The circumflex (◌̂) is a diacritic in the Latin and Greek scripts that is also used in the written forms of many languages and in various romanization and transcription schemes. It received its English name from Latin: circumflexus "bent around"—a translation of the Greek: περισπωμένη (perispōménē).
The circumflex in the Latin script is chevron-shaped (◌̂), while the Greek circumflex may be displayed either like a tilde (◌̃) or like an inverted breve (◌̑). For the most commonly encountered uses of the accent in the Latin alphabet, precomposed characters are available.
In English, the circumflex, like other diacritics, is sometimes retained on loanwords that used it in the original language (for example, crème brûlée).
In mathematics and statistics, the circumflex diacritic is sometimes used to denote a function and is called a hat operator.
A free-standing version of the circumflex symbol, ^, is encoded in ASCII and Unicode and has become known as caret and has acquired special uses, particularly in computing and mathematics. The original caret, ‸, is used in proofreading to indicate insertion.
The shape of the circumflex was originally a combination of the acute and grave accents (^), as it marked a syllable contracted from two vowels: an acute-accented vowel and a non-accented vowel (all non-accented syllables in Ancient Greek were once marked with a grave accent).[1][clarification needed] Later a variant similar to the tilde (~) was also used.
The term "circumflex" is also used to describe similar tonal accents that result from combining two vowels in related languages such as Sanskrit and Latin.
Since Modern Greek has a stress accent instead of a pitch accent, the circumflex has been replaced with an acute accent in the modern monotonic orthography.
In Afrikaans, the circumflex marks a vowel with a lengthened pronunciation, often arising from compensatory lengthening due to the loss of ⟨g⟩ from the original Dutch form. Examples of circumflex use in Afrikaans are sê "to say", wêreld "world", môre "tomorrow", brûe "bridges".
Akkadian. In the transliteration of this language, the circumflex indicates a long vowel resulting from an aleph contraction.
In western Cree, Sauk, and Saulteaux, the Algonquianist Standard Roman Orthography (SRO) indicates long vowels [aː eː iː oː~uː] either with a circumflex ⟨â ê î ô⟩ or with a macron ⟨ā ē ī ō⟩.
The PDA orthography for Domari uses circumflex-bearing vowels for length.
In Emilian, â î û are used to represent [aː, iː, uː]
Japanese. In the Nihon-shiki system of romanization, the circumflex is used to indicate long vowels. The Kunrei-shiki system, which is based on Nihon-shiki system, also uses the circumflex. The Traditional and Modified forms of the Hepburn system use the macron for this purpose, though some users may use the circumflex as a substitute if there are difficulties inputting the macron, as the two diacritics are visually similar.
In UNGEGN romanization system for Khmer, â is used to represent [ɑː], ê[ae] in first series and [ɛː] in second series, and ô for [ɔː]. There are also additional vowels which are diphthongs such as aô[ao], âu[ʔɨw], âm[ɑm], ŏâm[oəm] and aôh[ɑh].
Bilingual sign showing the use of the circumflex in Welsh as an indicator of length and stress: parêd [paˈreːd] "parade", as opposed to pared [ˈparɛd] "partition wall".
The circumflex accent marks the stressed vowel of a word in some languages:
Portugueseâ, ê, and ô are stressed "closed" vowels, opposed to their open counterparts á, é, and ó (see below).
Welsh: the circumflex, due to its function as a disambiguating lengthening sign (see above), is used in polysyllabic words with word-final long vowels. The circumflex thus indicates the stressed syllable (which would normally be on the penultimate syllable), since in Welsh, non-stressed vowels may not normally be long. This happens notably where the singular ends in an a, to, e.g. singular camera, drama, opera, sinema → plural camerâu, dramâu, operâu, sinemâu; however, it also occurs in singular nominal forms, e.g. arwyddocâd; in verbal forms, e.g. deffrônt, cryffânt; etc.
Vowel quality
In Breton, it is used on an e to show that the letter is pronounced open instead of closed.
In Bulgarian, the sound represented in Bulgarian by the Cyrillic letter ъ (er goljam) is usually transliterated as â in systems used prior to 1989. Although called a schwa (misleadingly suggesting an unstressed lax sound), it is more accurately described as a mid back unrounded vowel/ɤ/. Unlike English or French, but similar to Romanian and Afrikaans, it can be stressed.
