Cedilla (¸): What the Hook Under ç Means and Its Comma Twin

What is a cedilla, and what does it do under the c?

A cedilla is a small hook written under a letter. Oxford (opens in a new tab) defines it as the mark placed under c in French, Portuguese and some other languages to show that the letter is said like an s rather than a k, as in français. It adds a second use: a similar mark under s in Turkish.

The French rule is short. Larousse (opens in a new tab) defines the cédille as the sign set under c before a, o or u to give it the sound [s], and its examples are façade, leçon and reçu. Without the hook, the c in those words would be read as a k.

Wiktionary (opens in a new tab) adds Portuguese açaí and Catalan força. The rules for each language, and the Turkish and Albanian ç that reads as ch, are on the c with cedilla page.

People often look it up as the cedilla accent. French grammar keeps it apart from the accents while treating it the same way. The Académie française (opens in a new tab) says an accent has full spelling value in French, and that the same goes for the tréma and the cédille. So all three stay on capital letters.

Unicode's name for the mark is U+0327 COMBINING CEDILLA, with no ACCENT in it. Its names list (opens in a new tab) notes French, Turkish and Azerbaijani beside the cedilla mark.

English can keep or drop the hook in borrowed words. Wiktionary's usage note gives façade and facade as both in use.

Where does the word cedilla come from?

From Spanish, where it means little z. Oxford dates the English word to the late 16th century and traces it to an obsolete Spanish form of zedilla, a diminutive of zeda, the letter z. Larousse gives the same root for the French cédille: Spanish cedilla, a little z, from Latin zeta. Wiktionary puts the English borrowing at about 1600.

The name describes where the shape came from. In Wiktionary's account, Spanish first used cedilla for the letter Ç itself, which grew out of Ꝣ, a Visigothic form of z. Over time the lower part of that z shrank into the hook. The upper part began as an ornament over the z and grew until it looked like a c, and readers came to take it for one.

So the story runs the other way from how the letter looks today: the hook is the old z, and the c is the part that was added.

Unicode encodes that older letter as U+A762 and U+A763, LATIN CAPITAL and SMALL LETTER VISIGOTHIC Z. Its names list sends readers from the small form straight to ç. Wiktionary's entry (opens in a new tab) describes Ꝣ as the letter Z in the Visigothic script.

Spanish no longer writes the mark it named. Wiktionary's entry for ç (opens in a new tab) marks the Spanish letter obsolete and says z has replaced it. The word stayed behind, and other languages borrowed it, as these names from Wiktionary's cedilla entry show:

Language	Name of the mark
French	cédille
Portuguese	cedilha
Italian	cediglia
German	Cedille
Romanian	sedilă
Turkish, Azerbaijani, Uzbek	sedil
Catalan	trencat

What job does the cedilla do, from French to romanized Arabic?

The keys at the top of this page are every letter in AccentPad's catalogue whose Unicode decomposition holds U+0327, and the usage map shows which CLDR alphabets write them. What the mark does changes from row to row. No single meaning holds across the table:

Letters	Where	What the cedilla does	Source
ç	French, Portuguese, Catalan	keeps the c an s before a, o and u	Larousse, Oxford
ç	Turkish, Azerbaijani, Albanian	a letter of its own for the ch sound	Wiktionary ç
ş	Turkish, Azerbaijani	a letter of its own for /ʃ/, the sh sound	Wiktionary ş
ģ ķ ļ ņ	Latvian	palatal consonants, drawn with a comma	NamesList, Wiktionary
ŗ ḑ	Livonian	/rʲ/ and /ɟ/, drawn with a comma	NamesList, Wiktionary
ţ	Gagauz	/t͡s/	Wiktionary ţ
m̧ o̧	Marshallese	/mˠ/ and /ɒ/; no single code point exists for either	Wiktionary, Unicode ch. 7
ş ţ ḑ ḩ	romanized Arabic	stand for ص, ط, ض and ح	Wiktionary
ḝ	Hebrew transliteration	an e with a cedilla and a breve	NamesList
ȩ	Slovene dialect transcription	an open e, [ɛ]	Wiktionary ȩ

The Arabic row comes from Wiktionary's entries for ş (opens in a new tab), ţ (opens in a new tab), ḑ (opens in a new tab) and ḩ (opens in a new tab), each marked as a romanization. The Livonian sounds come from the ḑ (opens in a new tab) and ŗ (opens in a new tab) entries, and the Marshallese ones from m̧ (opens in a new tab) and o̧ (opens in a new tab). The Slovene row is the ȩ (opens in a new tab) entry, which explains the mark as a sign of the more open vowel.

Several rows have a fuller page. The Turkish alphabet covers ç and ş as letters, the Latvian alphabet the sounds of ģ, ķ, ļ and ņ, and the r with accent page the Livonian ŗ.

Cedilla or comma below: why do ş and ș, ţ and ț look the same?

