#

Chapter 6

Writing Systems and Punctuation

This chapter begins the portion of the Unicode Standard devoted to the detailed description of each script or other related group of Unicode characters. Each of the subsequent chapters presents a historically or geographically related group of scripts. This chapter presents a general introduction to writing systems, explains how they can be used to classify scripts, and then presents a detailed discussion of punctuation characters that are shared across scripts.

#Scripts and Blocks. The codespace of the Unicode Standard is divided into subparts called blocks (see D10b in Section 3.4, Characters and Encoding). Character blocks generally contain characters from a single script, and in many cases, a script is fully represented in its block; however, some scripts are encoded using several blocks, which are not always adjacent. Discussion of scripts and other groups of characters are structured by blocks. Corresponding subsection headers identify each block and its associated range of Unicode code points. The Unicode code charts are also organized by blocks.

#Scripts and Writing Systems. There are many different kinds of writing systems in the world. Their variety poses some significant issues for character encoding in the Unicode Standard as well as for implementers of the standard. Those who first approach the Unicode Standard without a background in writing systems may find the huge list of scripts bewilderingly complex. Therefore, before considering the script descriptions in detail, this chapter first presents a brief introduction to the types of writing systems. That introduction explains basic terminology about scripts and character types that will be used again and again when discussing particular scripts.

#Punctuation. The rest of this chapter deals with a special case: punctuation marks, which tend to be scattered about in different blocks and which may be used in common by many scripts. Punctuation characters occur in several widely separated places in the blocks, including Basic Latin, Latin-1 Supplement, General Punctuation, Supplemental Punctuation, and CJK Symbols and Punctuation. There are also occasional punctuation characters in blocks for specific scripts.

Most punctuation characters are intended for common usage with any script, although some of them are script-specific. Some scripts use both common and script-specific punctuation characters, usually as the result of recent adoption of standard Western punctuation marks. While punctuation characters vary in details of appearance and function between different languages and scripts, their overall purpose is shared: they serve to separate or otherwise organize units of text, such as sentences and phrases, thereby helping to clarify the meaning of the text. Certain punctuation characters also occur in mathematical and scientific formulae.

#6.1 Writing Systems

This section presents a brief introduction to writing systems. It describes the different kinds of writing systems and relates them to the encoded scripts found in the Unicode Standard. This framework may help to make the variety of scripts, modern and historic, a little less daunting. The terminology used here follows that developed by Peter T. Daniels, a leading expert on writing systems of the world.

The term writing system has two mutually exclusive meanings in this standard. As used in this section, “writing system” refers to a way that families of scripts may be classified by how they represent the sounds or words of human language. For example, the writing system of the Latin script is alphabetic. In other places in the standard, “writing system” refers to the way a particular language is written. For example, the modern Japanese writing system uses four scripts: Han ideographs, Hiragana, Katakana and Latin (Romaji).

#Alphabets. A writing system that consists of letters for the writing of both consonants and vowels is called an alphabet. The term “alphabet” is derived from the first two letters of the Greek script: alpha, beta. Consonants and vowels have equal status as letters in such a system. The Latin alphabet is the most widespread and well-known example of an alphabet, having been adapted for use in writing thousands of languages.

The correspondence between letters and sounds may be either more or less exact. Many alphabets do not exhibit a one-to-one correspondence between distinct sounds and letters or groups of letters used to represent them; often this is an indication of original spellings that were not changed as the language changed. Not only are many sounds represented by letter combinations, such as “th” in English, but the language may have evolved since the writing conventions were settled. Examples range from cases such as Italian or Finnish, where the match between letter and sound is rather close, to English, which has notoriously complex and arbitrary spelling.

Phonetic alphabets, in contrast, are used specifically for the precise transcription of the sounds of languages. The best known of these alphabets is the International Phonetic Alphabet, an adaptation and extension of the Latin alphabet by the addition of new letters and marks for specific sounds and modifications of sounds. Unlike normal alphabets, the intent of phonetic alphabets is that their letters exactly represent sounds. Phonetic alphabets are not used as general-purpose writing systems per se, but it is not uncommon for a formerly unwritten language to have an alphabet developed for it based on a phonetic alphabet.

#Abjads. A writing system in which only consonants are indicated is an abjad. The main letters are all consonants (or long vowels), with other vowels either left out entirely or optionally indicated with the use of secondary marks on the consonants. The Phoenician script is a prototypical abjad; a better-known example is the Arabic writing system. The term “abjad” is derived from the first four letters of the traditional order of the Arabic script: alef, beh, jeem, dal. Abjads are often, although not exclusively, associated with Semitic languages, which have word structures particularly well suited to the use of consonantal writing. Some abjads allow consonant letters to mark long vowels, as the use of waw and yeh in Arabic for /u:/ or /i:/.

Hebrew and Arabic are typically written without any vowel marking at all. The vowels, when they do occur in writing, are referred to as points or harakat, and are indicated by the use of diacritic dots and other marks placed above and below the consonantal letters.

#Syllabaries. In a syllabary, each symbol of the system typically represents both a consonant and a vowel, or in some instances more than one consonant and a vowel. One of the best-known examples of a syllabary is Hiragana, used for Japanese, in which the units of the system represent the syllables ka, ki, ku, ke, ko, sa, si, su, se, so, and so on. In general parlance, the elements of a syllabary are not called letters, but rather syllables.This can lead to some confusion, however, because letters of alphabets and units of other writing systems are also used, singly or in combinations, to write syllables of languages. So in a broad sense, the term “letter” can be used to refer to the syllables of a syllabary.

In syllabaries such as Cherokee, Hiragana, Katakana, and Yi, each symbol has a unique shape, with no particular shape relation to any of the consonant(s) or vowels of the syllables. In other cases, however, the syllabic symbols of a syllabary are not atomic; they can be built up out of parts that have a consistent relationship to the phonological parts of the syllable. The best example of this is the Hangul writing system for Korean. Each Hangul syllable is made up of a part for the initial consonant (or consonant cluster), a part for the vowel (or diphthong), and an optional part for the final consonant (or consonant cluster). The relationship between the sounds and the graphic parts to represent them is systematic enough for Korean that the graphic parts collectively are known as jamos and constitute a kind of alphabet on their own.

The jamos of the Hangul writing system have another characteristic: their shapes are not completely arbitrary, but were devised with intentionally iconic shapes relating them to articulatory features of the sounds they represent in Korean. The Hangul writing system has thus also been classified as a featural syllabary.

#Abugidas. Abugidas represent a kind of blend of syllabic and alphabetic characteristics in a writing system. The Ethiopic script is an abugida. The term “abugida” is derived from the first four letters of the Ethiopic script in the Semitic order: alf, bet, gaml, dant. The order of vowels (-ä -u -i -a) is that of the traditional vowel order in the first four columns of the Ethiopic syllable chart. Historically, abugidas spread across South Asia and were adapted by many languages, often of phonologically very different types.

This process has also resulted in many extensions, innovations, and/or simplifications of the original patterns. The best-known example of an abugida is the Devanagari script, used in modern times to write Hindi and many other Indian languages, and used classically to write Sanskrit. See Section 12.1, Devanagari, for a detailed description of how Devanagari works and is rendered.

