Verb Phrase Ellipsis in English: Why “Mei Can Solve It, and Amir Can Too” Leaves a Whole Verb Phrase Silent

> Mei can solve the problem, and Amir can too. The second clause contains: > Amir can too. Where is: > solve the problem? It is not pronounced. Yet the sentence is complete. A reader automatically reconstructs: > Amir can solve the problem too. English has omitted a whole verb phrase while leaving an auxiliary behind. This construction is called **verb phrase ellipsis**, usually shortened to **VP ellipsis** or **VPE**. It is one of the clearest examples of an important language principle: > English can leave structure silent when the surrounding grammar and discourse make the missing meaning recoverable. The result is not careless incompleteness. It is licensed compression. ## Quick answer: what is VP ellipsis? VP ellipsis occurs when a verb phrase is unspoken but understood from an antecedent. Full: > Lina will submit the form, and Amir will submit the form too. With VP ellipsis: > Lina will submit the form, and Amir will too. The surviving auxiliary **will** helps license the missing verb phrase: > submit the form. Another: > Mei likes the book, and Amir does too. The first clause has no auxiliary. English uses **do-support** in the elliptical clause: > does [like the book]. Another: > Priya has finished the task, and Jun has too. Understood: > has finished the task. ## English often leaves the auxiliary visible Common survivors include: > can > could > will > would > should > may > might > have > be > do Examples: > She can swim, and I can too. > They will attend, and we will too. > He has finished, and she has too. > Mei doesn’t agree, but Amir does. The auxiliary carries grammatical information such as tense, modality, polarity and agreement. The repeated lexical predicate can remain silent. ## Why “Amir too” is not the same construction Compare: > Mei likes coffee, and Amir does too. VP ellipsis: > does [like coffee]. Now: > Mei likes coffee, and Amir tea. That is more naturally analysed as **gapping**, because the repeated predicate is missing inside a coordinated clause with contrasting remnants. Now: > Mei likes coffee, and Amir too. This may be analysed as stripping or another reduced coordination pattern depending on the structure. The general lesson is: > not every missing verb is VP ellipsis. The surviving auxiliary is one of the clearest English clues. ## VP ellipsis and the existing eduKateSG ellipsis page eduKateSG already has: [Ellipsis and Substitution in English](https://edukatesg.com/2026/08/29/ellipsis-substitution-cohesion-english/) That article owns the broad reader job: > how English omits or substitutes repeated material for cohesion. This article owns the narrower mechanism: > auxiliary remains + verb phrase disappears + antecedent supplies the interpretation. The specialised page deepens the existing estate rather than replacing it. ## Do-support appears when the first clause has a lexical verb Consider: > Mei understands the chapter. To avoid repetition: > Amir does too. The word **does** does not mean: > performs an unspecified action. It is grammatical support. The understood content is: > understands the chapter. Now past tense: > Mei understood the chapter, and Amir did too. **Did** carries past tense. The lexical predicate remains silent. This is why English auxiliary grammar is central to VP ellipsis. ## Modal auxiliaries make the structure especially visible > Mei can solve it, but Amir cannot. Understood: > Amir cannot solve it. Now: > Mei might attend, and Amir might too. Understood: > Amir might attend. The modal survives because it contributes information that cannot simply disappear. The repeated lower predicate is recoverable. ## Negation can remain while the VP disappears > Lina will attend, but Amir will not. Understood: > Amir will not attend. Now: > Lina submitted the form, but Amir did not. Understood: > Amir did not submit the form. The negative marker belongs with the surviving auxiliary structure. This is useful because English can contrast outcomes with very little repetition. ## Short answers use the same economy Question: > Can you swim? Answer: > I can. The missing material is: > swim. Question: > Did Mei submit the form? Answer: > She did. Understood: > submitted the form. In classroom grammar, these may simply be called short answers. They illustrate the same larger English ability to recover a silent predicate from context. ## “To” can also stand before ellipsis Consider: > I wanted to finish the chapter, but I wasn’t able