TOEIC Link Grammar — Aspectual Coercion and Lexical Aspect Mismatch Repair Recognition: The Hidden Aspect-Calculus That Separates Band-22 From Band-25

Aspectual coercion — the silent process by which English repairs a mismatch between a verb's lexical aspect (state, activity, accomplishment, or achievement in the Vendler taxonomy) and the aspectual operator the syntax has imposed on it (progressive, perfect, durative adverbial, frequency adverbial) — is one of the most under-decoded grammar phenomena at the CEFR B2-to-C1 transition. The TOEIC Link grammar module tests coercion directly through items where the surface syntax is well-formed but the lexical aspect of the head verb does not match the constructional aspect of the operator, and the test taker is asked to identify what reading the sentence actually licenses. Band-22 candidates check syntactic licensing alone and pick the surface-correct but aspectually incoherent reading. Band-25 candidates decode the coercion the construction performs on the verb and pick the reading that emerges after coercion.

This guide formalizes the four coercion types the TOEIC Link grammar module systematically exploits, catalogues the four failure modes that hold candidates at band-22, and outlines a four-week drill routine that installs aspectual coercion recognition to automatic processing. For adjacent grammar-module preparation, see the grammar quantifier and negation scope ambiguity resolution discipline guide and the grammar focus particle even only also and association-with-focus recognition discipline guide.

Why aspectual coercion is the right test instrument for band-22-to-band-25 discrimination

The TOEIC Link grammar module operates on a deliberate design principle: above band-22 the rater is no longer scoring whether the candidate can identify syntactically licensed structures but whether the candidate can compute what reading the structure actually delivers when the lexical content does not align with the constructional template. Aspectual coercion is the canonical site at which this misalignment occurs because English permits an extremely wide range of lexical-aspect-to-constructional-aspect mismatches to surface without overt syntactic protest, and the language repairs the mismatch through one of four well-defined coercion operations rather than through error signaling.

The result is that a candidate who has installed only syntactic-licensing checks will accept a sentence like The package arrived for three hours as licensed because the syntax assembles without violation, and will pick the answer choice that reads the duration as the duration of the arrival event itself. The candidate who has installed aspectual coercion recognition will see that arrive is a punctual achievement that cannot host a duration argument directly, will identify that the construction has coerced the verb into an iterative reading (arrived repeatedly across a three-hour window) or into a result-state reading (was in the post-arrival state for three hours), and will pick the answer choice that captures the coerced reading the construction actually licenses. The discrimination between the surface and the coerced reading is exactly what the band-22-vs-band-25 grammar items target.

The four coercion types the TOEIC Link grammar module systematically exploits

Coercion type 1 — State-to-inchoative coercion under progressive operator

The first coercion type arises when the progressive operator is applied to a stative verb. States (know, own, resemble, contain, belong, understand) lexically resist progressive marking because they do not denote a process with internal stages, and the surface form I am knowing the answer is therefore unlicensed in steady-state contexts. The construction coerces the state into an inchoative reading — the state is being entered rather than being held — and the licensed reading becomes I am coming to know the answer or I am in the process of recognizing the answer. The TOEIC Link item that targets this coercion asks the candidate to identify which reading the progressive licenses, and the band-25 answer is the inchoative reading rather than the stative reading.

The repair drill is to install a state-inventory list (know, own, resemble, belong, understand, consist, contain, possess, depend) and to apply inchoative coercion automatically whenever the progressive operator is detected on a verb from the list.

Coercion type 2 — Achievement-to-iterative coercion under durative adverbial

The second coercion type arises when a durative adverbial (for three hours, all day, throughout the meeting) is applied to a punctual achievement verb. Achievements (arrive, depart, find, recognize, notice, discover, reach, win, lose, start) lexically resist durative measurement because they denote a transition at a single point and have no internal duration. The construction coerces the achievement into one of two licensed readings — the iterative reading (the event happened repeatedly across the measured interval) or the result-state reading (the post-transition state persisted across the measured interval). The TOEIC Link item that targets this coercion asks the candidate to identify which reading the construction licenses given the surrounding context, and the band-25 answer is the iterative or result-state reading rather than the impossible direct-duration reading.

The repair drill is to install an achievement-inventory list and to compute, given the lexical aspect and the discourse context, which of the two coerced readings the construction is delivering.

Coercion type 3 — Accomplishment-to-process coercion under in-adverbial vs for-adverbial selection

The third coercion type arises in the systematic alternation between in-adverbials (in three hours) that measure the duration to telos and for-adverbials (for three hours) that measure the duration of the process. Accomplishments (build the house, write the report, solve the puzzle) are telic and combine naturally with in-adverbials, while activities (run, read, swim) are atelic and combine naturally with for-adverbials. When an accomplishment is combined with a for-adverbial, the construction coerces the accomplishment into a process reading that strips the telic component, and the licensed reading becomes "the activity was performed for the measured duration but the telos was not necessarily reached." When an activity is combined with an in-adverbial, the construction coerces the activity into a telic reading by introducing an implicit endpoint, and the licensed reading becomes "the activity reached an inferred endpoint within the measured interval."

The TOEIC Link item that targets this coercion frequently embeds a workplace-realistic scenario (a deadline, a project milestone, a budget cycle) and asks the candidate to identify whether the construction is committing the speaker to telos achievement or to process duration only. The band-25 answer is the reading the adverbial actually licenses given the verb's lexical aspect.

The repair drill is to install the in/for distinction as a default decoding lens and to identify, given the verb's Vendler class, which adverbial preserves the lexical aspect and which forces a coercion.