In an abugida, each consonant letter carries an inherent vowel, usually /a/. There are also vowel letters, often distinguished between a set of independent vowel letters, which occur on their own, and dependent vowel letters, or matras, which are subordinate to consonant letters. When a dependent vowel letter follows a consonant letter, the vowel overrides the inherent vowel of the consonant. This is shown schematically in Figure 6-1.

#Figure 6-1. Overriding Inherent Vowels
ka + i → ki ka + u → ku
ka + e → ke ka + o → ko

Abugidas also typically contain a special element usually referred to as a halant, virama, or killer, which, when applied to a consonant letter with its inherent vowel, has the effect of removing the inherent vowel, resulting in a bare consonant sound.

In Brahmi-derived scripts, text often needs to be interpreted as a sequence of orthographic syllables, each of which is a two-dimensional visual arrangement of components that form a unit. At the core of an orthographic syllable is a base character, which can be a consonant, an independent vowel, a numeric character, or a ligature formed from base characters and other characters. Attached to this core may be dependent forms (such as half-forms, subjoined forms, repha forms, medial forms) of consonants or independent vowels, as well as nukta marks, virama marks, dependent vowel marks, register shifter marks, tone marks, final consonant marks, and other marks. It is common for different components of orthographic syllables to form ligatures. Orthographic syllables often do not correspond to phonological syllables; it is common for the final consonants of phonological syllables to become the base characters, or sometimes dependent forms, of subsequent orthographic syllables.

Because of legacy practice, three distinct approaches have been taken in the Unicode Standard for the encoding of abugidas: the Devanagari model, the Tibetan model, and the Thai model. The Devanagari model, used for most abugidas, represents text in primarily phonetic order and encodes a virama character that can combine with adjacent consonants to create conjunct forms. The Tibetan model also uses the primarily phonetic order, but its subjoined consonants are encoded directly rather than as virama-consonant sequences. The Thai model represents text in primarily visual display order, based on the typewriter legacy; neither Thai nor the other scripts using this model have conjunct forms.

The Ethiopic script is traditionally analyzed as an abugida, because the base character for each consonantal series is understood as having an inherent vowel. However, Ethiopic lacks some of the typical features of Brahmi-derived scripts, such as halants and matras. Historically, it was derived from early Semitic scripts and in its earliest form was an abjad. In its traditional presentation and its encoding in the Unicode Standard, it is now treated more like a syllabary.

#Logosyllabaries. The final major category of writing system is known as the logosyllabary. In a logosyllabary, the units of the writing system are used primarily to write words and/or morphemes of words, with some subsidiary usage to represent syllabic sounds per se.

The best example of a logosyllabary is the Han script, used for writing Chinese and borrowed by a number of other East Asian languages for use as part of their writing systems. The term for a unit of the Han script is hànzì 漢字 in Chinese, kanji 漢字 in Japanese, and hanja 漢字 in Korean. In many instances this unit also constitutes a word, but more typically, two or more units together are used to write a word.

The basic unit of a logosyllabary has variously been referred to as an ideograph (also ideogram), a logograph (also logogram), or a sinogram. Other terms exist as well, and especially for poorly understood or undeciphered writing systems, the units of writing may simply be called signs. Notionally, a logograph (or logogram) is a unit of writing which represents a word or morpheme, whereas an ideograph (or ideogram) is a unit of writing which represents an idea or concept. However, the lines between these terms are often unclear, and usage varies widely. The Unicode Standard makes no principled distinction between these terms, but rather follows the customary usage associated with a given script or writing system. For the Han script, the term CJK ideograph (or Han ideograph) is used.

There are a number of other historical examples of logosyllabaries, such as Tangut. They vary in the degree to which they combine logographic writing principles, where the symbols stand for morphemes or entire words, and syllabic writing principles, where the symbols come to represent syllables per se, divorced from their meaning as morphemes or words. In some notable instances, as for Sumero-Akkadian cuneiform, a logosyllabary may evolve through time into a syllabary or alphabet by shedding its use of logographs. In other instances, as for the Han script, the use of logographic characters is very well entrenched and persistent. However, even for the Han script a small number of characters are used purely to represent syllabic sounds, so as to be able to represent such things as foreign personal names and place names.

Egyptian hieroglyphs constitute another mixed example. The majority of the hieroglyphs are logographs, but Egyptian hieroglyphs also contain a well-defined subset that functions as an alphabet, in addition to other signs that represent sequences of consonants. And some hieroglyphs serve as semantic determinatives, rather than logographs in their own right—a function which bears some comparison to the way radicals work in CJK ideographs. To simplify the overall typology of Unicode scripts, Egyptian hieroglyphs and other hieroglyphic systems are lumped together with true logosyllabaries such as Han, but there are many differences in detail. For more about Egyptian hieroglyphs, in particular, see Section 11.4, Egyptian Hieroglyphs.

The classification of a writing system is often rendered somewhat ambiguous by complications in the exact ways in which it matches up written elements to the phonemes or syllables of a language. For example, although Hiragana is classified as a syllabary, it does not always have an exact match between syllables and written elements. Syllables with long vowels are not written with a single element, but rather with a sequence of elements. Thus the syllable with a long vowel kū is written with two separate Hiragana symbols, {ku}+{u}.

There may also be complications when a writing system deviates from the historical model from which it derives. For example, Mahajani and Multani are both based on the Brahmi model, but are structurally simpler than an abugida. These writing systems do not contain a virama. They also do not have matras and consonant conjunct formation characteristic to abugidas. Instead, Mahajani and Multani behave respectively as an alphabet and an abjad, and are encoded and classified accordingly in the Unicode Standard.

Because of these kinds of complications, one must always be careful not to assume too much about the structure of a writing system from its nominal classification.

#Typology of Scripts in the Unicode Standard. Table 6-1 lists all of the scripts currently encoded in the Unicode Standard, showing the writing system type for each. The list is an approximate guide, rather than a definitive classification, because of the mix of features seen in many scripts. The writing systems for some languages may be quite complex, mixing more than one type of script together in a composite system. Japanese is the best example; it mixes a logosyllabary (Han), two syllabaries (Hiragana and Katakana), and one alphabet (Latin, for romaji). In some instances, there is not a one-to-one correspondence between a script and its type, because the type may depend on how the script is used. For example, the Sunuwar script is used as an alphabet by the community in Nepal, but is used as an abugida in Sikkim.