In Pinyin romanized Mandarin Chinese, ê is used to represent the sound /ɛ/ in isolation, which occurs sometimes as an exclamation.
In French, the letter ê is normally pronounced open, like è. In the usual pronunciations of central and northern France, ô is pronounced close, like eau; in Southern France, no distinction is made between close and openo.
Portugueseâ/ɐ/, ê/e/, and ô/o/ are stressed high vowels, in opposition to á/a/, é/ɛ/, and ó/ɔ/, which are stressed low vowels.
In Romanian, the circumflex is used on the vowels â and î to mark the vowel /ɨ/, similar to Russian yery. The names of these accented letters are â din a and î din i, respectively. (The letter â only appears in the middle of words; thus, its majuscule version appears only in all-capitals inscriptions.)
In Swedishdialect and folkloreliterature the circumflex is used to indicate the phonemes /a(ː)/ or /æ(ː)/(â), /ɶ(ː)/ or /ɞ(ː)/ (ô) and /ɵ(ː)/ (û) in dialects and regional accents where these are distinct from /ɑ(ː)/ (a), /ø(ː)/ (ö) or /o(ː)/ (o or å) and /ʉ(ː)/ (u) respectively, unlike Standard Swedish where [a] and [ɑː], [ɵ] and [ʉː] are short and long allophones of the phonemes /a/ and /ʉ/ respectively, and where Old Swedish short /o/ (ŏ) has merged with /o(ː)/ from Old Swedish /ɑː/ (ā, Modern Swedish å) instead of centralizing to [ɞ] or fronting to [ɶ] and remaining a distinct phoneme (ô) as in the dialects in question. Different methods can be found in different literature, so some author may use æ instead of â, or use â where others use å̂ (å with a circumflex; for a sound between /ɑ(ː)/ and /o(ː)/).
Vietnameseâ/ə/, ê/e/, and ô/o/ are higher vowels than a/ɑ/, e/ɛ/, and o/ɔ/. The circumflex can appear together with a tone mark on the same vowel, as in the word Việt. Vowels with circumflex are considered separate letters from the base vowels.
Nasality
In Luxembourgishm̂ n̂ can be used to indicate nasalisation of a vowel. Also, the circumflex can be over the vowel to indicate nasalisation. In either case, the circumflex is rare.
In Emilian, ê ô[eː, oː] denote both length and height. In Romagnol, they are used to represent the diphthongs /eə, oə/, whose specific articulation varies between dialects, e.g. sêl[seəl~seɛl~sæɛl~sɛɘl] "salt".
In Philippine languages, the circumflex (pakupyâ) is used to represent the simultaneous occurrence of a stress and a glottal stop in the last vowel of the word.[4][5]
In Old Tupi, the circumflex changed a vowel into a semivowel: î[j], û[w], and ŷ[ɰ].
In Rusyn, the letter ŷ[ɨ] is sometimes used to transliterate the Cyrillicы.
In Turkish, the circumflex over a and u is sometimes used in words of Arabic or Persian derivation to indicate when a preceding consonant (k, g, l) is to be pronounced as a palatal plosive; [c], [ɟ] (kâğıt, gâvur, mahkûm, Gülgûn). The circumflex over i is used to indicate a nisba suffix (millî, dinî).[6]
Visual discrimination between homographs
In Serbo-Croatian the circumflex can be used to distinguish homographs, and it is called the "genitive sign" or "length sign". Examples include sam "am" versus sâm "alone". For example, the phrase "I am alone" may be written Ja sam sâm to improve clarity. Another example: da "yes", dâ "gives".[7]
Turkish. According to Turkish Language Association orthography, düzeltme işareti "correction mark" over a, i and u marks a long vowel to disambiguate similar words. For example, compare ama "but" and âmâ "blind", şura 'that place, there' and şûra "council".[6] In general, circumflexes occur only in Arabic and Persianloanwords as vowel length in early Turkish was not phonemic. However, this standard was never applied entirely consistently[8] and by the late 20th century many publications had stopped using circumflexes almost entirely.[9]
Welsh. The circumflex is known as hirnod "long sign" or acen grom "crooked accent", but more usually and colloquially as to bach "little roof". It lengthens a stressed vowel (a, e, i, o, u, w, y), and is used particularly to differentiate between homographs; e.g. tan and tân, ffon and ffôn, gem and gêm, cyn and cŷn, or gwn and gŵn. However the circumflex is only required on elongated vowels if the same word exists without the circumflex - "nos" (night), for example, has an elongated "o" sound but a circumflex is not required as the same word with a shortened "o" doesn't exist.
