Unicode subscripts and superscripts

Unicode has subscripted and superscripted versions of a number of characters including a full set of Arabic numerals.^[1] These characters allow any polynomial, chemical and certain other equations to be represented in plain text without using any form of markup like HTML or TeX.

The World Wide Web Consortium and the Unicode Consortium have made recommendations on the choice between using markup and using superscript and subscript characters:

When used in mathematical context (MathML) it is recommended to consistently use style markup for superscripts and subscripts […] However, when super and sub-scripts are to reflect semantic distinctions, it is easier to work with these meanings encoded in text rather than markup, for example, in phonetic or phonemic transcription.^[2]

The intended use^[2] when these characters were added to Unicode was to produce true superscripts and subscripts so that chemical and algebraic formulas could be written without markup. Thus "H₂O" (using a subscript 2 character) is supposed to be identical to "H₂O" (with subscript markup).

In reality, many fonts that include these characters ignore the Unicode definition, and instead design the digits for mathematical numerator and denominator glyphs,^[3][4] which are aligned with the cap line and the baseline, respectively. When used with the solidus, these glyphs are a common substitute for diagonal fractions, such as ³/₄ for the ¾ glyph. This change was made because using markup does not give a good graphic approximation of fractions (compare markup ³/₄ with super/sub-script ³/₄). The change also makes the superscript letters useful for ordinal indicators, more closely matching the ª and º characters. However, it makes them incorrect for normal superscript and subscript, and so chemical and algebraic formulas are better rendered by using markup.

Unicode intended that diagonal fractions be rendered by a different mechanism: the fraction slash U+2044 is visually similar to the solidus, but when used with the ordinary digits (not the superscripts and subscripts), it instructs the layout system that a fraction such as ¾ is to be rendered using automatic glyph substitution.^[5][a] User-end support was quite poor for a number of years, but fonts, browsers,^[b] word processors,^[c] desktop publishing software^[d] and others increasingly support the intended Unicode behavior.

A selection of supporting fonts is displayed in the table below. (These will not display properly if you do not have the fonts installed, or if your browser does not support this behavior.)

The most common superscript digits (1, 2, and 3) were in ISO-8859-1 and were therefore carried over into those positions in the Latin-1 range of Unicode. The rest were placed in a dedicated section of Unicode at U+2070 to U+209F. The two tables below show these characters. Each superscript or subscript character is preceded by a normal x to show the subscripting/superscripting. The table on the left contains the actual Unicode characters; the one on the right contains the equivalents using HTML markup for the subscript or superscript.

Unicode version 16.0 also includes subscript and superscript characters that are intended for semantic usage, in the following blocks:^[1][6]