#Table 6-1. Typology of Scripts in the Unicode Standard
Alphabets Adlam, Armenian, Avestan, Bassa Vah, Carian, Caucasian Albanian, Coptic, Cyrillic, Deseret, Elbasan, Garay, Georgian, Glagolitic, Gothic, Greek, Hanifi Rohingya, Kayah Li, Latin, Lisu, Lycian, Lydian, Mahajani, Mandaic, Medefaidrin, Meroitic Cursive, Meroitic Hieroglyphs, Mongolian, Mro, Nag Mundari, N’Ko, Nyiakeng Puachue Hmong, Ogham, Ol Chiki, Ol Onal, Old Hungarian, Old Italic, Old Permic, Old Persian, Old Turkic, Osage, Osmanya, Pahawh Hmong, Pau Cin Hau, Runic, Shavian, Sunuwar, Tangsa, Thaana, Tifinagh, Todhri, Toto, Ugaritic, Vithkuqi, Wancho, Warang Citi, Yezidi
Abjads Arabic, Chorasmian, Elymaic, Hatran, Hebrew, Imperial Aramaic, Inscriptional Pahlavi, Inscriptional Parthian, Manichaean, Multani, Nabataean, Old North Arabian, Old Sogdian, Old South Arabian, Old Uyghur, Palmyrene, Phoenician, Psalter Pahlavi, Samaritan, Sogdian, Syriac
Abugidas Ahom, Balinese, Batak, Bengali, Bhaiksuki, Brahmi, Buginese, Buhid, Chakma, Cham, Devanagari, Dives Akuru, Dogra, Grantha, Gujarati, Gunjala Gondi, Gurmukhi, Gurung Khema, Hanunóo, Javanese, Kaithi, Kannada, Kawi, Kharoshthi, Khmer, Khojki, Khudawadi, Kirat Rai, Lao, Lepcha, Limbu, Makasar, Malayalam, Marchen, Masaram Gondi, Meetei Mayek, Modi, Myanmar, Nandinagari, New Tai Lue, Newa, Oriya, Phags-pa, Rejang, Saurashtra, Sharada, Siddham, Sinhala, Sora Sompeng, Soyombo, Sundanese, Syloti Nagri, Tagalog, Tagbanwa, Tai Le, Tai Tham, Tai Viet, Takri, Tamil, Telugu, Thai, Tibetan, Tirhuta, Tulu-Tigalari, Zanabazar Square
Logosyllabaries Anatolian Hieroglyphs, Egyptian Hieroglyphs, Han, Khitan Small Script, Linear A, Nüshu, Sumero-Akkadian, Tangut
Simple Syllabaries Bamum, Bopomofo, Canadian Aboriginal Syllabics, Cherokee, Cypriot, Cypro-Minoan, Ethiopic, Hiragana, Katakana, Linear B, Mende Kikakui, Miao, Vai, Yi
Featural Syllabaries Hangul

#Notational Systems. In addition to scripts for written natural languages, there are notational systems for other kinds of information. Some of these more closely resemble text than others. The Unicode Standard encodes symbols for use with mathematical notation, Western and Byzantine musical notation, Duployan shorthand, Sutton SignWriting notation for sign languages, and Braille, as well as symbols for use in divination, such as the Yijing hexagrams. Notational systems can be classified by how closely they resemble text. Even notational systems that do not fully resemble text may have symbols used in text. In the case of musical notation, for example, while the full notation is two-dimensional, many of the encoded symbols are frequently referenced in texts about music and musical notation.

#6.2 General Punctuation

Punctuation characters—for example, U+002C COMMA and U+2022 BULLET—are encoded only once, rather than being encoded again and again for particular scripts; such general-purpose punctuation may be used for any script or mixture of scripts. In contrast, punctuation principally used with a specific script is found in the block corresponding to that script, such as U+058A ARMENIAN HYPHEN, U+061B “؛ARABIC SEMICOLON, or the punctuation used with CJK ideographs in the CJK Symbols and Punctuation block. Script-specific punctuation characters may be unique in function, have different directionality, or be distinct in appearance or usage from their generic counterparts.

Punctuation intended for use with several related scripts is often encoded with the principal script for the group. For example, U+1735 PHILIPPINE SINGLE PUNCTUATION is encoded in a single location in the Hanunóo block, but it is intended for use with all four of the Philippine scripts.

#Use and Interpretation. The use and interpretation of punctuation characters can be heavily context dependent. For example, U+002E FULL STOP can be used as sentence-ending punctuation, an abbreviation indicator, a decimal point, and so on.

Many Unicode algorithms, such as the Bidirectional Algorithm and Line Breaking Algorithm, both of which treat numeric punctuation differently from text punctuation, resolve the status of any ambiguous punctuation mark depending on whether it is part of a number context.

Legacy character encoding standards commonly include generic characters for punctuation instead of the more precisely specified characters used in printing. Examples include the single and double quotes, period, dash, and space. The Unicode Standard includes these generic characters, but also encodes the unambiguous characters independently: various forms of quotation marks, em dash, en dash, minus, hyphen, em space, en space, hair space, zero width space, and so on.

#Rendering. Punctuation characters vary in appearance with the font style, just like the surrounding text characters. In some cases, where used in the context of a particular script, a specific glyph style is preferred. For example, U+002E FULL STOP should appear square when used with Armenian, but is typically circular when used with Latin. For mixed Latin/Armenian text, two fonts (or one font allowing for context-dependent glyph variation) may need to be used to render the character faithfully.

#Writing Direction. Punctuation characters shared across scripts have no inherent directionality. In a bidirectional context, their display direction is resolved according to the rules in Unicode Standard Annex #9, “Unicode Bidirectional Algorithm.” Certain script-specific punctuation marks have an inherent directionality that matches the writing direction of the script. For an example, see “Dandas” later in this section. The image of certain paired punctuation marks, specifically those that are brackets, is mirrored when the character is part of a right-to-left directional run (see Section 4.7, Bidi Mirrored). Mirroring ensures that the opening and closing semantics of the character remains independent of the writing direction. The same is generally not true for other punctuation marks even when their image is not bilaterally symmetric, such as slash or the curly quotes. See also “Paired Punctuation” later in this section.

In vertical writing, many punctuation characters have special vertical glyphs. Normally, fonts contain both the horizontal and vertical glyphs, and the selection of the appropriate glyph is based on the text orientation in effect at rendering time. However, see “CJK Compatibility Forms: Vertical Forms” later in this section.

Figure 6-2 shows a set of three common shapes used for ideographic comma and ideographic full stop. The first shape in each row is that used for horizontal text, the last shape is that for vertical text. The centered form may be used with both horizontal and vertical text. See also Figure 6-4 for an example of vertical and horizontal forms for quotation marks.

#Figure 6-2. Forms of CJK Punctuation
Horizontal Centered Vertical

#Layout Controls. A number of characters in the blocks described in this section are not graphic punctuation characters, but rather affect the operation of layout algorithms. For a description of those characters, see Section 23.2, Layout Controls.

#Encoding Characters with Multiple Semantic Values. Some of the punctuation characters in the ASCII graphic character range (U+0020..U+007F) have multiple uses, either through ambiguity in the original standards or through accumulated reinterpretations of a limited code set. For example, 2716 is defined in ANSI X3.4 as apostrophe (closing single quotation mark; acute accent), and 2D16 is defined as hyphen-minus. In general, the Unicode Standard provides the same interpretation for the equivalent code points, without adding to or subtracting from their semantics. The Unicode Standard supplies unambiguous codes elsewhere for the most useful particular interpretations of these ASCII values; the corresponding unambiguous characters are cross-referenced in the character names list for this block. For more information, see “Apostrophes,” “Space Characters,” and “Dashes and Hyphens” later in this section.