In Pinyin, the romanized writing of Mandarin Chinese, ẑ, ĉ, and ŝ are, albeit rarely, used to represent zh[tʂ], ch[tʂʰ], and sh[ʂ], respectively.
In Esperanto, the circumflex is used on ĉ[tʃ], ĝ[dʒ], ĥ[x], ĵ[ʒ], ŝ[ʃ]. Each indicates a different consonant from the unaccented form, and is considered a separate letter for purposes of collation. (See Esperanto orthography.)
In Chichewa, ŵ (present for example in the name of the country Malaŵi) used to denote the voiced bilabial fricative/β/; nowadays, however, most Chichewa-speakers pronounce it as a regular [w].[10]
In the African language Venda, a circumflex below d, l, n, and t is used to represent dental consonants: ḓ, ḽ, ṋ, ṱ.
In the 18th century, the Real Academia Española introduced the circumflex accent in Spanish to mark that a ch or x were pronounced /k/ and /ɡs/ respectively (instead of /tʃ/ and /x/, which were the default values): châracteres, exâcto (spelled today caracteres, exacto). This usage was quickly abandoned during the same century, once the RAE decided to use ch and x with one assigned pronunciation only: /tʃ/ and /ɡs/ respectively.
In Domari (according to the Pan-Domari Alphabet orthography), the circumflex is used on the letters <ĉ ĝ ĵ ŝ ẑ> to represent the sounds of /t͡ʃ ɣ d͡ʒ ʃ ʒ/. It is also used above vowels to indicate length.
Abbreviation, contraction, and disambiguation
English
In 18th century British English, before the cheap Penny Post and while paper was taxed, the combination ough was occasionally shortened to ô when the gh was not pronounced, to save space: thô for though, thorô for thorough, and brôt for brought.
In French, the circumflex generally marks the former presence of a consonant (usually s) that was deleted and is no longer pronounced. (The corresponding Norman French words, and consequently the words derived from them in English, frequently retain the lost consonant.) For example:
ancêtre "ancestor"
hôpital "hospital"
hôtel "hostel"
forêt "forest"
rôtir "to roast"
côte "rib, coast, slope"
pâté "paste"
août "August"
dépôt (from the Latin depositum 'deposit', but now referring to both a deposit or a storehouse of any kind)[12]
Some homophones (or near-homophones in some varieties of French) are distinguished by the circumflex. However, â, ê and ô distinguish different sounds in most varieties of French, for instance cote[kɔt] "level, mark, code number" and côte[kot] "rib, coast, hillside".
In handwritten French, for example in taking notes, an m with a circumflex (m̂) is an informal abbreviation for même "same".
In February 2016, the Académie française decided to remove the circumflex from about 2,000 words, a plan that had been outlined since 1990. However, usage of the circumflex would not be considered incorrect.[13]
Italian
In Italian, î is occasionally used in the plural of nouns and adjectives ending with -io[jo] as a crasis mark. Other possible spellings are -ii and obsolete -j or -ij. For example, the plural of vario[ˈvaːrjo] "various" can be spelt vari, varî, varii; the pronunciation will usually stay [ˈvaːri] with only one [i]. The plural forms of principe[ˈprintʃipe] "prince" and of principio[prinˈtʃiːpjo] "principle, beginning" can be confusing. In pronunciation, they are distinguished by whether the stress is on the first or on the second syllable, but principi would be a correct spelling of both. When necessary to avoid ambiguity, it is advised to write the plural of principio as principî or as principii.[citation needed]
Latin
In Neo-Latin, circumflex was used most often to disambiguate between forms of the same word that used a long vowel, for example ablative of first declension and genitive of fourth declension, or between second and third conjugation verbs. It was also used for the interjection ô.[14]
Norwegian
In Norwegian, the circumflex differentiates fôr "lining, fodder" from the preposition for. From a historical point of view, the circumflex also indicates that the word used to be spelled with the letter ð in Old Norse – for example, fôr is derived from fóðr, lêr 'leather' from leðr, and vêr "weather, ram" from veðr (both lêr and vêr only occur in the Nynorsk spelling; in Bokmål these words are spelled lær and vær). After the ð disappeared, it was replaced by a d (fodr, vedr).
