The origin of the question mark is a Latin manuscript, not a single inventor. Its ancestor, the punctus interrogativus, a wavy stroke over a point, appears in books made at Charlemagne's court in the late eighth century, among them the Godescalc Evangelistary of 781 to 783. Our reading of the evidence: no one person can be shown to have invented it, and the popular stories about Alcuin of York, a Latin abbreviation and an Egyptian cat are unproven at best.
EdCitation publishes this guide, written from sources read on 2 October 2026: the manuscript itself, the scholars and dictionaries who trace the mark, the Spanish Academy, Unicode, APA and MLA.
EdCitation never writes any part of anyone's work, and we say so with pride: every question in your paper, and the mark that ends it, stays yours. It finds, cites and checks instead. Cite a source builds a reference from a DOI in six styles, free, and Mechanics QA reads a finished paper for formatting and the slips a spell-checker lets through.
Who invented the question mark?
Nobody can be named. The earliest Latin example O'Conner and Kellerman (2022) could find is in the Godescalc Evangelistary, a book of Gospel readings made for Charlemagne and his wife Hildegard. The Bibliothèque nationale de France (n.d.) dates it 781 to 783, before Hildegard's death on 30 April 783, and places its copying at the court, perhaps then at Worms; O'Conner and Kellerman say Aachen.
What the manuscript shows
O'Conner and Kellerman found several in its first pages, among them folio 6v. We looked at that page in the library's digitised copy on Gallica. After "autem quid" (Peter's question in John 21:21) and "quid ad te" (the reply in 21:22), a small wavy stroke climbs to the right above the points that close each question. The text is in gold and silver uncials; only the dedication poem is in Caroline minuscule (Bibliothèque nationale de France, n.d.).
The scribe names himself Godescalc in that poem. Whether he devised the mark or brought a habit from elsewhere, no record says.
What the punctus interrogativus was for
It told someone reading aloud where the voice should rise. The palaeographer Malcolm Parkes, quoted by O'Conner and Kellerman (2022), counts it among the marks of the later eighth century made for liturgical texts that were read or sung, and says it "seems to have spread rapidly from the court of Charlemagne to other centres", reaching Corbie around 800. Albert Derolez traced it to a neume, a sign in musical notation for a rising voice, as they also report.
The names came later: the Oxford English Dictionary's earliest evidence in English is from 1952 for "punctus interrogativus" and from 1862 for "question mark" (Oxford University Press, n.d.-a, n.d.-b). Unicode now encodes the medieval shape as U+2E54 MEDIEVAL QUESTION MARK (Unicode Consortium, 2026e).
Did Alcuin of York invent the question mark?
No document shows that he did. O'Conner and Kellerman (2022) date Alcuin's time running Charlemagne's palace school at Aachen from 782 to 793, after the Godescalc manuscript was begun. His interest in points is documented: in a letter of 799 he told Charlemagne that distinctiones and subdistinctiones, marks for pauses, had nearly fallen out of use through the "rusticity" of scribes. Yet they found no text of his that uses or mentions the punctus interrogativus, not even his question-and-answer Quaestiones in Genesim.
Lexico (n.d.), the Oxford dictionary site that closed in August 2022, read in an archived copy, called Alcuin's authorship "the story accepted by most" and cited nothing for it. Treat Alcuin as the court's documented champion of punctuation, not its proven inventor.
Is the question mark Egyptian cat myth true?
No evidence supports it. The tale says the curl of the question mark copies a curious cat's tail, in ancient Egypt or from a monk's pet; Lexico (n.d.) told it, then said "there is no evidence to back up this tale". We found no Egyptologist, museum or manuscript scholar who links any Egyptian sign to the mark, and could not trace where the story began.
The quaestio and "Qo" story
The second legend says scholars wrote quaestio, Latin for "question", at the end of a question, shortened it to qo, and stacked the q on the o until it became ?. Lexico (n.d.) reports that there is no manuscript evidence for this either. The earliest Latin examples point elsewhere: the Godescalc marks are a stroke over points, not letters.
The Syriac double dot, a scholar's claim
One older candidate is real, though not our mark's ancestor. In 2011 Chip Coakley of Cambridge University Library argued that the zagwa elaya, two stacked dots over a word near the start of a sentence in fifth-century Syriac Bible manuscripts, marks a question; the same report says Greek and Latin question marks appear from the eighth century, probably independently (University of Cambridge, 2011).
What is documented about the question mark, and what is legend?