#6.2.1 Blocks Devoted to Punctuation

For compatibility with widely used legacy character sets, the Basic Latin (ASCII) block (U+0000..U+007F) and the Latin-1 Supplement block (U+0080..U+00FF) contain several of the most common punctuation signs. They are isolated from the larger body of Unicode punctuation, signs, and symbols only because their relative code locations within ASCII and Latin-1 are so widely used in standards and software. The Unicode Standard has a number of blocks devoted specifically to encoding collections of punctuation characters.

The General Punctuation block (U+2000..U+206F) contains the most common punctuation characters widely used in Latin typography, as well as a few specialized punctuation marks and a large number of format control characters. All of these punctuation characters are intended for generic use, and in principle they could be used with any script.

The Supplemental Punctuation block (U+2E00..U+2E7F) is devoted to less commonly encountered punctuation marks, including those used in specialized notational systems or occurring primarily in ancient manuscript traditions.

The CJK Symbols and Punctuation block (U+3000..U+303F) has the most commonly occurring punctuation specific to East Asian typography—that is, typography involving the rendering of text with CJK ideographs.

The Vertical Forms block (U+FE10..U+FE1F), the CJK Compatibility Forms block (U+FE30..U+FE4F), the Small Form Variants block (U+FE50..U+FE6F), and the Halfwidth and Fullwidth Forms block (U+FF00..U+FFEF) contain many compatibility characters for punctuation marks, encoded for compatibility with a number of East Asian character encoding standards. Their primary use is for round-trip mapping with those legacy standards. For vertical text, the regular punctuation characters are used instead, with alternate glyphs for vertical layout supplied by the font.

The punctuation characters in these various blocks are discussed below in terms of their general types.

#6.2.2 Format Control Characters

Format control characters are special characters that have no visible glyph of their own, but that affect the display of characters to which they are adjacent, or that have other specialized functions such as serving as invisible anchor points in text. All format control characters have General_Category = Cf. A significant number of format control characters are encoded in the General Punctuation block, but their descriptions are found in other sections.

Cursive joining controls, as well as U+200B ZERO WIDTH SPACE, U+2028 LINE SEPARATOR, U+2029 PARAGRAPH SEPARATOR, and U+2060 WORD JOINER, are described in Section 23.2, Layout Controls. Bidirectional ordering controls are also discussed in Section 23.2, Layout Controls, but their detailed use is specified in Unicode Standard Annex #9, “Unicode Bidirectional Algorithm.”

Invisible operators are explained in Section 22.6, Invisible Mathematical Operators. Deprecated format characters related to obsolete models of Arabic text processing are described in Section 23.3, Deprecated Format Characters.

The reserved code points U+2065 and U+FFF0..U+FFF8, as well as any reserved code points in the range U+E0000..U+E0FFF, are reserved for the possible future encoding of other format control characters. Because of this, they are treated as default ignorable code points. For more information, see Section 5.21, Ignoring Characters in Processing.

#6.2.3 Space Characters

Space characters are found in several blocks in the Unicode Standard. The list of space characters appears in Table 6-2.

#Table 6-2. Unicode Space Characters
Code Name
U+0020 SPACE
U+00A0 NO-BREAK SPACE
U+1680 OGHAM SPACE MARK
U+2000 EN QUAD
U+2001 EM QUAD
U+2002 EN SPACE
U+2003 EM SPACE
U+2004 THREE-PER-EM SPACE
U+2005 FOUR-PER-EM SPACE
U+2006 SIX-PER-EM SPACE
U+2007 FIGURE SPACE
U+2008 PUNCTUATION SPACE
U+2009 THIN SPACE
U+200A HAIR SPACE
U+202F NARROW NO-BREAK SPACE
U+205F MEDIUM MATHEMATICAL SPACE
U+3000 IDEOGRAPHIC SPACE

The space characters in the Unicode Standard can be identified by their General Category, (gc = Zs), in the Unicode Character Database. One exceptional “space” character is U+200B ZERO WIDTH SPACE. This character, although called a “space” in its name, does not actually have any width or visible glyph in display. It functions primarily to indicate word boundaries in writing systems that do not actually use orthographic spaces to separate words in text. It is given the General Category (gc = Cf) and is treated as a format control character, rather than as a space character, in implementations. Further discussion of U+200B ZERO WIDTH SPACE, as well as other zero-width characters with special properties, can be found in Section 23.2, Layout Controls.

The most commonly used space character is U+0020 SPACE. In ideographic text, U+3000 IDEOGRAPHIC SPACE is commonly used because its width matches that of the ideographs.

The main difference among other space characters is their width. U+2000..U+2006 are standard quad widths used in typography. U+2007 FIGURE SPACE has a fixed width, known as tabular width, which is the same width as digits used in tables. U+2008 PUNCTUATION SPACE is a space defined to be the same width as a period. U+2009 THIN SPACE and U+200A HAIR SPACE are successively smaller-width spaces used for narrow word gaps and for justification of type. The fixed-width space characters (U+2000..U+200A) are derived from conventional (hot lead) typography. Algorithmic kerning and justification in computerized typography do not use these characters. However, where they are used (for example, in typesetting mathematical formulae), their width is generally font-specified, and they typically do not expand during justification. The exception is U+2009 THIN SPACE, which sometimes gets adjusted.

In addition to the various fixed-width space characters, there are a few script-specific space characters in the Unicode Standard. U+1680 OGHAM SPACE MARK is unusual in that it is generally rendered with a visible horizontal line, rather than being blank.

#No-Break Space. U+00A0 NO-BREAK SPACE (NBSP) is the nonbreaking counterpart of U+0020 SPACE. It has the same width, but behaves differently for line breaking. For more information, see Unicode Standard Annex #14, “Unicode Line Breaking Algorithm.”

Unlike U+0020, U+00A0 NO-BREAK SPACE behaves as a numeric separator for the purposes of bidirectional layout. See Unicode Standard Annex #9, “Unicode Bidirectional Algorithm,” for a detailed discussion of the Unicode Bidirectional Algorithm.

U+00A0 NO-BREAK SPACE has an additional, important function in the Unicode Standard. It may serve as the base character for displaying a nonspacing combining mark in apparent isolation. Versions of the standard prior to Version 4.1 indicated that U+0020 SPACE could also be used for this function, but SPACE is no longer recommended, because of potential interactions with the handling of SPACE in XML and other markup languages. See Section 2.11, Combining Characters, for further discussion.

#Narrow No-Break Space. U+202F NARROW NO-BREAK SPACE (NNBSP) is a narrow version of U+00A0 NO-BREAK SPACE. The NNBSP can be used to represent the narrow space occurring around punctuation characters in French typography, which is called an “espace fine insécable.”

Prior to Unicode Version 16.0, NNBSP was recommended as a special format character in Mongolian text. That role has been taken over by U+180E MONGOLIAN VOWEL SEPARATOR. See Section 13.5, Mongolian, for more information.