Portuguese
Circumflexes are used in many common words of the language, such as você (you), ânimo (cheer), and avô (grandfather). In early literacy classes in school, it is commonly nicknamed chapéu ("hat").
In mathematics, the circumflex is used to modify variable names; it is usually read "hat", e.g., î is "i hat". The Fourier transform of a function ƒ is often denoted by .
In the notation of sets, a hat above an element signifies that the element was removed from the set, such as in , the set containing all elements except .
In geometry, a hat is sometimes used for an angle. For instance, the angles or .
In statistics, the hat is used to denote an estimator or an estimated value, as opposed to its theoretical counterpart. For example, in errors and residuals, the hat in indicates an observable estimate (the residual) of an unobservable quantity called (the statistical error). It is read x-hat or x-roof, where x represents the character under the hat.
The precomposed charactersÂ/â, Ê/ê, Î/î, Ô/ô, and Û/û (which incorporate the circumflex) are included in the ISO-8859-1 character set, and dozens more are available in Unicode. In addition, Unicode has U+0302◌̂COMBINING CIRCUMFLEX ACCENT and U+032D◌̭COMBINING CIRCUMFLEX ACCENT BELOW which in principle allow adding the diacritic to any base letter.
The Greek diacriticπερισπωμένη, perispōménē, 'twisted around' is encoded as U+0342͂COMBINING GREEK PERISPOMENI.
For historical reasons, there is a similar but larger character, U+005E^CIRCUMFLEX ACCENT (^) (^ in HTML5[15]), which is also included in ASCII but often called a caret instead (though this term has a long-standing meaning as a proofreader's mark, with its own codepoints in Unicode). It is, however, unsuitable for use as a diacritic on modern computer systems, as it is a spacing character. Two other spacing circumflex characters in Unicode are the smaller modifier lettersU+02C6ˆMODIFIER LETTER CIRCUMFLEX ACCENT and U+A788ꞈMODIFIER LETTER LOW CIRCUMFLEX ACCENT, mainly used in phonetic notations or as a sample of the diacritic in isolation.
Typing the circumflex accent
You may need rendering support to display the uncommon Unicode characters in this section correctly.
In countries where the local language(s) routinely include letters with a circumflex, local keyboards are typically engraved with those symbols.
For users with American or British QWERTY keyboards, the characters â, ĉ, ê, ĝ, ĥ, î, ĵ, ô, ŝ, û, ẃ, ý (and their uppercase equivalents) may be obtained after installing the International or extended keyboard layout setting. Then, by using (US Int) ⇧ Shift+6 or (UK Ext) AltGr+6 (^), then release, then the base letter, produces the accented version. (With this keyboard mapping, ⇧ Shift+6 or AltGr+6 becomes a dead key that applies the diacritic to the subsequent letter, if such a precomposed character exists. For example, AltGr+6w produces ŵ as used in Welsh.) Alternatively for systems with a 'compose' function, compose^w, etc. may be used.
^Smyth, Herbert Weir (1920). A Greek Grammar for Colleges. New York: American Book Company – via ccel.org.: "155. The ancients regarded the grave originally as belonging to every syllable not accented with the acute or circumflex; and some Mss. show this in practice, e.g. πὰγκρὰτής. [...]"
^Halawa, T.; Harefa, A.; Silitonga, M. (1983). Struktur Bahasa Nias [Nias Language Structure] (PDF) (in Indonesian). Jakarta: Pusat Pembinaan dan Pengembangan Bahasa, Departemen Pendidikan dan Kebudayaan – via repositori.kemdikbud.go.id.
^"Dépôt". Larousse (in French). Retrieved 8 December 2016.