Most popular stories are legend; the manuscripts are documented.
| Claim | Documented or legend | Source |
|---|---|---|
| Punctus interrogativus in the Godescalc Evangelistary, 781 to 783 | Documented: the mark is visible on folio 6v | Bibliothèque nationale de France (n.d.); O'Conner and Kellerman (2022) |
| It spread from Charlemagne's court, reaching Corbie about 800 | Documented: a palaeographer's finding | Parkes, quoted by O'Conner and Kellerman (2022) |
| Alcuin of York invented it | Not documented: no writing of his uses it | O'Conner and Kellerman (2022); asserted without a source by Lexico (n.d.) |
| Alcuin urged scribes to punctuate | Documented: his letter of 799 | O'Conner and Kellerman (2022) |
| It copies an Egyptian cat's tail | Legend, no evidence | Lexico (n.d.) |
| It grew from qo, short for quaestio | Legend, no manuscript evidence | Lexico (n.d.) |
| Syriac marked questions by the fifth century | A scholar's claim, reported in 2011 | University of Cambridge (2011) |
| Henry Denham printed a reversed ? for rhetorical questions in 1575 | Documented by Parkes, as Houston reports | Houston (2011) |
| Spanish ¿ first recommended in 1754 | Documented by the Academy itself | Real Academia Española and ASALE (2010a) |
| "Question mark" as an English name, 1862 | Documented | Oxford University Press (n.d.-b) |
| Cite a source keeps a title's ? in APA and MLA | Our run, 2 October 2026 | Nagel (1974), below |
The modern shape came with print: the earliest example O'Conner and Kellerman (2022) had seen of a mark that looks and acts like today's is in Pietro Bembo's De Aetna, printed by Aldus Manutius in February 1496 (see our history of the semicolon). Houston (2011) reports that in 1575 the printer Henry Denham added a reversed question mark, the percontation point, for rhetorical questions; it lasted about fifty years, and Unicode keeps it as U+2E2E, alias punctus percontativus (Unicode Consortium, 2026e).
Where did the Spanish upside-down question mark come from?
From the Real Academia Española, in 1754. The Academy's Ortografía calls the punctus interrogativus a Carolingian addition to classical punctuation, at first a single closing mark for questions and exclamations alike, and says the opening sign was first recommended in academic spelling in 1754, though printed texts took it up generally only later (Real Academia Española & ASALE, 2010a). Ridao Rodrigo (2022) finds that the Academy's first Orthographía, of 1741, shows only the closing sign, which it calls the interrogante, and cites García Folgado for dating its general use to about 1870.
What the Academy asks now
- Both signs, every time. Spanish question marks are double signs, like brackets: ¿Quién le ha llamado? is right, and Quién le ha llamado? is wrong (Real Academia Española, n.d.).
- No full stop after the closing sign.
- A doubtful date takes both signs, as in (¿1576?-1636), or the closing one alone (Real Academia Española & ASALE, 2010b).
Referencing in Spanish-speaking countries is covered in Spanish citation styles.
What do Greek and Arabic question marks look like?
Greek uses a mark shaped like a semicolon, and Arabic a question mark whose hook faces the other way. Unicode's code charts list them:
| Mark | Code point and name | What Unicode's chart says |
|---|---|---|
| ? | U+003F QUESTION MARK | Cross-refers to each relative below (Unicode Consortium, 2026b) |
| ¿ | U+00BF INVERTED QUESTION MARK | "Spanish"; also "turned question mark" (2026c) |
| ; | U+037E GREEK QUESTION MARK | Also erotimatiko; equivalent to the semicolon, U+003B, the preferred character (2026d) |
| ؟ | U+061F ARABIC QUESTION MARK | Also used with Thaana and Syriac in modern text (2026a) |
| ⸮ | U+2E2E REVERSED QUESTION MARK | Also punctus percontativus (2026e) |
| ⹔ | U+2E54 MEDIEVAL QUESTION MARK | Listed beside the medieval exclamation mark (2026e) |
The Greek mark hides a trap. Unicode makes U+037E equivalent to the semicolon and names U+003B the preferred character for a Greek question (Unicode Consortium, 2026b, 2026d), so a Greek question in your source may end in a plain semicolon. That is correct, not a typing slip.
How do you cite a title ending in a question mark in APA and MLA?