#6.2.4 Dashes and Hyphens

Because of its prevalence in legacy encodings, U+002D HYPHEN-MINUS is the most common of the dash characters used to represent a hyphen. It has ambiguous semantic value and is rendered with an average width. U+2010 HYPHEN represents the hyphen as found in words such as “left-to-right.” It is rendered with a narrow width. When typesetting text, U+2010 HYPHEN is preferred over U+002D HYPHEN-MINUS. U+2011 NON-BREAKING HYPHEN has the same semantic value as U+2010 HYPHEN, but should not be broken across lines.

U+2012 FIGURE DASH has the same (ambiguous) semantic as the U+002D HYPHEN-MINUS, but has the same width as digits (if they are monospaced). U+2013 EN DASH is used to indicate a range of values, such as 1973–1984, although in some languages hyphen is used for that purpose. The en dash should be distinguished from the U+2212 MINUS SIGN, which is an arithmetic operator. Although it is not preferred in mathematical typesetting, typographers sometimes use U+2013 EN DASH to represent the minus sign, particularly a unary minus. When interpreting formulas, U+002D HYPHEN-MINUS, U+2012 FIGURE DASH, and U+2212 MINUS SIGN should each be taken as indicating a minus sign, as in “x = a - b”, unless a higher-level protocol precisely defines which of these characters serves that function.

U+2014 EM DASH is used to make a break—like this—in the flow of a sentence. (Some typographers prefer to use U+2013 EN DASH set off with spaces – like this – to make the same kind of break.) Like many other conventions for punctuation characters, such usage may depend on language. This kind of dash is commonly represented with a typewriter as a double hyphen. In older mathematical typography, U+2014 EM DASH may also used to indicate a binary minus sign. U+2015 HORIZONTAL BAR is used to introduce quoted text in some typographic styles.

U+2E3A TWO-EM DASH and U+2E3B THREE-EM DASH can be used to represent dashes even wider than an em dash. An extra-wide dash in contemporary Chinese typography, referred to as 破折號 (pòzhéhào), is used to indicate an abrupt change of thought, the insertion of new content, or the continuation of tone or sound. That wide dash is often represented with a sequence of two U+2014 EM DASH characters, but modern practice is transitioning to the use of U+2E3A TWO-EM DASH, instead, which behaves better typographically than the sequence of em dashes. Because U+2E3A and U+2E3B are so wide, the code charts use the dashed box convention for their representative glyphs; these are, however, just ordinary punctuation characters, and not format control characters.

Dashes and hyphen characters may also be found in other blocks in the Unicode Standard. The full list is provided in Table 6-3. That list is correlated with the Dash property in the Unicode Character Database. Characters with the Dash property consist of all characters with General_Category=Pd plus U+2053 SWUNG DASH, U+2212 MINUS SIGN, and characters with compatibility decompositions to U+2212 MINUS SIGN.

#Table 6-3. Unicode Dash Characters
Code Name
U+002D HYPHEN-MINUS
U+058A ARMENIAN HYPHEN
U+05BE HEBREW PUNCTUATION MAQAF
U+1400 CANADIAN SYLLABICS HYPHEN
U+1806 MONGOLIAN TODO SOFT HYPHEN
U+2010 HYPHEN
U+2011 NON-BREAKING HYPHEN
U+2012 FIGURE DASH
U+2013 EN DASH
U+2014 EM DASH
U+2015 HORIZONTAL BAR (= quotation dash)
U+2053 SWUNG DASH
U+207B SUPERSCRIPT MINUS
U+208B SUBSCRIPT MINUS
U+2212 MINUS SIGN
U+2E17 DOUBLE OBLIQUE HYPHEN
U+2E1A HYPHEN WITH DIAERESIS
U+2E3A TWO-EM DASH
U+2E3B THREE-EM DASH
U+2E40 DOUBLE HYPHEN
U+2E5D OBLIQUE HYPHEN
U+301C WAVE DASH
U+3030 WAVY DASH
U+30A0 KATAKANA-HIRAGANA DOUBLE HYPHEN
U+FE31 PRESENTATION FORM FOR VERTICAL EM DASH
U+FE32 PRESENTATION FORM FOR VERTICAL EN DASH
U+FE58 SMALL EM DASH
U+FE63 SMALL HYPHEN-MINUS
U+FF0D FULLWIDTH HYPHEN-MINUS
U+10EAD YEZIDI HYPHENATION MARK

For a description of the line breaking behavior of dashes and hyphens, see Unicode Standard Annex #14, “Unicode Line Breaking Algorithm.”

#Soft Hyphen. Despite its name, U+00AD SOFT HYPHEN is not a hyphen, but rather an invisible format character used to indicate optional intraword breaks. As described in Section 23.2, Layout Controls, its effect on the appearance of the text depends on the language and script used.

#Tilde. Although several shapes are commonly used to render U+007E “~TILDE, modern fonts generally render it with a center line glyph, as shown here and in the code charts. However, it may also appear as a raised, spacing tilde, serving as a spacing clone of U+0303 “ ̃COMBINING TILDE (see “Spacing Clones of Diacritical Marks” in Section 7.9, Combining Marks). This is a form common in older implementations, particularly for terminal emulation and typewriter-style fonts.

Some of the common uses of a tilde include indication of alternation, an approximate value, or, in some notational systems, indication of a logical negation. In the latter context, it is really being used as a shape-based substitute character for the more precise U+00AC “¬NOT SIGN. A tilde is also used in dictionaries to repeat the defined term in examples. In that usage, as well as when used as punctuation to indicate alternation, it is more appropriately represented by a wider form, encoded as U+2053 “SWUNG DASH. U+02DC “˜SMALL TILDE is a modifier letter encoded explicitly as the spacing form of the combining tilde as a diacritic. For mathematical usage, U+223C “TILDE OPERATOR should be used to unambiguously encode the operator.

#Dictionary Abbreviation Symbols. In addition to the widespread use of tilde in dictionaries, more specialized dictionaries may make use of symbols consisting of hyphens or tildes with dots or circles above or below them to abbreviate the representation of inflected or derived forms (plurals, case forms, and so on) in lexical entries. U+2E1A HYPHEN WITH DIAERESIS, for example, is typically used in German dictionaries as a short way of indicating that the addition of a plural suffix also causes placement of an umlaut on the main stem vowel. U+2E1B TILDE WITH RING ABOVE indicates a change in capitalization for a derived form, and so on. Such conventions are particularly widespread in German dictionaries, but may also appear in other dictionaries influenced by German lexicography.

#6.2.5 Paired Punctuation

#Mirroring of Paired Punctuation. Paired punctuation marks such as parentheses (U+0028, U+0029), square brackets (U+005B, U+005D), and braces (U+007B, U+007D) are interpreted semantically rather than graphically in the context of bidirectional or vertical texts; that is, the orientation of these characters toward the enclosed text is maintained by the software, independent of the writing direction. In a bidirectional context, the glyphs are adjusted as described in Unicode Standard Annex #9, “Unicode Bidirectional Algorithm.” (See also Section 4.7, Bidi Mirrored.) During display, the software must ensure that the rendered glyph is the correct one in the context of bidirectional or vertical texts.