Coercion type 4 — Activity-to-habitual coercion under frequency adverbial

The fourth coercion type arises when a frequency adverbial (every Tuesday, three times a week, annually) is applied to a singular activity or accomplishment. Activities and accomplishments lexically denote single events, and the surface form John submitted the report annually is therefore aspectually ambiguous between two readings — the single-event reading (where the submission was performed once and the frequency adverbial is misanalysed) and the habitual reading (where the submission is being characterised as a recurring practice). The construction coerces the singular activity into the habitual reading when the frequency adverbial cannot be analysed as a one-shot specification, and the licensed reading becomes "the speaker is characterising John's submission practice across an extended period."

The TOEIC Link item that targets this coercion frequently embeds the habitual reading in a context where the candidate must distinguish between the speaker's report of a single event (which the frequency adverbial would not license) and the speaker's characterisation of a recurring practice (which the construction coerces the verb into). The band-25 answer is the habitual reading.

The repair drill is to install the habitual coercion lens and to identify, given the frequency adverbial's specification, which reading the construction is licensing in context.

The four failure modes that hold candidates at band-22

Failure 1 — Syntactic-licensing-only check

The first failure mode is checking whether the surface form is syntactically licensed and stopping there. The candidate accepts The package arrived for three hours as licensed because the assembly is well-formed and picks the answer choice that reads the duration as the duration of the arrival event itself. The repair is to install a default coercion check that runs whenever the lexical aspect of the head verb is detected to mismatch the aspectual operator of the construction.

Failure 2 — Vendler-taxonomy gap

The second failure mode is having no installed taxonomy of lexical aspect classes. The candidate cannot quickly identify whether a verb is a state, activity, accomplishment, or achievement, and therefore cannot detect which constructional operators will force coercion. The repair is to install a 200-verb Vendler-tagged inventory and drill the tagging until it is automatic, so that on every grammar item the candidate can name the verb's lexical-aspect class before reading the aspectual operator.

Failure 3 — Coercion-direction confusion

The third failure mode is detecting that a coercion is required but mis-identifying which direction the coercion runs. The candidate sees know under a progressive and incorrectly coerces it into a habitual reading rather than into an inchoative reading. The candidate sees an achievement under a durative adverbial and incorrectly coerces it into an accomplishment reading rather than into an iterative or result-state reading. The repair is to drill the four coercion types as a directional lookup table, so that whenever a specific mismatch is detected the correct repair direction is retrievable without computation.

Failure 4 — Context-discourse integration failure

The fourth failure mode is identifying the coercion in isolation but failing to integrate it with the surrounding discourse context. The TOEIC Link grammar item frequently embeds a coerced construction in a scenario where the discourse context disambiguates between two coerced readings (between the iterative and the result-state reading of an achievement under durative adverbial, for example). The candidate who decodes the coercion but does not integrate with context will pick the wrong reading at the disambiguation step. The repair is to drill coercion identification on full-context items rather than on isolated sentence fragments.

The four-week drill routine

Week 1 — Vendler-taxonomy tagging drill

The candidate works through a 200-verb inventory and tags each verb with its Vendler class (state / activity / accomplishment / achievement). The week's output is a tagged-verb log that surfaces which classes the candidate identifies confidently and which require additional drill.

Week 2 — Coercion-direction lookup drill

The candidate works through 80 surface-mismatch examples and identifies, for each one, which of the four coercion types applies and in which direction the coercion runs. The week's output is a coercion-direction log against a band-25 target hit rate of 88%.

Week 3 — Context-discourse integration drill

The candidate works through 50 full-context TOEIC-Link-style grammar items where the coerced reading must be disambiguated against the surrounding discourse and identifies the licensed reading. The week's output is a context-integration log against a band-25 target hit rate of 84%.

Week 4 — Timed-response drill

The candidate works through 30 timed grammar items under TOEIC Link test pace and applies the full coercion-recognition discipline in under 25 seconds per item. The week's output is a timed-hit-rate log against a band-23 baseline of 72% and a band-25 baseline of 86%.

Aspectual coercion under timing pressure

The TOEIC Link grammar item is timed at approximately 30 seconds per item, and any decoding the candidate performs must run within that budget. The four-week drill works specifically because it pushes Vendler tagging, coercion-direction lookup, and context-discourse integration into automatic processing where the cost is sub-second per item. A candidate who has completed the drill performs the lexical-aspect tag, detects the mismatch, retrieves the coercion direction, and integrates the reading with discourse context as a single fused operation that runs within the test pace.

The interaction with the broader grammar resultative secondary predicate and depictive construction recognition discipline matters: resultative constructions frequently embed accomplishment-to-process coercion or achievement-to-result-state coercion as a sub-component, and the resultative reading the construction licenses depends on the coercion the construction has performed on the head verb. The two disciplines interlock, and the candidate who has installed both decodes the joint structure correctly under test pace.

Closing — Coercion-decoded grammar as a band-25 marker

Aspectual coercion recognition is one of the cleanest illustrations of the TOEIC Link grammar principle that above band-22 the rater is testing not whether the candidate can identify syntactic well-formedness but whether the candidate can compute what reading the well-formed surface actually licenses. The four-coercion taxonomy (state-to-inchoative / achievement-to-iterative / accomplishment-to-process / activity-to-habitual) is one such decoding dimension. Installing it over four weeks produces a robust band-25 floor on the set of grammar items that target lexical-aspect-vs-constructional-aspect mismatch repair.

For adjacent grammar-module disciplines, see the grammar quantifier and negation scope ambiguity resolution discipline guide and the grammar focus particle even only also and association-with-focus recognition discipline guide.