Keep the question mark and drop the full stop that would follow it. APA finishes the title element with a period unless the title ends in a question mark or exclamation point, which then replaces it; its example is Late-onset unexplained epilepsy: What are we missing? (American Psychological Association, 2025). MLA agrees for a source's title, since the stronger mark supersedes a period, but keeps the comma after a container title that ends in one, as in What Is Philosophy?, by Gilles Deleuze and Félix Guattari (Modern Language Association, 2016); that post was written for the eighth edition. Capitals in such titles are covered in title case or sentence case, and question marks beside quotation marks in full stops inside or outside quotation marks.
A worked example: "What is it like to be a bat?"
On 2 October 2026 we gave EdCitation's Cite a source the DOI 10.2307/2183914, Thomas Nagel's 1974 article. It read Crossref's record, found no retraction, and returned, italics as it set them:
APA 7: Nagel, T. (1974). What is it like to be a bat? The Philosophical Review, 83(4), 435. https://doi.org/10.2307/2183914
MLA 9: Nagel, Thomas. “What Is It Like to Be a Bat?” The Philosophical Review, vol. 83, no. 4, Oct. 1974, p. 435, https://doi.org/10.2307/2183914.
The punctuation is right in both. APA adds no full stop after "bat?", and MLA keeps the question mark inside the closing quotation mark with nothing after it. APA's sentence case lowers the title's capitals; MLA keeps them.
One thing needs your hand. Crossref's record gives only the first page, 435, and the article runs to page 450, as its JSTOR cover sheet shows. Corrected:
- APA 7: Nagel, T. (1974). What is it like to be a bat? The Philosophical Review, 83(4), 435–450. https://doi.org/10.2307/2183914
- MLA 9: Nagel, Thomas. “What Is It Like to Be a Bat?” The Philosophical Review, vol. 83, no. 4, Oct. 1974, pp. 435-50, https://doi.org/10.2307/2183914.
MLA's own examples shorten the second page number, as in "pp. 250-58" (Modern Language Association, n.d.). The in-text forms came back as (Nagel, 1974) for APA and (Nagel) for MLA; add the page you quote.
How does EdCitation help with the question marks in a paper?
It checks the paper and references around your questions and leaves the questions to you. In our view EdCitation is the best partner for that last read, because it looks each reference up in the publisher's record and never writes one: in the run above, the punctuation after Nagel's title came out right in both styles, and the one gap was the record's, in plain sight.
Mechanics QA, for the finished file
Mechanics QA gives an uploaded Word document or PDF the first read a marker gives it, done by rule. It holds font, size, spacing, margins, title page, headings and reference list to the APA 7 student paper; finds comments, tracked changes, placeholders and template text left in the file; matches each citation to the list and looks each reference up in the publisher's record; flags figures that do not add up or disagree between sections; and catches the slips a spell-checker lets through. No model reads the paper, and each finding names the rule it broke and its fix, marked where it sits. Whether a sentence should end in a question mark is your judgement, not a rule. Mechanics QA is part of Pro, $8 a month or $72 a year, and a free account can run it three times a month (pricing).
Find sources and Verify references, free
EdCitation's Find sources searches about 300 million published works; on 2 October 2026, "punctus interrogativus" returned 58, led by the Oxford English Dictionary's entry. We pasted two references into Verify references. Nagel's corrected entry came back verified: "The DOI resolves to this record and the title matches." Our first entry for Ridao Rodrigo's article, dated 2021 from its volume, came back "check this": "The work was published in 2022, not 2021; the record has no date in 2021." The journal shows volume 27 (2021) published on 3 October 2022, so we cite 2022. Neighbouring marks have histories of their own: the interrobang, the em dash and the apostrophe.
Quick questions
Who invented the question mark?
No one person can be shown to have. Its ancestor, the punctus interrogativus, appears in manuscripts from Charlemagne's court, the earliest cited being the Godescalc Evangelistary of 781 to 783.
Did the question mark come from an Egyptian cat's tail?
No. The Egyptian cat story is a myth with no evidence behind it; the earliest question signs scholars have identified are Syriac, from the fifth century, then Latin and Greek, from the eighth.
Why does Spanish use an upside-down question mark?
The Real Academia Española first recommended the opening ¿ in 1754, and today it is obligatory, paired with the closing ?.
Do I add a full stop after a title ending in a question mark?
No. In APA the question mark replaces the full stop, and in MLA it supersedes it; MLA keeps the comma after a container title that ends in one.
Will EdCitation correct the question marks in my essay?
No: EdCitation never writes or rewrites your work. In our test, Cite a source kept a title's question mark correctly in APA and MLA, and Mechanics QA, in Pro, reports formatting and spell-checker slips for you to fix.
