where a topic-introducing sentence is followed by one or more topic-specifying sentences and the bridging inference is the elaboration-relation construction — "The compliance program covers three regulatory regimes. The data-protection framework, the financial-controls framework, and the operational-resilience framework each operate against quarterly attestation cycles" where the elaboration-bridge is the topic-specification construction, "The deployment pipeline has four gates. Gate one validates unit tests, gate two validates integration tests, gate three validates performance benchmarks, and gate four validates security scans" where the elaboration-bridge is the enumeration-specification construction, "The customer-success program operates against three engagement tiers. Tier one delivers self-serve enablement, tier two delivers structured onboarding, and tier three delivers dedicated executive sponsorship" where the elaboration-bridge is the tier-specification construction, "The product roadmap prioritizes five themes. The themes are reliability, performance, security, observability, and developer-experience" where the elaboration-bridge is the theme-specification construction — and the surface-sentential reading that processes these passages without the elaboration-bridge construction would either misparse the elaboration sentence as a topic-shift or misparse the topic-introducing sentence as a standalone assertion. The recovery procedure for this category constructs the elaboration-bridge inference by the topic-elaboration-default discourse rule, anchors the constructed elaboration relation as the inter-sentence coherence relation, and produces the elaboration-bridge-constructed representation that the elaboration-bridge-bearing item answer requires.

Category 4 — Contrast-bridge inference

The contrast-bridge inference category includes the inter-sentence coherence relations where the contrastive relation between two propositions is implicit and the bridging inference is the contrast construction — "The first-quarter forecast projected fifteen-percent growth. The realized growth was twenty-two percent" where the contrast-bridge is the projection-versus-realization construction, "The legacy platform supported six hundred concurrent users. The replatformed system handles four thousand without performance degradation" where the contrast-bridge is the legacy-versus-replatformed construction, "The pre-merger entity operated in eight markets. The combined entity operates in twenty-three" where the contrast-bridge is the pre-merger-versus-combined construction, "The previous renewal cycle delivered an eighty-four percent retention rate. The current cycle is tracking at ninety-three percent" where the contrast-bridge is the previous-versus-current construction — and the surface-sentential reading that processes these passages without the contrast-bridge construction would either misparse the contrast as an additive sequence or misparse the second proposition as topic-shift content. The recovery procedure for this category constructs the contrast-bridge inference by the contrastive-relation-default discourse rule, anchors the constructed contrast relation as the inter-sentence coherence relation, and produces the contrast-bridge-constructed representation that the contrast-bridge-bearing item answer requires.

The inter-sentence coherence-recovery procedure

The inter-sentence coherence-bridging procedure operates as a four-step inter-sentence parse that the candidate executes against each bridging-inference passage at the passage's first reading pass. The four-step procedure is the operational template that converts the surface-sentential parse into the discourse-integrated parse and is the prerequisite for the rubric-scored discourse-integration comprehension that the bridging-inference-bearing items require.

Step one identifies the bridging-inference category by the inter-sentence surface markers — adjacent propositions with omitted explicit connector signal for the implicit-connection category, event-and-consequence proposition pair signal for the causal-bridge category, topic-and-specification proposition sequence signal for the elaboration-bridge category, opposed-or-paired proposition signal for the contrast-bridge category — and is the necessary first step because the within-category recovery rule depends on the bridging-category identification.

Step two applies the within-category recovery rule — default-connector inference for the implicit-connection category, cause-effect-default rule for the causal-bridge category, topic-elaboration-default rule for the elaboration-bridge category, contrastive-relation-default rule for the contrast-bridge category — and produces the discourse-integrated representation that the integrated comprehension model operates against.

Step three integrates the constructed bridging-inference with the broader passage-level discourse-context — preceding discourse referents that confirm the bridging-relation assignment, passage-level rhetorical structure that confirms the bridging-inference's contribution to the passage-level argument, paragraph-organization signals that confirm the bridging-inference's role in the local-coherence structure — and produces the discourse-confirmed bridging representation that anchors the comprehension model against surface-sentential misperception fallback.

Step four validates the discourse-confirmed bridging representation against the answer-option set's discourse-integrated option, deflects the sentence-level-interpretation distractor option that the surface-sentential reading would select, and produces the rubric-scored item-answer selection that the inter-sentence coherence-bridging discipline rewards.

The four-week installation drill

The inter-sentence coherence-bridging discipline requires four weeks of installation drill that builds the bridging-construction execution from controlled to automatic to under-pacing automatic on the reading-section pacing constraint. The four-week drill is the operational schedule that converts the explicit four-step procedure into the implicit automatic parse that the reading-section pacing requires.

Week one installs the four-category bridging-inference recognition through forty bridging-identification items distributed across the four categories at ten items per category. The week-one drill operates at unrestricted pacing and produces the category-recognition fluency that the within-category recovery rule depends on.

Week two installs the within-category recovery-rule execution through eighty bridging-construction items distributed across the four categories at twenty items per category. The week-two drill operates at unrestricted pacing and produces the recovery-rule fluency that the integrated discourse representation depends on.

Week three installs the integrated bridging-and-passage representation through one hundred twenty bridging-inference-bearing-item drills distributed across the reading-section format profile. The week-three drill operates at the reading-section pacing constraint and produces the under-pacing bridging fluency that the rubric-scored item-answer selection requires.

Week four installs the bridging-and-non-bridging interleaved processing through one hundred sixty mixed-passage drills that interleave bridging-inference-bearing passages with explicit-connector-bearing passages at the reading-section pacing constraint. The week-four drill is the structural complement to the week-three drill; the week-three drill installs the bridging-inference construction under pacing, and the week-four drill installs the bridging-versus-explicit category-switching under pacing, which is the final installation step that the reading section's mixed-passage composition requires.

The four-week installation drill produces the inter-sentence coherence-bridging discipline at the rubric-rewarded automatic-execution level that the band-25 reading-section subscore depends on, and the discipline is the operational adaptation that the bridging-inference density requires for the band-25 reading-section performance ceiling.