Superscript

• The Latin-1 Supplement block contains the feminine and masculine ordinal indicators ª and º.
• The Latin Extended-C block contains one superscript, ⱽ.
• The Latin Extended-D block contains six superscripts: ꝰ ꟲ ꟳ ꟴ ꟸ ꟹ.
• The Latin Extended-E block contains five superscripts: ꭜ ꭝ ꭞ ꭟ ꭩ.
• The Latin Extended-F block is entirely superscript IPA letters: 𐞁 𐞂 𐞃 𐞄 𐞅 𐞇 𐞈 𐞉 𐞊 𐞋 𐞌 𐞍 𐞎 𐞏 𐞐 𐞑 𐞒 𐞓 𐞔 𐞕 𐞖 𐞗 𐞘 𐞙 𐞚 𐞛 𐞜 𐞝 𐞞 𐞟 𐞠 𐞡 𐞢 𐞣 𐞤 𐞥 𐞦 𐞧 𐞨 𐞩 𐞪 𐞫 𐞬 𐞭 𐞮 𐞯 𐞰 𐞲 𐞳 𐞴 𐞵 𐞶 𐞷 𐞸 𐞹 𐞺.
• The Spacing Modifier Letters block has superscripted letters and symbols used for phonetic transcription: ʰ ʱ ʲ ʳ ʴ ʵ ʶ ʷ ʸ ˀ ˁ ˠ ˡ ˢ ˣ ˤ.
• The Phonetic Extensions block has several superscripted letters and symbols: Latin/IPA ᴬ ᴭ ᴮ ᴯ ᴰ ᴱ ᴲ ᴳ ᴴ ᴵ ᴶ ᴷ ᴸ ᴹ ᴺ ᴻ ᴼ ᴽ ᴾ ᴿ ᵀ ᵁ ᵂ ᵃ ᵄ ᵅ ᵆ ᵇ ᵈ ᵉ ᵊ ᵋ ᵌ ᵍ ᵏ ᵐ ᵑ ᵒ ᵓ ᵖ ᵗ ᵘ ᵚ ᵛ, Greek ᵝ ᵞ ᵟ ᵠ ᵡ, Cyrillic ᵸ, other ᵎ ᵔ ᵕ ᵙ ᵜ. These are intended to indicate secondary articulation.
• The Phonetic Extensions Supplement block has several more: Latin/IPA ᶛ ᶜ ᶝ ᶞ ᶟ ᶠ ᶡ ᶢ ᶣ ᶤ ᶥ ᶦ ᶧ ᶨ ᶩ ᶪ ᶫ ᶬ ᶭ ᶮ ᶯ ᶰ ᶱ ᶲ ᶳ ᶴ ᶵ ᶶ ᶷ ᶸ ᶹ ᶺ ᶻ ᶼ ᶽ ᶾ, Greek ᶿ.
• The Cyrillic Extended-B block contains two Cyrillic superscripts: ꚜ ꚝ.
• The Cyrillic Extended-D block contains many Cyrillic superscripts: 𞀰 𞀱 𞀲 𞀳 𞀷 𞀵 𞀶 𞀷 𞀸 𞀹 𞀺 𞀻 𞀼 𞀽 𞀾 𞀿 𞁀 𞁁 𞁂 𞁃 𞁄 𞁅 𞁆 𞁇 𞁈 𞁉 𞁊 𞁋 𞁌 𞁍 𞁎 𞁏 𞁐 𞁫 𞁬 𞁭.
• The Georgian block contains one superscripted Mkhedruli letter: ჼ.
• The Kanbun block has superscripted annotation characters used in Japanese copies of Classical Chinese texts: ㆒ ㆓ ㆔ ㆕ ㆖ ㆗ ㆘ ㆙ ㆚ ㆛ ㆜ ㆝ ㆞ ㆟.
• The Tifinagh block has one superscript letter : ⵯ.
• The Unified Canadian Aboriginal Syllabics and its Extended blocks contain several mostly consonant-only letters to indicate syllable coda called Finals, along with some characters that indicate syllable medial known as Medials: Main block ᐜ ᐝ ᐞ ᐟ ᐠ ᐡ ᐢ ᐣ ᐤ ᐥ ᐦ ᐧ ᐨ ᐩ ᐪ ᑉ ᑊ ᑋ ᒃ ᒄ ᒡ ᒢ ᒻ ᒼ ᒽ ᒾ ᓐ ᓑ ᓒ ᓪ ᓫ ᔅ ᔆ ᔇ ᔈ ᔉ ᔊ ᔋ ᔥ ᔾ ᔿ ᕀ ᕁ ᕐ ᕑ ᕝ ᕪ ᕻ ᕯ ᕽ ᖅ ᖕ ᖖ ᖟ ᖦ ᖮ ᗮ ᘁ ᙆ ᙇ ᙚ ᙾ ᙿ; Extended block: ᣔ ᣕ ᣖ ᣗ ᣘ ᣙ ᣚ ᣛ ᣜ ᣝ ᣞ ᣟ ᣳ ᣴ ᣵ.

