TOEIC Link Speaking — Presupposition Trigger and Shared-Knowledge Anchoring Discipline Under Extended Response: Definite-Description, Factive-Predicate, and Cleft-Construction Triggers That Convert a Two-Minute Opinion Response into a Common-Ground-Calibrated High-Band Performance

Most TOEIC Link candidates who stall between band 25 and band 29 on the extended-response speaking task hit the same ceiling: every sentence is delivered as new information. There is no signal that the speaker treats any part of the prompt or the surrounding world as background. The rater hears a flat, evenly stressed wall of assertions, and the assertion-density score that the LINK rubric uses to discriminate between band 25 and band 29 cannot rise. The discipline that closes this gap is not vocabulary, not grammar, not fluency, and not pronunciation. It is the deliberate deployment of presupposition triggers — definite descriptions, factive predicates, change-of-state verbs, iterative adverbs, cleft constructions, and counterfactual conditionals — that anchor part of each sentence as shared knowledge between speaker and rater, freeing the asserted portion to carry genuinely new propositional content. For the underlying clause-construction mechanics that this discipline rides on, see the cleft and pseudo-cleft construction deployment guide, the meta-discourse and frame signaling guide, and the evidence attribution and source grounding guide.

Why presupposition triggers raise the band-25-to-band-29 ceiling

The LINK extended-response rubric assigns the top two bands not to candidates who produce more words per minute, but to candidates whose responses display the information-structure control that university-educated native speakers use without thinking. The single most diagnostic feature of that control is the use of backgrounding — the placement of supporting propositions inside grammatical structures that mark them as already-shared rather than newly asserted. A presupposition trigger is the pragmatic device that produces this backgrounding. When the candidate says the rising cost of urban commuting, the rising cost is treated as a fact the rater already knows; the assertion the sentence makes is about its consequences, not about the cost itself. When the candidate says what surprises me is the speed of adoption, the cleft construction treats something surprises me as background and asserts only the identifying clause. When the candidate says the report regrets the slow rollout, the factive predicate regret presupposes that the rollout was slow, and the assertion is the institutional attitude toward it.

This backgrounding does three things the LINK rubric rewards. First, it signals that the candidate and the rater share a common ground — a sociolinguistic move that confers authority and reduces the cognitive load on the rater because the rater is not asked to evaluate every clause as a fresh truth claim. Second, it compresses the per-minute information density of the response, because each sentence now carries presupposed plus asserted content rather than asserted content alone. Third, it gives the candidate more elaboration time per assertion, because the trigger word does the backgrounding work that would otherwise require an entire dedicated sentence. The band-25 candidate spends six seconds saying commuting in big cities has become more expensive recently, and this is the situation I want to talk about. The band-29 candidate spends one second saying the rising cost of urban commuting and then uses the remaining five seconds to advance the argument.

A candidate who walks into the LINK speaking section without explicit presupposition-trigger control will produce responses that the rater hears as competent but undifferentiated from band-23 work. The drill is finite, the trigger taxonomy is small, and the technique pays for itself in two weeks.

The six-trigger taxonomy

Presupposition triggers in English fall into six families that the TOEIC Link speaking section repeatedly rewards. Each family produces a distinct backgrounding pattern and a distinct slot in the extended-response template.

Family 1 — Definite descriptions

The definite article the and demonstratives this, that, these, those presuppose the existence and identifiability of the referent. A response that opens the central challenge facing remote-work coordination presupposes that remote-work coordination is something the rater already knows about and that it has a central challenge worth identifying. A response that opens a challenge for remote-work coordination leaves both the existence and the importance unargued. The definite-description trigger is the highest-frequency backgrounding move on the LINK extended-response task and the single trigger that the band-29 candidate uses two-to-three times in the first thirty seconds of any opinion response.

Family 2 — Factive predicates

Factive predicates — verbs and adjectives that presuppose the truth of their complement clause — include know that, realize that, regret that, forget that, be glad that, be aware that, be surprising that, be unfortunate that, resent that, appreciate that, acknowledge that, concede that, and admit that. When the candidate says most employers recognize that hybrid scheduling has become the dominant model, the factive recognize presupposes that hybrid scheduling has indeed become dominant and asserts only the recognition of that fact by employers. The factive family is the trigger family that produces the most rubric-rewarded backgrounding because each factive embeds a full proposition inside the presupposed layer.

Family 3 — Change-of-state verbs

Change-of-state verbs — stop, start, begin, continue, resume, cease, quit, finish, come to — presuppose that the prior state existed. When the candidate says firms have stopped relying on long commutes, the change-of-state verb stop presupposes that firms previously relied on long commutes and asserts only the change. The change-of-state family is the trigger family that compresses the most narrative content into the shortest assertion, which is the precise compression the LINK rubric rewards on time-pressured opinion responses.

Family 4 — Iterative and continuative adverbs

Iterative and continuative adverbs — again, still, another, once more, yet, anymore, no longer — presuppose that a prior event or state occurred. When the candidate says the policy is still under review, the adverb still presupposes that the policy was under review at an earlier moment and asserts only that this state continues. When the candidate says the budget is no longer adequate, the adverbial cluster presupposes that the budget was previously adequate and asserts only the change. The iterative-and-continuative family is the trigger family that adds temporal-stance backgrounding without committing extra syllables.