Paired punctuation marks containing the qualifier “LEFT” in their name are taken to denote opening; characters whose name contains the qualifier “RIGHT” are taken to denote closing. For example, U+0028 LEFT PARENTHESIS and U+0029 RIGHT PARENTHESIS are interpreted as opening and closing parentheses, respectively. In a right-to-left directional run, U+0028 is rendered as “)”. In a left-to-right run, the same character is rendered as “(”. In some mathematical usage, brackets may not be paired, or may be deliberately used in the reversed sense, such as ]a,b[. Mirroring assures that in a right-to-left environment, such specialized mathematical text continues to read ]b,a[ and not [b, a]. See also “Language-Based Usage of Quotation Marks” later in this section.

#Quotation Marks and Brackets. Like brackets, quotation marks occur in pairs, with some overlap in usage and semantics between these two types of punctuation marks. For example, some of the CJK quotation marks resemble brackets in appearance, and they are often used when brackets would be used in non-CJK text. Similarly, both single and double guillemets may be treated more like brackets than quotation marks. Unlike brackets, quotation marks are not mirrored in a bidirectional context.

Some of the editing marks used in annotated editions of scholarly texts exhibit features of both quotation marks and brackets. The particular convention employed by the editors determines whether editing marks are used in pairs, which editing marks form a pair, and which is the opening character.

Horizontal brackets—for example, those used in annotating mathematical expressions—are not paired punctuation, even though the set includes both top and bottom brackets. See “Horizontal Brackets” in Section 22.7, Technical Symbols, for more information.

#6.2.6 Language-Based Usage of Quotation Marks

The use of quotation marks differs systematically by language and by medium. As for many other punctuation marks, and in contrast to parentheses or bracket characters, quotation marks in the Unicode Standard are encoded by shape and not by how they are used in relation to the quoted text. The same character may be used to open a quote in one language, to close a quote in another, or to serve both functions in a third.

The most commonly used character for quotation mark is U+0022 " QUOTATION MARK, usually represented with a straight double quote glyph. This quotation mark is supported on most keyboard layouts, but typographically it is mainly appropriate for typewritten manuscripts, programming text, or similar usage. Editing implementations commonly offer a facility for converting the U+0022 QUOTATION MARK to a typographically appropriate glyph for the language and context, such as distinguishing between opening or closing a quote. Within each of the different standard typographical shapes for quotation marks, there exist common glyph variations depending on font design, as discussed in the following text.

The same issues apply to the character supported on keyboards for the single quote, U+0027 APOSTROPHE; however, as its name indicates, that character is further conflated with its function as an apostrophe. (See the discussion in “Apostrophes” in this section.)

#European Usage. In European typography, it is common to use guillemets (single or double angle quotation marks) for books and, except for some languages, curly quotation marks in office automation. Single guillemets may be used for quotes inside quotes. The following description does not attempt to be complete, but intends to document a range of known usages of quotation mark characters. Some of these usages are also illustrated in Figure 6-3. In this section, the words single and double are omitted from character names where there is no conflict or both are meant.

Dutch, English, Italian, Portuguese, Spanish, and Turkish use a left quotation mark and a right quotation mark for opening and closing quotations, respectively. It is typical to alternate single and double quotes for quotes within quotes. Whether single or double quotes are used for the outer quotes depends on local and stylistic conventions.

Czech, German, and Slovak use the low-9 style of quotation mark for opening instead of the standard open quotes. They employ the left quotation mark style of quotation mark for closing instead of the more common right quotation mark forms. When guillemets are used in German books, they point to the quoted text. This style is the inverse of French usage.

Danish, Finnish, Norwegian, and Swedish use the same right quotation mark character for both the opening and closing quotation character. This usage is employed both for office automation purposes and for books. Swedish books sometimes use the guillemet, U+00BB RIGHT-POINTING DOUBLE ANGLE QUOTATION MARK, for both opening and closing.

Hungarian and Polish usage of quotation marks is similar to the Scandinavian usage, except that they use low double quotes for opening quotations. Presumably, these languages avoid the low single quote so as to prevent confusion with the comma.

French, Greek, Russian, and Slovenian, among others, use the guillemets, but Slovenian usage is the same as German usage in their direction. Of these languages, at least French inserts space between text and quotation marks. In the French case, U+00A0 NO-BREAK SPACE can be used for the space that is enclosed between quotation mark and text; this choice helps line breaking algorithms.

#Figure 6-3. European Quotation Marks

#Glyph Variation in Curly Quotes. The glyphs for the quotation marks in the range U+2018..U+201F may vary significantly across fonts. The two most typical styles use curly or wedge-shaped glyphs. See Table 6-4.

#Table 6-4. Models of Visual Relationship between Quote Glyphs

The Unicode code charts use a curly style in a serifed, Times-like font. Because quotation marks are used in pairs, glyphs within a single style are expected to be in a certain visual relationship, and that relationship stands regardless of glyph style. The visual relationship follows either a rotated or a mirrored model. The rotated model is predominant in both curly and wedge glyph style fonts. These two models are illustrated in Table 6-4 using sample fonts with different glyph styles. The glyphs are enlarged for clarity.

In the rotated model, turning the ink of the glyph for U+201D RIGHT DOUBLE QUOTATION MARK 180 degrees results in the glyph for U+201C LEFT DOUBLE QUOTATION MARK; flipping it horizontally results in the glyph for U+201F DOUBLE HIGH-REVERSED-9 QUOTATION MARK. The same symmetries apply to the raised single quotation marks. Similarly, the glyphs for the low double quotation marks, U+201E DOUBLE LOW-9 QUOTATION MARK and U+2E42 DOUBLE LOW-REVERSED-9 QUOTATION MARK, are horizontally flipped images of each other.

Some fonts in widespread use instead follow the mirrored model, in which the glyph for U+201C looks like a mirrored image of the glyph for U+201D instead of a rotated image of it. Most fonts that follow the mirrored model use wedge style glyphs for quotation marks. In particular, in fonts such as Tahoma and Verdana, the glyph for U+201F is a rotated image of the glyph for U+201D, which makes the glyphs for U+201C and U+201F appear swapped compared to the typical design of wedge style quote glyphs using the rotated model. The sets of glyphs which show these swapped appearances are highlighted by a light background in Table 6-4.

#East Asian Usage. The glyph for each quotation mark character for an Asian character set occupies predominantly a single quadrant of the character cell. The quadrant used depends on whether the character is opening or closing and whether the glyph is for use with horizontal or vertical text.

The pairs of quotation characters are listed in Table 6-5.

#Table 6-5. East Asian Quotation Marks
Style Opening Closing
Corner bracket 300C 300D
White corner bracket 300E 300F
Double prime 301D 301F

#Glyph Variation in East Asian Usage. In East Asian usage, the glyphs for “double-prime” quotation marks U+301D REVERSED DOUBLE PRIME QUOTATION MARK and U+301F LOW DOUBLE PRIME QUOTATION MARK consist of a pair of wedges, slanted either forward or backward, with the tips of the wedges pointing either up or down. In a pair of double-prime quotes, the closing and the opening character of the pair slant in opposite directions. Two common variations exist, as shown in Figure 6-4. To confuse matters more, another form of double-prime quotation marks is used with Western-style horizontal text, in addition to the curly single or double quotes.

