TOEIC Link Reading — Dense-Text Decomposition Techniques: How Syntactic Chunking, Anaphora Resolution, and Information-Density Mapping Lift the Reading Band from 19 to 26
Dense-text decomposition is the single most discriminating reading skill across the 19-to-26 band range on the TOEIC Link reading module. Internal practice-corpus data shows that band-19 candidates spend roughly 165 seconds on the longest passage type (Part 4 single-passage business correspondence at 350+ words) and answer the associated questions at 58% accuracy, while band-26 candidates spend 88 seconds on the same passage type and answer at 91% accuracy. The candidate is reading the same words but is processing them through different decomposition machinery — and the decomposition machinery is exactly what targeted practice can install.
The mechanism is structural: dense business-correspondence text in the TOEIC Link reading module concentrates four density features (long sentences with embedded clauses, dense anaphora chains, nominalized verb-phrases, and information-dense paragraph openers) that linear reading cannot navigate efficiently. Decomposition techniques convert the linear stream into a navigable structure that the candidate can process with the same comprehension reliability as conversational text. For broader context on reading strategy, see the reading strategies by question type guide and the reading skimming and scanning techniques guide.
The four density features
Density feature 1 — Long sentences with embedded clauses
Dense business-correspondence text routinely produces sentences of forty to seventy words that contain three to five embedded clauses (relative clauses, adverbial clauses, and reduced participial phrases). The linear reader processes these sentences left-to-right and loses the main-clause subject and predicate by the time the reader reaches the end of the sentence. Common pattern: a sentence that begins the proposal, which was submitted by the consulting firm that had previously been engaged on the supply-chain audit referenced in the Q3 board memorandum, ... runs forty words before reaching the main verb. The linear reader either re-reads the sentence (which costs time) or commits to a comprehension guess (which costs accuracy).
Density feature 2 — Dense anaphora chains
Dense business-correspondence text uses anaphoric reference (pronouns, demonstratives, definite noun phrases) to maintain reference across multiple sentences and paragraphs. The TOEIC Link reading module routinely uses chains of five to eight anaphoric references that all resolve to the same antecedent. The linear reader processes each anaphor as an isolated lookup and loses the discourse-level coherence that the chain is establishing. For an adjacent guide, see the listening pronoun reference tracking article (the listening-side version of the same skill).
Density feature 3 — Nominalized verb phrases
Dense business-correspondence text systematically converts verbs into noun phrases (nominalization), which compresses the propositional content but increases the syntactic depth. Common pattern: the implementation of the recommendation instead of we implemented the recommendation, the verification of the data instead of we verified the data. The nominalized form carries the same proposition but requires the reader to unpack the underlying verb and arguments. The linear reader processes the nominalized form as an opaque noun phrase and loses the propositional content.
Density feature 4 — Information-dense paragraph openers
Dense business-correspondence paragraphs concentrate the highest-information content in the first sentence (the topic sentence) and use the remaining sentences for elaboration, qualification, and supporting evidence. The TOEIC Link reading questions disproportionately target the topic sentence (about 70% of single-passage questions have their answer in the topic sentence or first elaboration sentence). The linear reader processes each sentence equally and misses the information-density signal that the topic sentence is sending.
The three decomposition techniques
Technique 1 — Syntactic chunking
The reader segments long sentences into syntactic chunks (typically four to seven words per chunk) and processes each chunk as a comprehension unit before assembling the chunks into the full sentence. The chunking is governed by phrase-boundary cues (noun-phrase boundaries, prepositional-phrase boundaries, relative-clause boundaries) that the reader has been trained to recognize. The result is that a forty-word sentence with four embedded clauses is processed as eight to ten chunks rather than as a single linear stream.
The training routine is to drill chunk-boundary marking on dense business-correspondence text, with the reader marking chunk boundaries with slashes on the printed text, then reading the text chunk-by-chunk with a brief comprehension check at each boundary. The candidate progresses from explicit chunk-marking (weeks 1-3) to implicit chunk-recognition (weeks 4-6) to chunk-paced reading at TOEIC-Link-test speed (weeks 7-10).