Combining superscript

• The Combining Diacritical Marks block contains medieval superscript letter diacritics. These letters are written directly above other letters appearing in medieval Germanic manuscripts, and so these glyphs do not include spacing, for example uͤ. They are shown here over the dotted circle placeholder ◌: ◌ͣ ◌ͤ ◌ͥ ◌ͦ ◌ͧ ◌ͨ ◌ͩ ◌ͪ ◌ͫ ◌ͬ ◌ͭ ◌ͮ ◌ͯ.
• The Combining Diacritical Marks Extended block contains three combining insular letters for the Middle English Ormulum, ◌ᫌ ◌ᫍ ◌ᫎ.^[7]
• The Combining Diacritical Marks Supplement block contains additional medieval superscript letter diacritics, enough to complete the basic lowercase Latin alphabet except for j, q and y, a few small capitals and ligatures (ae, ao, av), and additional letters: ◌᷒ ◌ᷓ ◌ᷔ ◌ᷕ ◌ᷖ ◌ᷗ ◌ᷘ ◌ᷙ ◌ᷚ ◌ᷛ ◌ᷜ ◌ᷝ ◌ᷞ ◌ᷟ ◌ᷠ ◌ᷡ ◌ᷢ ◌ᷣ ◌ᷤ ◌ᷥ ◌ᷦ ◌ᷧ ◌ᷨ ◌ᷪ ◌ᷫ ◌ᷬ ◌ᷭ ◌ᷮ ◌ᷯ ◌ᷰ ◌ᷱ ◌ᷲ ◌ᷳ ◌ᷴ, Greek ◌ᷩ.
• The Cyrillic Extended-A and -B blocks contains multiple medieval superscript letter diacritics, enough to complete the basic lowercase Cyrillic alphabet used in Church Slavonic texts, also includes an additional ligature (ст): ◌ⷠ ◌ⷡ ◌ⷢ ◌ⷣ ◌ⷤ ◌ⷥ ◌ⷦ ◌ⷧ ◌ⷨ ◌ⷩ ◌ⷪ ◌ⷫ ◌ⷬ ◌ⷭ ◌ⷮ ◌ⷯ ◌ⷰ ◌ⷱ ◌ⷲ ◌ⷳ ◌ⷴ ◌ⷵ ◌ⷶ ◌ⷷ ◌ⷸ ◌ⷹ ◌ⷺ ◌ⷻ ◌ⷼ ◌ⷽ ◌ⷾ ◌ⷿ ◌ꙴ ◌ꙵ ◌ꙶ ◌ꙷ ◌ꙸ ◌ꙹ ◌ꙺ ◌ꙻ ◌ꚞ ◌ꚟ.
• The Cyrillic Extended-D block has one additional combining character, that being і: ◌𞂏.

Subscript

• The Latin Extended-C block contains one subscript, ⱼ.
• The Phonetic Extensions block has several subscripted letters and symbols: Latin/IPA ᵢ ᵣ ᵤ ᵥ and Greek ᵦ ᵧ ᵨ ᵩ ᵪ.
• The Cyrillic Extended-D block also contains many Cyrillic subscripts: 𞁑 𞁒 𞁓 𞁔 𞁕 𞁖 𞁗 𞁘 𞁙 𞁚 𞁛 𞁜 𞁝 𞁞 𞁟 𞁠 𞁡 𞁢 𞁣 𞁤 𞁥 𞁦 𞁧 𞁨 𞁩 𞁪.

Combining subscript

• The Combining Diacritical Marks Supplement block contains a combining subscript: ◌᷊.
• The Combining Diacritical Marks Extended block contains two combining letters for linguistic transcriptions of Scots, ◌ᪿ ◌ᫀ.

Consolidated, the Unicode standard contains superscript and subscript versions of a subset of Latin, Greek and Cyrillic letters. Here they are arranged in alphabetical order for comparison (or for copy and paste convenience). Since these characters appear in different Unicode ranges, they may not appear to be the same size or position due to font substitution in the browser. Shaded cells mark small capitals that are not very distinct from minuscules, and Greek letters that are indistinguishable from Latin, and so would not be expected to be supported by Unicode.

Little punctuation is encoded. Parentheses are shown above in the basic block above, and the exclamation mark ⟨ꜝ⟩ is shown in the IPA table below. A question mark may be created with a superscript gelded question mark and a combining dot: ⟨ˀ̣⟩, although some fonts do not render it properly.