#Figure 6-4. Asian Quotation Marks

Three pairs of quotation marks are used with Western-style horizontal text, as shown in Table 6-6.

#Table 6-6. Opening and Closing Forms
Style Opening Closing Comment
Single 2018 2019 Rendered as “wide” character
Double 201C 201D Rendered as “wide” character
Double prime 301D 301E

#Overloaded Character Codes. The character codes for standard quotes can refer to regular narrow quotes from a Latin font used with Latin text as well as to wide quotes from an Asian font used with other wide characters. This situation can be handled with some success where the text is marked up with language tags. For more information on narrow and wide characters, see Unicode Standard Annex #11, “East Asian Width.”

#Consequences for Semantics. The semantics of U+00AB LEFT-POINTING DOUBLE ANGLE QUOTATION MARK, U+00BB RIGHT-POINTING DOUBLE ANGLE QUOTATION MARK, and U+201D RIGHT DOUBLE QUOTATION MARK are context dependent. By contrast, the semantics of U+201A SINGLE LOW-9 QUOTATION MARK and U+201B SINGLE HIGH-REVERSED-9 QUOTATION MARK are always opening. That usage is distinct from that of U+301F LOW DOUBLE PRIME QUOTATION MARK, which is unambiguously closing. All other quotation marks may represent opening or closing quotation marks depending on the usage.

#6.2.7 Apostrophes

U+0027 APOSTROPHE is the most commonly used character for apostrophe. For historical reasons, U+0027 is a particularly overloaded character. In ASCII, it is used to represent a punctuation mark (such as right single quotation mark, left single quotation mark, apostrophe punctuation, vertical line, or prime) or a modifier letter (such as apostrophe modifier or acute accent). Punctuation marks generally break words; modifier letters generally are considered part of a word.

When text is set, U+2019 RIGHT SINGLE QUOTATION MARK is preferred as apostrophe, but only U+0027 is present on most keyboards. Software commonly offers a facility for automatically converting the U+0027 APOSTROPHE to a contextually selected curly quotation glyph. In these systems, a U+0027 in the data stream is always represented as a straight vertical line and can never represent a curly apostrophe or a right quotation mark.

#Letter Apostrophe. U+02BC MODIFIER LETTER APOSTROPHE is preferred where the apostrophe is to represent a modifier letter (for example, in transliterations to indicate a glottal stop). In the latter case, it is also referred to as a letter apostrophe.

#Punctuation Apostrophe. U+2019 RIGHT SINGLE QUOTATION MARK is preferred where the character is to represent a punctuation mark, as for contractions: “We’ve been here before.” In this latter case, U+2019 is also referred to as a punctuation apostrophe.

An implementation cannot assume that users’ text always adheres to the distinction between these characters. The text may come from different sources, including mapping from other character sets that do not make this distinction between the letter apostrophe and the punctuation apostrophe/right single quotation mark. In that case, all of them will generally be represented by U+2019.

The semantics of U+2019 are therefore context dependent. For example, if surrounded by letters or digits on both sides, it behaves as an in-text punctuation character and does not separate words or lines.

#6.2.8 Hyphenation Point and Dictionary Syllabification

This subsection discusses several punctuation characters typically used in dictionaries to indicate syllabification and hyphenation, with a number of illustrative examples primarily from English-language dictionaries.

U+2027 HYPHENATION POINT is a raised dot used to indicate correct word breaking, as in “dic‧tio‧nary.” It is a punctuation mark, to be distinguished from U+00B7 MIDDLE DOT, which has multiple semantics.

Table 6-7 gives examples of syllabification conventions in a number of English dictionaries from the 19th century to modern usage. These illustrate the usage of U+2027 HYPHENATION POINT in modern conventions, but also show the use of dashes and the interaction of syllabification punctuation marks with various modifier letters indicating syllabic stress.

#Table 6-7. Dictionary Syllabification Conventions
Example Source Notes
SYʹLLABLE Dictionary of the English Language (Samuel Johnson, 1843) This early dictionary uses U+02B9 MODIFIER LETTER PRIME to indicate syllabic stress, placing it after the vowel letter of a syllable, rather than at the end of a syllable.
si·lă'bl Oxford English Dictionary (1st Edition) The middle dot indicates the vowel of the stressed syllable, similar to Johnson's use of a prime. The break between unstressed syllables is indicated with an apostrophe.
ˈsɪləb(ə)l Oxford English Dictionary (2nd Edition) The 2nd Edition has shifted to IPA, and indicates syllabic stress with a pre-posed U+02C8 MODIFIER LETTER VERTICAL LINE. The parentheses indicate optional omission of the schwa in pronunciation.
silʹə‐bl Chambers English Dictionary (7th Edition) This dictionary shows syllabic breaks, where the stressed syllable is followed by U+02B9 MODIFIER LETTER PRIME and the dash is U+2010 HYPHEN, to indicate a preferred hyphenation location. When splitting a word like abateʹ‐ment, the stress mark goes after the stressed syllable, followed directly by the hyphen.
sɪ̲ləbl BBC English Dictionary This dictionary uses IPA without indication of syllabic breaks in the pronunciation. The vowel of a stressed syllable is underscored, hence ɪ̲ can be represented as <U+026A, U+0332>.
sɪ̲ləbə⁰l Collins Cobuild English Language Dictionary This convention is similar to the BBC English Dictionary. The raised zero is represented by U+2070 SUPERSCRIPT ZERO and indicates the optional omission of the schwa.
syl‧la‧ble (sílləb'l) Readers Digest Great Illustrated Dictionary This exemplifies a modern convention of separating representation of syllabification and of pronunciation. The syllabification for word break is indicated by the use of U+2027 in the spelling. The primary stress is indicated with an accent on the vowel, instead of a modifier letter, and the apostrophe indicates the omitted schwa.
syl‧la‧ble
/ˈsiləbəl/
Webster’s 3rd New International Dictionary This edition of Webster's uses broad IPA for the pronunciation, shown between phonemic slashes. It splits words at the end of a line with a normal hyphen. When a word is split at the end of a line at the position of a dash in the spelling, that is indicated in the dictionary by the use of U+2E17 DOUBLE OBLIQUE HYPHEN.
sy‧lla‧ble
/ˈsɪl.ə.bəl/
[ˈsɪl.ə.bɫ̩]
Wiktionary Online dictionaries construct their entries quite differently than traditional printed dictionaries, and typically separate hyphenation from pronunciation. Wiktionary often omits hyphenation, but when listed it uses the hyphenation point to show hyphenation opportunities. It may separately list a broad IPA pronunciation and a close IPA pronunciation of the word. In the IPA pronunciation, a U+002E FULL STOP indicates phonological syllable boundaries.

