About Quotation Mark
Quotation Mark is the Unicode character " at codepoint U+0022 in the Basic Latin block of the Basic Multilingual Plane (BMP).
Reference
The straight double quote: one glyph doing the work of an opening mark and a closing one. Unicode's chart calls it neutral (vertical), used as opening or closing quotation mark, and states outright that the preferred characters in English are U+201C and U+201D. It then gives it ten cross-references — to primes, accents, Hebrew gershayim and the curly pair, which is a fair measure of how much it is asked to stand in for. Its neutrality is the point: the same character at both ends is useless for showing where a quotation turns around and exactly right where a machine reads the text. Basic Latin, Unicode 1.1, general category Po, with the Quotation_Mark and Pattern_Syntax properties.
Typical uses
- String delimiters in JSON, most programming languages and shell commands
- Quoted fields in CSV, where a literal one inside a field is escaped by doubling it
- Quotation marks in plain text, and in publications that have chosen not to curl
At a glance
Historical Notes
Origins
Mechanical typewriters economised on keys: one vertical mark served as opening quote, closing quote and, in some layouts, more besides. ASCII inherited the key and the compromise, and every keyboard since has kept it.
Unicode
Unicode 1.1, Basic Latin, inherited wholesale from ASCII. What is interesting is the annotation, which is unusually opinionated — the standard describes the glyph as neutral and then names the characters English should use instead. It also singles out Hebrew, where U+05F4 GERSHAYIM is preferred over this mark.
The Pattern_Syntax property
Unicode gives this character the Pattern_Syntax property, which marks codepoints reserved for use as syntax in formal languages. That is the standard acknowledging, in machine-readable form, what the character has actually become: not a typographic mark that computing borrowed, but a piece of syntax that also happens to look like punctuation.
Developer Notes
| Unicode Codepoint | |
| Decimal | |
| UTF-8 (hex) | |
| UTF-16 (hex) | |
| HEX Entity | |
| HTML Entity | |
| CSS (\xxxx) | |
| JavaScript | |
| Python |
How to Use Quotation Mark
To use the Quotation Mark symbol (") in HTML, insert the
entity " directly into your markup. The decimal entity " works in any HTML-serialised document.
In CSS pseudo-elements use content: '\22';.
In JavaScript and TypeScript the escape '\u0022' evaluates to ". In Python source, write '\u0022'.
The character encodes to 22 as UTF-8 bytes and 0022 as UTF-16 code units. Its decimal codepoint is 34.
This is a text-based Unicode character. It will render in the current font on most platforms. If a font lacks this glyph, the browser may show a placeholder or fall back to a system font.
Things to know
- In CSV under RFC 4180, a field containing a comma or line break is wrapped in these marks and a literal one inside the field is doubled. An unbalanced quote thousands of rows up is the classic reason a file suddenly loads as one enormous column.
- A curly quote pasted into JSON, a JavaScript literal or a shell command is a syntax error, not a typographic infelicity.
- " is one of the few named entities that is also valid without its semicolon, and it has an uppercase form, ". The apostrophe has neither.
- Pattern_Syntax means the character is reserved for syntax in formal languages. Treat it as structure first.
- Smart-quote conversion in editors and CMSes silently turns this into U+201C or U+201D. Text that has passed through one and text that has not are different strings.
Comparisons
In the Wild
BBC News, on 4 August 2026, quotes with straight marks and does not curl at all: The star has taken a "step back" from the spotlight after appearances led to "endless, ongoing public scrutiny". A Fifa story on the same page reports that a decision was "absolutely necessary". The page carried 14 of these marks, 44 straight apostrophes, and zero curly quotes of any kind.
Der Spiegel's home page carried 1,019 of them alongside only four German „…“ pairs, which suggests the straight marks are largely inside the page's own machinery rather than in its prose.
English Wikipedia sets them in running text on purpose: 396 on the Electron article, 112 on Absolute zero, with zero curly double quotes on either. Both articles use the real minus sign for negative numbers, so the straight quotes are a house-style choice rather than indifference to typography. Pages change; this is what those showed on the day.
FAQ
Is it wrong to use straight quotes in prose?
Unicode names the curly pair as preferred for English. It is a preference, not a rule, and BBC News is one publication that does not follow it.
Why must I keep them in code?
Because parsers read this character. JSON has exactly one string delimiter and this is it.
How do I put one inside a CSV field?
Double it. Under RFC 4180 two of these marks inside a quoted field mean one literal character.
Is " the only entity?
There is also ", and both are valid without the semicolon — a legacy allowance that ' does not share.
What is the difference from a double prime?
U+2033 DOUBLE PRIME means inches, arcseconds and minutes of time, and is drawn as two slanted strokes. Unicode cross-references the two from this entry.