Family 5 — Cleft and pseudo-cleft constructions

Cleft and pseudo-cleft constructions — it is X that, what X is is Y, the reason X is Y is Z, the one thing X is Y — explicitly mark the existence of the focused element as presupposed and assert only the identification. When the candidate says what concerns me most is the equity dimension, the pseudo-cleft presupposes that something concerns the candidate most and asserts only the identification of that thing as the equity dimension. The cleft family is the trigger family that produces the strongest information-structure-control signal because the construction is grammatically marked as a backgrounding device. The cleft-and-pseudo-cleft mechanics are covered in depth in the cleft construction deployment guide; this discipline asks the candidate to deploy at least one cleft per minute of extended response.

Family 6 — Counterfactual conditionals

Counterfactual conditionals — if X had Y, had X done Y, were X to do Y, but for X, without X — presuppose that the antecedent did not in fact occur. When the candidate says had the rollout been staged in phases, the disruption would have been smaller, the counterfactual presupposes that the rollout was not staged in phases and asserts only the projection. The counterfactual family is the trigger family that the LINK rubric uses to discriminate between band-27 and band-29 candidates because the construction simultaneously displays backgrounding control, tense backshift, and hypothetical projection — three rubric features in one sentence.

The extended-response template that deploys all six families

The discipline asks the candidate to allocate the two-minute extended-response slot across an opening, three argument moves, and a closing, with each segment carrying at least one presupposition trigger from a different family.

Seconds 0–15 — Opening with a definite-description anchor

The opening uses a definite-description trigger to introduce the topic as shared knowledge: the central tension in the prompt is between flexibility and accountability. The opener takes one sentence, presupposes that the prompt has a central tension, and asserts only the candidate's identification of that tension.

Seconds 15–45 — First argument move with a factive predicate

The first argument move deploys a factive predicate: most organizations acknowledge that the productivity argument for remote work has been over-stated. The factive presupposes the over-statement and asserts only its recognition by organizations. This frees the next twenty seconds for a numerical or anecdotal elaboration that the rater treats as building on a presupposed foundation.

Seconds 45–75 — Second argument move with a change-of-state verb and an iterative adverb

The second argument move deploys a change-of-state verb plus an iterative adverb: firms have started rethinking the in-office mandate, and the debate is still unresolved. The change-of-state verb presupposes a prior state of no rethinking, and the iterative adverb presupposes a prior moment of unresolved debate. Two presuppositions stack inside one sentence.

Seconds 75–105 — Third argument move with a cleft construction

The third argument move deploys a cleft construction: what makes this question difficult is the asymmetry between salaried and hourly workers. The pseudo-cleft presupposes that something makes the question difficult and asserts only the identification of that thing. The cleft is the rubric-rewarded high-information-structure-control move that separates band-27 from band-29.

Seconds 105–120 — Closing with a counterfactual conditional

The closing deploys a counterfactual conditional: had the policy been designed with hourly workers in mind, the equity gap would not have widened. The counterfactual presupposes the absence of inclusive design and asserts only the projection of the alternative outcome. The closing simultaneously summarizes the candidate's position and displays the hypothetical-projection register that the rubric rewards.

The trigger-density target

The discipline asks for a presupposition-trigger density of one trigger per ten seconds of extended response, distributed across at least four of the six families inside a two-minute slot. A candidate who hits this density and this distribution will not score below band-27 on information-structure-control, regardless of pronunciation or fluency variation. A candidate who hits half this density will not score above band-25, regardless of vocabulary breadth.

The recognition drill that closes the gap inside two weeks

The drill has three stages and should run for fifteen minutes per day across ten consecutive days.

Stage 1 — Trigger spotting in transcribed model responses (days 1–3)

The candidate listens to a band-29 model response, transcribes it in full, and highlights every presupposition trigger by family. The target is to find six-to-eight triggers in a two-minute response and to classify each by family. This stage builds the recognition reflex that the production stage depends on.

Stage 2 — Trigger insertion into one-minute prompted responses (days 4–7)

The candidate records a one-minute response to a LINK-style opinion prompt, transcribes it, and identifies every assertion that could have been backgrounded with a presupposition trigger. The candidate then re-records the same response with the triggers inserted and compares the two for information density.

Stage 3 — Full two-minute responses with target trigger density (days 8–10)

The candidate records two-minute responses on novel prompts and self-audits against the trigger-density target (one per ten seconds, at least four families represented). A trigger count below ten or a family count below four is treated as a failed run and triggers a re-record.

A candidate who completes this ten-day drill arrives at the LINK speaking section with the trigger-deployment reflex preloaded and will not need to consciously plan triggers during the timed response. The reflex moves the speaker from band 25 to band 27 in week one and from band 27 to band 29 in week two for candidates whose other rubric dimensions are already at band 27.

For the broader extended-response discipline that this trigger taxonomy sits inside, see the extended discourse and multi-turn coherence control guide, the stance modulation and commitment calibration guide, and the elaboration depth and supporting detail deployment guide.