to. The final **to** can be understood as: > to finish the chapter. Or: > I did not want to apologise, but I knew I had to. Understood: > had to apologise. The exact analysis of infinitival ellipsis can be treated separately in advanced syntax, but for readers it belongs to the same family of recoverable predicate omission. The overt marker tells us a missing action is still structurally expected. ## The antecedent is the source of the missing meaning Sentence: > Mei polished the table, and Amir did too. Antecedent VP: > polished the table. Ellipsis: > did [polish the table]. The reader needs to locate the right antecedent. This becomes harder when several possible actions are nearby. ## Multiple possible antecedents can create ambiguity Consider: > Mei said that Lina would submit the form, and Amir did too. What did Amir do? A natural interpretation is: > Amir said that Lina would submit the form. But complex sentences can create competing antecedents. The sentence is risky because **did too** may force the reader to decide which predicate is being reused. Good writing avoids making the reader search through several candidates. ## Parallelism helps recovery Clear: > Mei can solve the first problem, and Amir can too. The clauses are parallel: > Mei / Amir > can / can > solve the first problem / [silent] Now: > Mei can solve the first problem because her teacher explained it, and after the break Amir can too. Still recoverable. But more intervening material increases processing load. Ellipsis works best when the antecedent remains accessible. ## Distance matters to processing Research on VP ellipsis processing has examined how readers retrieve an antecedent from memory. The practical lesson is important: > the more competing material lies between antecedent and ellipsis, the more carefully a writer should test clarity. A grammatical sentence can still be hard to process. Reader-first writing distinguishes: > licensed from: > easy to recover. ## Strict and sloppy readings Consider: > Mei washed her car, and Lina did too. There are at least two possible readings. ### Sloppy reading Lina washed: > Lina’s own car. ### Strict reading Lina washed: > Mei’s car. The silent VP contains a pronoun relationship that can be reconstructed in more than one way. This is called the **strict/sloppy ambiguity**. The sentence is a reminder that ellipsis does not always reproduce surface words mechanically. Reference relations matter. ## Why the strict/sloppy contrast matters for students Suppose: > Amir called his mother, and Daniel did too. Did Daniel call: – Daniel’s mother? – Amir’s mother? Ordinary context may favour one. Grammar allows both kinds of interpretation in many classic cases. If the distinction matters, a writer should name the referent: > Daniel called his own mother too. or: > Daniel also called Amir’s mother. Clarity beats elegant compression. ## VP ellipsis can preserve a different subject Full: > Mei will revise the essay. Elliptical: > Amir will too. The silent VP is reused with a new subject: > Amir will revise the essay. So the antecedent is not copied as one frozen string. Some grammatical dependencies update with the new clause. This is why ellipsis is a serious syntax-semantics phenomenon, not simple deletion from printed text. ## Auxiliary choice must match the grammar Antecedent: > Mei is studying. Elliptical: > Amir is too. Antecedent: > Mei has studied. Elliptical: > Amir has too. Antecedent: > Mei studied. Elliptical: > Amir did too. English chooses an auxiliary form that carries the required tense, aspect and agreement. Learners who omit the auxiliary may be using an ellipsis strategy that is natural elsewhere but not licensed in the same way in Standard English. ## A Singapore multilingual connection Singapore students may use English alongside Mandarin, Malay, Tamil or other languages and varieties. Languages differ in how ellipsis is licensed. A pattern that is natural without an overt auxiliary in one language may require: > do > can > will > have in Standard English. The teaching goal is not to label multilingual transfer as careless. It is to show which overt grammatical cue English expects. ## VP ellipsis and gapping Compare: > Mei ordered tea, and Amir coffee. Gapping: > [ordered] is missing from the second coordination. Now: > Mei ordered tea, and Amir did too. VP ellipsis: > did [order tea]. The difference is visible in what remains. Gapping leaves contrastive remnants. VP ellipsis leaves an auxiliary or related licensor. See: [Gapping in