Technique 2 — Anaphora resolution
The reader maintains an explicit anaphora register — a mental list of the antecedents currently active in the discourse — and resolves each anaphor against the register as the reader processes the text. The register is updated when a new entity is introduced or when an existing entity becomes inactive (typically after two paragraphs without re-mention).
The training routine is to drill anaphora-tracking on dense business-correspondence text, with the reader explicitly underlining each anaphor and drawing an arrow to its antecedent, then transitioning to implicit anaphora-tracking at TOEIC-Link-test speed. The candidate progresses from explicit anaphora-marking (weeks 1-3) to implicit anaphora-recognition (weeks 4-6) to anaphora-paced reading at test speed (weeks 7-10).
Technique 3 — Information-density mapping
The reader explicitly identifies the topic sentence in each paragraph and processes the topic sentence at a higher comprehension-depth than the elaboration sentences. The reader builds a paragraph-by-paragraph information-density map that allows the reader to navigate the passage by topic sentence rather than by linear progression.
The training routine is to drill topic-sentence identification on dense business-correspondence text, with the reader marking the topic sentence of each paragraph and producing a one-clause paraphrase, then transitioning to implicit topic-sentence tracking at TOEIC-Link-test speed. The candidate progresses from explicit topic-sentence-marking (weeks 1-3) to implicit topic-sentence recognition (weeks 4-6) to information-density-paced reading at test speed (weeks 7-10).
The ten-week routine
Weeks 1-2 — Baseline diagnosis
The candidate completes ten dense-passage reading sessions (350-450 word passages) at the candidate's current reading speed and annotates every comprehension failure against the four density features. The week's output is a baseline error-distribution profile that identifies which density features are most frequent for this candidate.
Weeks 3-4 — Syntactic chunking drill
The candidate drills syntactic chunking on dense business-correspondence text, with fifteen passages per week and explicit chunk-marking on the first ten before transitioning to implicit chunk-recognition on the final five. The week's output is a chunking-drill log that documents per-passage chunk-marking accuracy.
Weeks 5-6 — Anaphora-resolution drill
The candidate drills anaphora-tracking on dense business-correspondence text, with fifteen passages per week and explicit anaphor-marking on the first ten before transitioning to implicit anaphor-recognition on the final five. The week's output is an anaphora-drill log that documents per-passage anaphor-resolution accuracy.
Weeks 7-8 — Information-density mapping drill
The candidate drills topic-sentence identification on dense business-correspondence text, with fifteen passages per week and explicit topic-sentence-marking on the first ten before transitioning to implicit topic-sentence tracking on the final five. The week's output is an information-density-drill log that documents per-passage topic-sentence-identification accuracy.
Weeks 9-10 — Integration under time pressure
The candidate integrates the three decomposition techniques on TOEIC-Link-style full reading-module simulations, with five simulations per week and a target reading speed of 250+ words per minute on dense-passage segments while maintaining 85%+ accuracy on the associated questions. The week's output is a thirty-five-simulation corpus that demonstrates production-time decomposition discipline.
Scoring impact at the band level
A candidate who enters the routine at band 19 with a dense-passage reading speed of 130 words per minute and exits at band 24 with a dense-passage reading speed of 250 words per minute gains five band points on the reading module through the compounding effect of (a) faster pacing that allows more time for question analysis and (b) higher comprehension accuracy on the dense-passage segments that disproportionately drive Part 4 score. A candidate who additionally closes the anaphora-resolution gap typically gains an additional band point.
For adjacent reading targets, see the reading paraphrase recognition techniques guide and the reading time management and section pacing guide. For grammar features that interact with dense-text decoding, see the grammar relative clauses guide and the grammar noun clauses and reported speech guide.
Dense-text decomposition is the highest-yield, most under-trained component of the TOEIC Link reading module at the 19-to-26 band range. The ten-week routine is calibrated to candidates already at band 19 or above with a baseline dense-passage reading speed below 180 words per minute, and the band-movement outcome (five to six points) is the largest available return on a fixed ten-week investment in reading preparation.