Factive Verbs and Presupposition in English Vocabulary: Why “She Realised the Door Was Open” Treats It as True

Compare: > She thought the door was open. with: > She realised the door was open. The surface structure is similar. Both sentences contain: > the door was open. But the two verbs do not treat that proposition in the same way. With **thought**, the sentence tells us what she believed. The door may or may not actually have been open. With **realised**, ordinary English presents **the door was open** as something the speaker takes to be true. The new information is that she came to recognise it. That difference belongs to a family of words known as **factive predicates**, especially **factive verbs**. Typical examples include: > know > realise > discover > regret > remember These words matter because they do more than add a dictionary definition. They change the **truth commitments** carried by the sentence. For a student, that is a vocabulary problem. For a reader, it is an inference problem. For a writer, it is a claim-strength problem. ## Quick answer: what is a factive verb? A **factive verb** is a verb whose complement is normally treated as true. Consider: > Amir knows that the laboratory is closed. The sentence does two things. It presents: > the laboratory is closed as true or taken for granted. Then it says: > Amir knows this. Now negate the main clause: > Amir does not know that the laboratory is closed. The speaker still presents: > the laboratory is closed as true. That persistence is one of the classic signs of **presupposition**. The main assertion changes. The backgrounded proposition remains. ## Assertion and presupposition are different jobs Take: > Mei regrets missing the deadline. The main assertion is approximately: > Mei feels regret about it. The backgrounded content is: > Mei missed the deadline. Now: > Mei does not regret missing the deadline. The attitude changes. But the sentence still normally treats: > Mei missed the deadline as true. This distinction is subtle and extremely useful. A sentence can contain information that is not presented as its main point. ## Presupposition survives negation surprisingly well Compare: > He realised that the calculation was wrong. and: > He did not realise that the calculation was wrong. Both normally take for granted: > the calculation was wrong. What changes is whether he recognised that fact. This is why negation alone does not always cancel the information carried by a sentence. Students who read only for positive or negative verb can miss the deeper meaning. ## Questions can preserve the same background assumption Consider: > Did she regret leaving early? The question asks about regret. But it normally presents: > she left early as already given. Likewise: > Did he realise that the train had left? The question concerns his awareness. The wording still tends to treat: > the train had left as a fact. This is one reason presupposition matters in interviews, examination questions, journalism and everyday conversation. A question can quietly carry assumptions inside it. ## “Know” is different from “think” Compare: > I think the answer is correct. and: > I know the answer is correct. The second is stronger in ordinary discourse. **Think** presents a belief or judgement. **Know** is factive: the construction normally commits the speaker to the truth of the complement. This does not mean every person who says **I know** is actually right. People can be mistaken, careless or dishonest. The linguistic point is about how the sentence is presented. A speaker who says: > I know that p normally presents **p** as factive background, not as a mere possibility. Language structure and real-world truth must still be separated. ## “Discover” packages truth plus change of knowledge Consider: > The researchers discovered that the sensor had failed. The sentence typically carries two layers: > the sensor had failed and: > the researchers came to know this. The verb therefore connects fact + change in epistemic state. This is useful in Science writing because **discover** is stronger than: > suspect > suggest > hypothesise. A writer should not upgrade: > researchers suspected contamination to: > researchers discovered contamination unless the evidence genuinely supports that change. Vocabulary precision becomes evidence precision. ## “Regret” adds attitude to a taken-for-granted event Compare: > She regrets sending the email. The event: > she sent the email is backgrounded as true. The sentence foregrounds her evaluation of it. Now: > She is pleased that she sent the email. This also behaves factively in many ordinary uses. The event is treated as true. The attitude differs. Factive language therefore often carries: > fact + stance. This connection matters in comprehension questions about tone and attitude. ## Factivity is not one perfectly simple box Linguistic research has debated exactly where factive presuppositions come from. Some analyses emphasise the predicate. Others show that complement type, syntax and discourse can affect factive behaviour. There are also uses that are less straightforward than textbook examples. For students, the safest working principle is: > words such as know, realise, regret and discover often present the content of their complement as true, but context and construction still matter. That keeps the mechanism useful without pretending the theory has no edge cases. ## Factive verbs are presupposition triggers A **presupposition trigger** is an expression that tends to introduce information as already accepted or taken for granted. Factive verbs are one major class. Other triggers include expressions involving: > stop > continue > again > definite descriptions > certain cleft structures For example: > Lina stopped playing tennis. normally carries: > Lina used to play tennis. But this article owns the factive-verb job. For the broader topic, see: [Presupposition in English: How Small Words Hide Assumptions Inside a Sentence](https://edukatesg.com/2026/08/28/presupposition-hidden-assumptions-english-reading/) That page owns presupposition generally; this page isolates the