English](https://edukatesg.com/2026/08/29/gapping-ellipsis-omit-repeated-verb-english/) ## VP ellipsis and sluicing Compare: > Someone submitted the form, but I don’t know who. That is **sluicing**: > who [submitted the form]? The surviving element is a wh-expression. Now: > Mei submitted the form, and Amir did too. VP ellipsis. Both use recoverable silence. The grammatical architecture differs. See: [Sluicing in English](https://edukatesg.com/2026/08/29/sluicing-ellipsis-someone-called-dont-know-who-english/) ## VP ellipsis and substitution Traditional school grammar sometimes describes: > do so as substitution. Compare: > Mei completed the analysis, and Amir did so too. Here **did so** contains overt material **so** that stands in for the event description. Now: > Mei completed the analysis, and Amir did too. No **so** appears. The VP is silent. This distinction is useful because English has several ways to avoid repetition. ## Why writers use VP ellipsis Full repetition: > The first group can explain the mechanism, and the second group can explain the mechanism too. Ellipsis: > The first group can explain the mechanism, and the second group can too. The second is lighter. The repeated information disappears. The contrast **first group / second group** becomes more visible. That is efficient when the antecedent is clear. ## But repetition can be better Suppose a long paragraph has introduced several tasks: > analyse the graph > evaluate the source > calculate the rate > explain the anomaly Then: > Group B can too. Too what? The writer may know. The reader may not. Repeat the predicate: > Group B can explain the anomaly too. Economy is useful only when recovery is safe. ## Composition and dialogue Dialogue naturally uses VP ellipsis: > “Can you come tomorrow?” > “I can.” > “Did she apologise?” > “She did.” This keeps speech realistic. But a composition should not use short elliptical dialogue so heavily that the reader loses who is agreeing with what. The mechanism supports naturalism. It does not replace scene clarity. ## Comprehension questions Passage: > Mei could not identify the cause, but her partner could. A student must reconstruct: > her partner could identify the cause. The auxiliary **could** does not stand alone semantically. The missing VP supplies the content. This is especially important in dense passages where the antecedent is several words away. ## Science > Trial A did not produce a measurable change, but Trial B did. Understood: > Trial B produced a measurable change. The construction creates a clean contrast. But Science readers should check whether the antecedent predicate is exactly the same. If Trial B merely showed a different effect, ellipsis may conceal a meaningful distinction. ## Mathematics > Mei can factorise the expression mentally, and Amir can too. Clear. Now: > Mei can factorise the expression and solve the equation, and Amir can too. What exactly can Amir do? Potentially both actions. If only one ability is intended, the sentence is under-specified. Ellipsis can inherit coordination scope. Mathematical writing should restore the specific predicate when the distinction affects capability. ## Humanities > One historian accepts the economic explanation; another does not. Understood: > another historian does not accept the economic explanation. This is compact and useful. But **another does not** can hide what exactly is rejected if several claims precede it. Critical reading reconstructs the proposition before evaluating disagreement. ## VP ellipsis and AI language understanding Computational work on ellipsis resolution treats VP ellipsis as a retrieval problem: 1. detect that a VP is missing; 2. locate a plausible antecedent; 3. reconstruct the intended content; 4. resolve reference; 5. handle ambiguity. This mirrors the human reader’s task. The presence of **did**, **will** or **can** is only the first clue. Meaning requires discourse. ## Diagnosis before prescription ### Gap 1: missing auxiliary Learner produces an under-supported second clause where Standard English expects: > does / did / can / will / has. **Repair:** reconstruct the full clause and identify which auxiliary carries tense or modality. ### Gap 2: wrong antecedent The learner reconstructs the nearest verb even when the discourse points elsewhere. **Repair:** test semantic parallelism, not distance alone. ### Gap 3: every omission called VP ellipsis The learner labels gapping, sluicing and substitution as one structure. **Repair:** identify what survives: auxiliary, wh-word, remnant pair or substitute