Additional superscript capitals are ᴭ ᴯ ᴲ ᴻ. Some of these are small caps in the source documents in the Unicode proposals.
Superscript capital s has been proposed for a future version of the Unicode Standard.^[8][9]
Superscript versons of small capital A and E have been proposed for a future version of the Unicode Standard.^[10][11][9]

Superscript versons of Greek psi and omega have been proposed for a future version of the Unicode Standard.^[10][9]

Many of the Cyrillic characters were added to the Cyrillic Extended-D block, which was added to the free Gentium Plus and Andika fonts with version 6.2 in February 2023.

See also small caps in Unicode.

The Latin Extended-F block was created for the remaining superscript IPA letters. They are supported by the free Gentium Plus and Andika fonts. Additional superscript characters for historical and para-IPA letters have been proposed for future versions of the Unicode Standard.^[11][9]

The Unicode characters for superscript (modifier) IPA and extIPA consonant letters are as follows. The entire Latin Extended-F block is dedicated to superscript IPA. Characters for sounds with secondary articulation are set off in parentheses and placed below the base letters.

The spacing diacritic for ejective consonants, U+2BC, works with superscript letters despite not being superscript itself: ⟨ᵖʼ ᵗʼ ᶜʼ ᵏˣʼ⟩. If a distinction needs to be made, the combining apostrophe U+315 may be used: ⟨ᵖ̕ ᵗ̕ ᶜ̕ ᵏˣ̕⟩. The spacing diacritic should be used for a baseline letter with a superscript release, such as [tˢʼ] or [kˣʼ], where the scope of the apostrophe includes the non-superscript letter, but the combining apostrophe U+315 might be used to indicate a weakly articulated ejective consonant like [ᵗ̕] or [ᵏ̕], where the whole consonant is written as a superscript, or together with U+2BC when separate apostrophes have scope over the base and modifier letters, as in ⟨pʼᵏˣ̕⟩.^[14]

Spacing diacritics, as in ⟨tʲ⟩, cannot be secondarily superscripted in plain text: ⟨ᵗʲ⟩. (In this instance, the old IPA letter for [tʲ], ⟨ƫ⟩, has a superscript variant in Unicode, U+1DB5 ⟨ᶵ⟩, but that is not generally the case.)

Among older letters, ⟨ꜧ⟩ (U+A727) was a graphic variant of ⟨ɮ⟩. Its superscript is supported at ⟨ꭜ⟩ (U+AB5C). The most common letters with palatal hook are also supported; they are displayed in the table above. IPA once had an idiosyncratic curl on some of the palatalized letters: these are the fricative letters ⟨ʆ ʓ⟩. Their superscript forms have been proposed for a future version of the Unicode Standard.^[11][9] The retired letters ⟨ƞ⟩ and ⟨ɼ⟩ have also been proposed for a future version of the Unicode Standard.^[11][9]

Among para-IPA letters, Sinological superscript ⟨ȡ ȴ ȵ ȶ⟩ have been proposed for a future version of the Unicode Standard.^[10][9] Superscripts of the Bantuist labio-dental plosives ⟨ȹ⟩ and ⟨ȸ⟩ have been proposed for a future version of the Unicode Standard.^[10][9] The central semivowels ⟨ɉ⟩, ɥ̶, and w̶ have also been proposed for a future version of the Unicode Standard.^[10][9]

Old-style click letters have been proposed for a future version of the Unicode Standard.^[15][9]

The Unicode characters for superscript (modifier) IPA vowel letters, plus a pair of extended letters ⟨ᵻ ᵿ⟩ found in English dictionaries, are as follows. Recently retired alternative letters such as ⟨ɩ ɷ⟩ are also supported; they are set off in parentheses and placed below the standard IPA letters:

The precomposed Unicode rhotic vowel letters ⟨ɚ ɝ⟩ are not directly supported. The rhotic diacritic U+02DE ◌˞ should be used instead: ⟨ᵊ˞ ᶟ˞⟩.^[16]

⟨ɜ⟩ and ⟨ᶟ⟩ are reversed ɛ. The older IPA turned ɛ, ⟨ᴈ⟩, is also supported, at U+1D4C ⟨ᵌ⟩. However, the briefly resurrected vowel letter ⟨ʚ⟩ (U+029A) is not supported, only its reversed replacement ⟨ɞ⟩ is.