lexical behaviour of factive verbs. ## “Stop” is related, but it is not a factive verb in the same sense Consider: > He stopped smoking. This presupposes or strongly backgrounds: > he smoked before. But there is no complement proposition like **that he smoked** selected by a factive verb. The mechanism belongs to the broader presupposition family. That distinction helps students see that: > presupposition > factivity. Factivity is one route into backgrounded meaning. ## Presupposition is not the same as lexical entailment eduKateSG already has: [Lexical Entailment in English Vocabulary](https://edukatesg.com/2026/08/29/lexical-entailment-snore-sleep-vocabulary/) Consider: > He snored. This entails: > he slept. Now: > She realised that the door was open. The proposition **the door was open** appears inside the complement clause and is presented as taken for granted. A useful classroom contrast is: > entailment = what must follow from asserted meaning > presupposition = what the sentence treats as backgrounded truth The formal boundary is more technical, but this contrast is highly useful for readers. ## Why presuppositions are powerful in public language Compare: > Did the company falsify the figures? with: > When did the company stop falsifying the figures? The second question contains a much stronger hidden commitment. It presupposes something like: > the company had been falsifying the figures. The question appears to ask about time. The embedded assumption is more serious. This is why careful readers inspect not only the answer options but the assumptions inside the wording. ## A question can be loaded without being grammatically unusual Suppose a student is asked: > Why did the writer finally admit that the policy had failed? Before answering **why**, inspect the wording. The sentence may background: > the writer admitted something and: > the policy had failed. Depending on the original passage, one or both assumptions may exceed the evidence. This is an excellent reading habit: > unpack the presupposed content before accepting the question’s frame. ## Factivity and journalism Words such as: > confirm > reveal > discover > acknowledge > admit > know can make a proposition sound established. Compare: > The report suggests that the system is vulnerable. with: > The report reveals that the system is vulnerable. The second is rhetorically stronger. It may imply that the vulnerability is real and was uncovered. A responsible writer should choose the verb that matches the evidence. This connects to: [Cause, Correlation and Contribution in English](https://edukatesg.com/2026/08/28/cause-correlation-contribution-english-writing/) The broader principle is the same: > stronger verbs create stronger evidence obligations. ## Factivity and legal/civic language Consider: > The minister acknowledged that the process had failed. Ordinary reading: > the process had failed is presented as accepted fact. Now: > The minister claimed that the process had failed. This is weaker. It reports a claim without automatically endorsing its truth. One verb can change the epistemic architecture of the sentence. Students reading news, legal commentary or GP passages need to notice that difference. ## Factivity and Science Compare: > The experiment showed that the material expanded. with: > The researchers suspected that the material expanded. These sentences do not make equal evidential commitments. A Science answer should not replace **suggest** with **prove** for stylistic variety. Nor should a summary replace **discovered** with **believed** without checking what happened to the truth status. Vocabulary controls the relationship between evidence and proposition. ## Factivity and Mathematics Consider: > The student realised that the two expressions were equivalent. This sentence treats: > the two expressions were equivalent as true. The new event is the student’s change in understanding. Now: > The student thought that the two expressions were equivalent. The mathematical equivalence is no longer guaranteed by the linguistic construction. One sentence reports knowledge. The other reports belief. This is a small but powerful distinction in mathematical explanation. ## Presupposition can fail Imagine: > Mei regrets owning a yacht. But Mei has never owned a yacht. The sentence’s background assumption conflicts with shared reality. This is **presupposition failure**. Listeners may respond: > She does not own a yacht. Notice that the correction targets the background assumption, not the attitude of regret. ## Listeners can accommodate new presuppositions Suppose someone says: > I have to pick up my brother from the airport. You did not previously know the speaker had a brother. You may simply update your mental model: > the speaker has a brother. This is often called **accommodation**. Language does not require every background assumption to have been explicitly stated before. Listeners can add plausible information to the discourse model. This makes communication efficient. It also creates room for manipulation if an unsupported assumption is smuggled into the conversation. ## Presupposition and narrative perspective A controlled narrative sentence can reveal a character’s internal knowledge without explaining everything explicitly. > Only then did she realise that the key had been missing all morning. The sentence backgrounds: > the key had been missing all morning. It foregrounds: > the moment of recognition. The writer therefore controls two times at once: – event time; – knowledge time. That is one reason factive verbs are powerful in narrative. They let a writer move the reader from what was true to when someone understood it. ## Factive verbs and character knowledge Compare: > He knew that the letter was forged. with: > He suspected that the letter was forged. The first places the character in a world where the forged letter is presented as accepted fact. The second leaves uncertainty open. This difference changes suspense, reliability, evidence and what later actions appear rational. Vocabulary can therefore alter the reader’s model of the story world. ## Factivity and