References
- American Psychological Association. (2025). Elements of reference list entries. APA Style. https://apastyle.apa.org/style-grammar-guidelines/references/elements-list-entry
- Bibliothèque nationale de France. (n.d.). Evangeliarium [Évangéliaire dit de Charlemagne ou de Godescalc], Nouvelle acquisition latine 1203 [Catalogue record]. BnF Archives et manuscrits. Retrieved 2 October 2026, from https://archivesetmanuscrits.bnf.fr/exportBranchePdf.html?eadCid=FRBNFEAD000034648
- Houston, K. (2011, September 4). Irony & sarcasm marks, part 1 of 3. Shady Characters. https://shadycharacters.co.uk/2011/09/irony-sarcasm-marks-part-1-of-3/
- Lexico. (n.d.). What is the origin of the question mark? Archived copy of 25 August 2022. https://web.archive.org/web/20220825044524/https://www.lexico.com/explore/origin-of-question-mark
- Modern Language Association. (n.d.). How to cite an online work. MLA Style Center. Retrieved 2 October 2026, from https://style.mla.org/works-cited/citations-by-format/online-works/
- Modern Language Association. (2016, October 11). When citing a work whose title ends in a question mark or exclamation point, should I also include a period? MLA Style Center. https://style.mla.org/punctuation-and-titles/
- Nagel, T. (1974). What is it like to be a bat? The Philosophical Review, 83(4), 435–450. https://doi.org/10.2307/2183914
- O'Conner, P. T., & Kellerman, S. (2022, February 28). Who invented the question mark? The Grammarphobia Blog. https://grammarphobia.com/blog/2022/02/question-mark.html
- Oxford University Press. (n.d.-a). Punctus interrogativus, n. In Oxford English dictionary. Retrieved 2 October 2026, from https://doi.org/10.1093/oed/4203815571
- Oxford University Press. (n.d.-b). Question mark, n. In Oxford English dictionary. Retrieved 2 October 2026, from https://www.oed.com/dictionary/question-mark_n
- Real Academia Española. (n.d.). Ortografía de los signos de interrogación y exclamación. Español al día. Retrieved 2 October 2026, from https://www.rae.es/espanol-al-dia/ortografia-de-los-signos-de-interrogacion-y-exclamacion
- Real Academia Española & Asociación de Academias de la Lengua Española. (2010a). Los signos de interrogación y exclamación (§ 3.4.9). In Ortografía de la lengua española [Online edition]. https://www.rae.es/ortograf%C3%ADa/los-signos-de-interrogaci%C3%B3n-y-exclamaci%C3%B3n
- Real Academia Española & Asociación de Academias de la Lengua Española. (2010b). Usos especiales de los signos de interrogación y exclamación (§ 3.4.9.2). In Ortografía de la lengua española [Online edition]. https://www.rae.es/ortograf%C3%ADa/usos-especiales-de-los-signos-de-interrogaci%C3%B3n-y-exclamaci%C3%B3n
- Ridao Rodrigo, S. (2022). Los signos de puntuación de la interrogación y la exclamación en la ortografía de la Real Academia Española: Consideraciones didácticas e historiográficas. Moenia, 27. https://doi.org/10.15304/moenia.id7620
- Unicode Consortium. (2026a). Arabic: Range 0600–06FF [Code chart]. The Unicode Standard, Version 18.0. https://www.unicode.org/charts/PDF/U0600.pdf
- Unicode Consortium. (2026b). C0 controls and Basic Latin: Range 0000–007F [Code chart]. The Unicode Standard, Version 18.0. https://www.unicode.org/charts/PDF/U0000.pdf
- Unicode Consortium. (2026c). C1 controls and Latin-1 Supplement: Range 0080–00FF [Code chart]. The Unicode Standard, Version 18.0. https://www.unicode.org/charts/PDF/U0080.pdf
- Unicode Consortium. (2026d). Greek and Coptic: Range 0370–03FF [Code chart]. The Unicode Standard, Version 18.0. https://www.unicode.org/charts/PDF/U0370.pdf
- Unicode Consortium. (2026e). Supplemental punctuation: Range 2E00–2E7F [Code chart]. The Unicode Standard, Version 18.0. https://www.unicode.org/charts/PDF/U2E00.pdf
- University of Cambridge. (2011, July 21). The riddle of the Syriac double dot: It's the world's earliest question mark. https://www.cam.ac.uk/research/news/the-riddle-of-the-syriac-double-dot-its-the-worlds-earliest-question-mark