Among older letters, ⟨ᴜ⟩ (U+1D1C), a graphic variant of ⟨ʊ⟩, is supported at ⟨ᶸ⟩ (U+1DB8).

Among para-IPA letters, Sinological superscript ⟨ɿ ʅ ʮ ʯ⟩ have been proposed for a future version of the Unicode Standard.^[10][9]

The two length marks are also supported:

These are used to add length to another superscript, such as long aspiration.

Superscript wildcards (full caps) are largely supported: e.g. ᴺC (prenasalized consonant), ꟲN (prestopped nasal), Pꟳ (fricative release), NᴾF (epenthetic plosive), CVNᵀ (tone-bearing syllable), Cᴸ (liquid or lateral release), Cᴿ (rhotic or resonant release), Vᴳ (off-glide/diphthong), Cⱽ (fleeting vowel). Superscript S for sibilant release has been proposed for a future version of the Unicode Standard;^[8][9] superscript Ʞ for fleeting/epenthetic click has not. Other basic Latin superscript wildcards for tone and weak indeterminate sounds, as described in the article on the International Phonetic Alphabet, are mostly supported. (See table in previous section.)

In addition, a very few IPA letters beyond the basic Latin alphabet have combining forms or are supported as subscripts:

Primarily for compatibility with earlier character sets, Unicode contains a number of characters that compose super- and subscripts with other symbols.^[1] In most fonts these render much better than attempts to construct these symbols from the above characters or by using markup.

• The Latin-1 Supplement block contains the precomposed fractions ½, ¼, and ¾. The copyright © and registered trademark signs ® are also in this block.
• The General Punctuation block contains the permille sign ‰ and the per-ten-thousand sign ‱, and Basic Latin has the percent sign %.
• The Number Forms block contains several precomposed fractions: ⅐ ⅑ ⅒ ⅓ ⅔ ⅕ ⅖ ⅗ ⅘ ⅙ ⅚ ⅛ ⅜ ⅝ ⅞ ⅟ ↉.
• The Letterlike Symbols block contains a few symbols composed of subscript and superscript characters: ℀ ℁ ℅ ℆ № ℠ ™ ⅍.
• The Enclosed Alphanumeric Supplement block contains three superscript abbreviations 🅪 🅫 🅬: MC for marque de commerce (trademark), MD for marque déposée (registered trademark), both used in Canada; MR for marca registrada (registered trademark) in Spanish and Portuguese speaking countries.^[17]
• The Miscellaneous Technical block has one additional subscript, a subscript 10 (⏨), for the purpose of scientific notation.
• The Unified Canadian Aboriginal Syllabics and its Extended blocks contain several letters composed with superscripted letters to indicate extended sound values: Main block ᐂ ᐫ ᐬ ᐭ ᐮ ᐰ ᑍ ᑧ ᑨ ᑩ ᑪ ᑬ ᒅ ᒆ ᒇ ᒈ ᒊ ᒤ ᓁ ᓔ ᓮ ᔌ ᔍ ᔎ ᔏ ᔧ ᕅ ᕔ ᕿ ᖀ ᖁ ᖂ ᖃ ᖄ ᖎ ᖏ ᖐ ᖑ ᖒ ᖓ ᖔ ᙯ ᙰ ᙱ ᙲ ᙳ ᙴ ᙵ ᙶ, Extended block ᢰ ᢱ ᢲ ᢳ ᢴ ᢵ ᢶ ᢷ ᢸ ᢹ ᢺ ᢻ ᢼ ᢽ ᢾ ᢿ ᣀ ᣁ ᣂ ᣃ ᣄ ᣅ.