Because Unicode lets the cedilla take more than one shape. Chapter 7 of the Unicode Standard (opens in a new tab) says U+0327 COMBINING CEDILLA can be displayed in a wide variety of forms, commas below among them. The same freedom holds for every precomposed letter that breaks down into it. The chapter also admits that many fonts draw no clear difference between U+0326 COMBINING COMMA BELOW and the cedilla.

Each language has a preferred form. Per the same chapter, Latvian and Romanian prefer a comma below, while Turkish and Marshallese prefer a cedilla. When a font has no language to go on, Table 7-1 sets a default shape by the base letter:

Base letter	Default shape of U+0327 with no language given
c, e, h, s	a cedilla, the hook
d, g, k, l, n, r, t	a comma below

So with no language given, Unicode prefers a comma for ţ, the t with cedilla, which is the same look as Romanian ț, and a hook for ş. A shape on the screen is a font's choice, not proof of the character underneath.

Romanian is the language that got letters of its own: ș U+0219 and ț U+021B, built on the comma below. We normalized both pairs with Node.js 26.0.0 (ICU 78.3, Unicode 17.0) on 2026-10-14. s plus U+0326 composed into ș, and s plus U+0327 into ş. The names list tells writers to use ș instead of ş for Romanian.

Old code pages had one character for both: Unicode's mapping of ISO 8859-2 (opens in a new tab) gives byte BA as the s with cedilla, which Romanian text used for its comma s. Chapter 7 therefore asks software that processes Romanian to treat U+0219 and U+015F as equivalent. The s with accent and t with accent pages cover the code pages and how to repair a file.

Latvian took the other path and kept the cedilla code points. The names list says that despite their names, ķ, ļ and ņ should normally be displayed with a comma below; it says the same of Livonian ŗ and ḑ. The g is the odd one. Chapter 7 says good Latvian typography puts a rotated comma over the small ģ, because a cedilla cannot hang under the loop of the g. The capital Ģ always takes the cedilla.

Unicode has a mark in that shape, U+0312 COMBINING TURNED COMMA ABOVE, which the names list also calls cedilla above. It marks it Latvian but not used in decomposition. In our test, g plus U+0327 composed into ģ, while g plus U+0312 stayed two characters. A Latvian word built with U+0312 can look right on screen and still fail an exact search for the real ģ.

You are writing	Type	Its mark	What a good font shows
Turkish or Azerbaijani ş	U+015F	U+0327 cedilla	a hook
Romanian ș and ț	U+0219, U+021B	U+0326 comma below	a comma
Latvian ķ, ļ, ņ	U+0137, U+013C, U+0146	U+0327 cedilla	a comma below
Latvian ģ	U+0123	U+0327 cedilla	a turned comma above the g
Marshallese m̧	m, then U+0327	U+0327 cedilla	a hook

What is the cedilla symbol ¸, and how do you type it on its own?

The cedilla symbol by itself is U+00B8 CEDILLA. It is a spacing character: it takes up a width of its own and sits on no letter. The names list says so and points to U+0327 as the combining form. Chapter 7 explains the split: U+00B8 came from ISO/IEC 8859-1, where it was unclear whether it was a mark or a symbol, and Unicode uses it only as a spacing character. For accented letters it provides the combining marks.

Use ¸ when you write about the mark, as this page does, and never to build a letter. We checked what normalization does to each form, with the same Node.js 26.0.0 on 2026-10-14. NFC joins a letter and its combining mark; NFKC also folds compatibility forms.

Characters you have	After NFC	After NFKC
c, then U+0327	ç, one character, U+00E7	ç
c, then ¸ U+00B8	still c and ¸	c, a space and U+0327
¸ U+00B8 alone	unchanged	a space and U+0327
g, then U+0312	still g and the turned comma	unchanged
m, then U+0327	still m and U+0327: there is no single m̧	unchanged

As a control, n followed by U+0303 composed into ñ in the same run. So a c followed by a pasted ¸ is never ç, even after normalization: NFKC splits the symbol into a space and the mark, and the space keeps the mark off the c.

To type the symbol itself:

Where	How to get ¸	Source
Windows, on the number pad	hold Alt, type 0184, let go	cp1252 byte B8 is U+00B8
Mac, U.S. or ABC layout	hold Shift and Option, press Z	measured with UCKeyTranslate
Mac, any layout	the Character Viewer: search cedilla	Apple
Microsoft Word	type 00b8, then Alt+X	Microsoft
HTML	¸ or ¸	WHATWG

The Windows code is read from Unicode's mapping of cp1252 (opens in a new tab), where byte B8, decimal 184, is the cedilla. The Mac row is AccentPad's own measurement of Apple's U.S. and ABC layouts on macOS 26.6.2, and Apple's Keyboard Viewer (opens in a new tab) shows what each key types while you hold modifier keys such as Option and Shift. Word's code-then-Alt+X method is in Microsoft's guide to Unicode characters (opens in a new tab), and the entity name is in the WHATWG list (opens in a new tab).