reported speech A writer says: > She said that the bridge was unsafe. Does the writer commit to the bridge being unsafe? Not necessarily. Now: > She discovered that the bridge was unsafe. The sentence normally does. This is useful when reading reported speech because different reporting verbs transmit different levels of commitment. The reader should ask: > Is the writer merely reporting someone’s words, or presenting the embedded proposition as true? ## Diagnosis before prescription ### Gap 1: belief and fact collapse The learner treats **think**, **know** and **realise** as interchangeable. **Repair:** compare whether the complement is presented as true. ### Gap 2: negation blindness The learner assumes: > did not realise that p means: > p is false. **Repair:** separate the negated awareness from the backgrounded proposition. ### Gap 3: assertion-only reading The learner reads only the main clause and misses presupposed content. **Repair:** write two lines: > asserted > taken for granted ### Gap 4: factivity overgeneralisation The learner assumes every verb with a **that**-clause is factive. **Repair:** compare: > think that > hope that > know that > regret that ### Gap 5: evidence-strength inflation The learner replaces **suggest** with **reveal** for variety. **Repair:** ask what truth commitment the new verb adds. ### Gap 6: question-frame acceptance The learner accepts presuppositions embedded in a question without checking the passage. **Repair:** unpack the question before answering it. ## A practical factivity test Take: > Lina realised that the answer was wrong. ### Negation > Lina did not realise that the answer was wrong. Does the wording still treat **the answer was wrong** as true? Usually yes. ### Question > Did Lina realise that the answer was wrong? Again, yes. ### Modal > Lina might realise that the answer is wrong. The behaviour becomes more context-sensitive, but the complement is still often presented as factive background. This persistence is the clue. ## A comparison table for vocabulary notebooks | Verb | Complement normally presented as true? | Main contribution | |—|—|—| | know | yes | knowledge | | realise | yes | coming to awareness | | discover | yes | finding out | | regret | yes | negative attitude | | remember | often yes in relevant uses | retained/recovered memory | | think | no | belief/judgement | | suspect | no | tentative belief | | hope | no | desired possibility | The table is not a substitute for context. It is a map of default lexical behaviour. ## Internal-link opportunities This article can connect to: – [Presupposition in English: How Small Words Hide Assumptions Inside a Sentence](https://edukatesg.com/2026/08/28/presupposition-hidden-assumptions-english-reading/) – [Lexical Entailment in English Vocabulary](https://edukatesg.com/2026/08/29/lexical-entailment-snore-sleep-vocabulary/) – [Lexical Coercion in English Vocabulary](https://edukatesg.com/2026/08/29/lexical-coercion-begin-the-book-vocabulary/) – [Cause, Correlation and Contribution in English](https://edukatesg.com/2026/08/28/cause-correlation-contribution-english-writing/) – [Semantic Frames in Vocabulary](https://edukatesg.com/2026/08/29/semantic-frames-vocabulary-event-roles/) – [How to Paraphrase Without Changing Meaning](https://edukatesg.com/2026/08/28/how-to-paraphrase-without-changing-meaning/) – [How Language Works](https://edukatesg.com/2026/08/26/how-language-works/) The distinct reader intent is: > factive verb + complement → backgrounded truth commitment → reader checks what is asserted versus taken for granted ## Connections eduKateAI can learn **Vocabulary ↔ truth status:** verbs do not only describe mental states; some also change how the truth of a complement is presented. **Factivity ↔ presupposition:** predicates such as **know**, **realise** and **regret** often background the truth of the proposition they embed. **Negation ↔ projection:** denying the main verb can leave the presupposed complement intact. **Reading ↔ hidden commitments:** a reader should separate what a sentence asserts from what it quietly treats as already accepted. **Writing ↔ evidence strength:** choosing **discover**, **reveal** or **know** can make a stronger epistemic commitment than choosing **suspect**, **think** or **suggest**. **Questions ↔ framing:** interrogative wording can contain assumptions that should be checked rather than accepted automatically. **Narrative ↔ perspective:** factive verbs can separate the time when something was true from the time when a character recognised it. **Subjects ↔ disciplinary precision:** Science, Mathematics, Humanities and civic reading all depend on distinguishing fact, belief, inference and report. **AI language understanding ↔ background meaning:** robust interpretation requires tracking presupposed content as well as overt assertion. ## Final checkpoint Compare: > She thought that the door was open. with: > She realised that the door was open. Which sentence normally presents the door’s being open as true? Now negate both: > She did not think that the door was open. > She did not realise that the door was open. What happens to the embedded proposition? If the learner can separate belief, factivity and presupposition, these verbs have become usable vocabulary rather than mere dictionary entries. ## Research basis This article is informed by the public research pass, including: – Stanford Encyclopedia of Philosophy, **Presupposition**: https://plato.stanford.edu/archives/spr2024/entries/presupposition/ – Lauri Karttunen, **Implicative Verbs**, *Language* / Cambridge Core: https://www.cambridge.org/core/journals/language/article/abs/implicative-verbs/89B3158B4BCEBEF8CD54E24FD70958B6 – Schueler, **Factivity and Complement-Types**, *Studia Linguistica*: https://onlinelibrary.wiley.com/doi/10.1111/stul.12050 – Ölmezer-Öztürk, **A Corpus-Based Study on “Regret” as a Factive Verb and Its Complements**: https://zenodo.org/records/805967 The article uses factivity as a reader-facing model and preserves the research caution that presupposition behaviour can depend on predicate, complement type and discourse context.