expression. ### Gap 4: pronoun ambiguity ignored The learner assumes one strict/sloppy interpretation. **Repair:** expand both possible readings. ### Gap 5: over-compression The writer uses **did too** after several candidate predicates. **Repair:** repeat the predicate when antecedent retrieval is unsafe. ### Gap 6: tense/aspect mismatch The learner chooses the wrong auxiliary. **Repair:** recover the full clause first, then keep the correct auxiliary. ## A practical reconstruction method Sentence: > Lina has completed the graph, and Amir has too. ### Step 1: identify the licensor > has. ### Step 2: find antecedent VP > completed the graph. ### Step 3: reconstruct with new subject > Amir has completed the graph too. ### Step 4: test reference Any pronouns inside the VP? ### Step 5: test ambiguity Could another antecedent fit? If not, the ellipsis is clear. ## Internal-link opportunities This article can connect to: – [Ellipsis and Substitution in English](https://edukatesg.com/2026/08/29/ellipsis-substitution-cohesion-english/) – [Gapping in English](https://edukatesg.com/2026/08/29/gapping-ellipsis-omit-repeated-verb-english/) – [Sluicing in English](https://edukatesg.com/2026/08/29/sluicing-ellipsis-someone-called-dont-know-who-english/) – [Embedded Questions in English](https://edukatesg.com/2026/08/29/embedded-questions-indirect-question-word-order-english/) – [Sentence Ambiguity and Word Placement](https://edukatesg.com/2026/08/28/sentence-ambiguity-word-placement-clarity/) – [End-Weight and Extraposition in English](https://edukatesg.com/2026/08/29/end-weight-extraposition-information-structure-english/) – [How Language Works](https://edukatesg.com/2026/08/26/how-language-works/) The distinct reader intent is: > auxiliary survives → repeated VP is silent → antecedent and reference reconstruct the missing predicate ## Connections eduKateAI can learn **Ellipsis ↔ antecedent:** the silent verb phrase receives its interpretation from accessible earlier or contextual material. **Auxiliary ↔ licensing:** English commonly leaves an auxiliary or related marker overt while the lexical VP disappears. **Do-support ↔ ellipsis:** when no auxiliary is available in the antecedent, English can use **do** to carry tense and agreement. **Reference ↔ ambiguity:** pronouns inside the reconstructed VP can support strict and sloppy readings. **Writing ↔ cohesion:** VP ellipsis reduces repetition while preserving the relation that connects two clauses. **Processing ↔ distance:** antecedent retrieval becomes harder when competing predicates or long intervening material weaken accessibility. **Multilingualism ↔ grammar:** languages license silent predicates differently, so English auxiliary requirements may need explicit teaching. **Subjects ↔ comparison:** Mathematics, Science and Humanities use VP ellipsis to contrast abilities, results and positions compactly. **AI ↔ discourse resolution:** interpreting ellipsis requires detecting missing structure, retrieving an antecedent and resolving reference rather than reading only visible words. ## Final checkpoint Expand: > Mei can solve the problem, and Amir can too. Then explain why: > Mei washed her car, and Lina did too. may have two interpretations. Finally, distinguish: > Amir did too from: > Amir coffee in: > Mei ordered tea, and Amir coffee. If the learner can identify the licensor, antecedent and reference structure, VP ellipsis has become visible grammar. ## Research basis This article is informed by the public research pass, including: – Freidin, *Adventures in English Syntax*, Cambridge University Press, discussion of auxiliary **do** and VP ellipsis: https://www.cambridge.org/core/books/adventures-in-english-syntax/does-every-politician-who-cheats-instinctively-lie/6F4F3A044EB1FF269837EA8B3EC966D6 – McShane & Babkin, **Detection and Resolution of Verb Phrase Ellipsis**, ACL Anthology: https://aclanthology.org/2016.lilt-13.1/ – Hobbs & Kehler, **A Theory of Parallelism and the Case of VP Ellipsis**, ACL Anthology: https://aclanthology.org/P97-1051/ – Elliott, Nicolae & Sudo, **The Sticky Reading: VP Ellipsis Without Parallel Binding**, *Semantics and Linguistic Theory*: https://journals.linguisticsociety.org/proceedings/index.php/SALT/article/view/24.640 – Oxford Handbook research on cross-linguistic ellipsis: https://academic.oup.com/edited-volume/41718 The article presents English auxiliary-stranding VP ellipsis conservatively and treats strict/sloppy ambiguity as an advanced reading feature rather than a rule students must reproduce in every use.