Some dictionaries use a character that looks like a vertical series of four dots to indicate places where there is a syllable, but no allowable break, as for example, a⁞⁠plomb or hoar⁞⁠y. This convention can be represented by U+205E VERTICAL FOUR DOTS. To hint that this punctuation mark should not itself be an opportunity for a line break, it can be followed by U+2060 WORD JOINER.

See also the discussion of dictionary abbreviation symbols involving hyphens and tildes in Section 6.2.4, Dashes and Hyphens.

#Interaction with Line Breaking. Where possible, the line breaking properties for punctuation marks commonly used in dictionaries have been assigned so as to accommodate these and similar conventions by default. However, implementing the full conventions in dictionaries requires tailoring of line break classes and rules or other types of special support. See Unicode Standard Annex #14, Unicode Line Breaking Algorithm, for the specification of line break classes and how to tailor them.

#6.2.9 Other Punctuation

#Tironian Et. U+204A TIRONIAN SIGN ET acts as a punctuation mark meaning “and”. It can also function as a letter in some contexts. In some Medieval language materials, Tironian et can appear in uppercase and is represented by U+2E52 TIRONIAN SIGN CAPITAL ET. However, TIRONIAN SIGN ET and TIRONIAN SIGN CAPITAL ET are not case-mapped.

#Word Separator Middle Dot. Historic texts in many scripts, especially those that are handwritten (manuscripts), sometimes use a raised dot to separate words. Such word-separating punctuation is comparable in function to the use of space to separate words in modern typography.

U+2E31 WORD SEPARATOR MIDDLE DOT is a middle dot punctuation mark which is analogous in function to the script-specific character U+16EB RUNIC SINGLE PUNCTUATION, but is for use with any script that needs a raised dot for separating words. For example, it can be used for the word-separating dot seen in Avestan or Samaritan texts.

#Fraction Slash. U+2044 FRACTION SLASH is used between digits to form numeric fractions, such as 2/3 and 3/9. The standard form of a fraction built using the fraction slash is defined as follows: any sequence of one or more decimal digits (General Category = Nd), followed by the fraction slash, followed by any sequence of one or more decimal digits. Such a fraction should be displayed as a unit, such as ¾ or ¾. The precise choice of display can depend on additional formatting information.

If the displaying software is incapable of mapping the fraction to a unit, then it can also be displayed as a simple linear sequence as a fallback (for example, 3/4). If the fraction is to be separated from a previous number, then a space can be used, choosing the appropriate width (normal, thin, zero width, and so on). For example, 1 + THIN SPACE + 3 + FRACTION SLASH + 4 is displayed as 1¾.

#Spacing Overscores and Underscores. U+203E OVERLINE is the above-the-line counterpart to U+005F LOW LINE. It is a spacing character, not to be confused with U+0305 COMBINING OVERLINE. As with all overscores and underscores, a sequence of these characters should connect in an unbroken line. The overscoring characters also must be distinguished from U+0304 COMBINING MACRON, which does not connect horizontally in this way.

#Doubled Punctuation. Several doubled punctuation characters that have compatibility decompositions into a sequence of two punctuation marks are also encoded as single characters: U+203C DOUBLE EXCLAMATION MARK, U+2048 QUESTION EXCLAMATION MARK, and U+2049 EXCLAMATION QUESTION MARK. These doubled punctuation marks are included as an implementation convenience for East Asian and Mongolian text, when rendered vertically.

#Period or Full Stop. The period, or U+002E FULL STOP, can be circular or square in appearance, depending on the font or script. The hollow circle period used in East Asian texts is separately encoded as U+3002 IDEOGRAPHIC FULL STOP. Likewise, Armenian, Arabic, Ethiopic, and several other script-specific periods are coded separately because of their significantly different appearance.

In contrast, the various functions of the period, such as its use as sentence-ending punctuation, an abbreviation mark, or a decimal point, are not separately encoded. The specific semantic therefore depends on context.

In old-style numerals, where numbers vary in placement above and below the baseline, a decimal or thousands separator may be displayed with a dot that is raised above the baseline. Because it would be inadvisable to have a stylistic variation between old-style and new-style numerals that actually changes the underlying representation of text, the Unicode Standard considers this raised dot to be merely a glyphic variant of U+002E “.FULL STOP.

#Ellipsis. The omission of text is often indicated by a sequence of three dots “...”, a punctuation convention called ellipsis. Typographic traditions vary in how they lay out these dots. In some cases the dots are closely spaced; in other cases the dots are spaced farther apart. U+2026 “HORIZONTAL ELLIPSIS is the ordinary Unicode character intended for the representation of an ellipsis in text and typically shows the dots separated with a moderate degree of spacing. A sequence of three U+002E FULL STOP characters can also be used to indicate an ellipsis, in which case the space between the dots will depend on the font used for rendering.

In a monowidth font, a sequence of three full stops will be wider than horizontal ellipsis, and may be appropriate when following style guides that require more widely spaced dots. In this case, the spacing between the last dot and following punctuation would be as expected.

In contrast, for a typical proportional font, a full stop is very narrow and a sequence of three of them will be more tightly spaced than the dots in horizontal ellipsis. When adhering to style guides, such as the Chicago Manual of Style (CMOS), which call for more widely spaced dots in presentation, established practice calls for separating the dots (and any surrounding punctuation) by either a NBSP or NNBSP. These contrasts are illustrated in Table 6-8, using “dots” to refer to full stop characters, with or without intervening no break spaces.

#Table 6-8. Horizontal Ellipsis
Ellipsis dots dots + NBSP dots + NNBSP
Fixed abc…def abc...def abc . . . def abc . . . def
abc.…Def abc....Def abc. . . . Def abc. . . . Def
Proportional abc…def abc...def abc . . . def abc . . . def
abc.…Def abc....Def abc. . . . Def abc. . . . Def

There are conventions that use four dots for an ellipsis in certain grammatical contexts, such as elided content following the end of a sentence. These conventions can represent the four dots either as a sequence of <full stop, horizontal ellipsis> or <horizontal ellipsis, full stop> or simply as a sequence of four full stop characters, depending on the requirements of those conventions. The usual CMOS convention of placing a full stop initially in these sequences is also illustrated in Table 6-8.

In East Asian typographic traditions, particularly in Japan, an ellipsis is raised to the center line of text. When an ellipsis is represented by U+2026 HORIZONTAL ELLIPSIS or by sequences of full stops, this effect requires specialized rendering support. In practice, it is relatively common for authors of East Asian text to substitute U+22EF MIDLINE HORIZONTAL ELLIPSIS for this. Because the midline ellipsis is a mathematical symbol, intended to represent column elision in matrix notation, it is typically used with layout on a mathematical center line. With appropriate font design to harmonize with East Asian typography, this midline ellipsis can produce the desired appearance without having to support contextual shifting of the baseline for U+2026 HORIZONTAL ELLIPSIS.

#Vertical Ellipsis. When text is laid out vertically, the ellipsis is normally oriented so that the dots run from top to bottom. Most commonly, an East Asian font will contain a vertically oriented glyph variant of U+2026 for use in vertical text layout. U+FE19 PRESENTATION FORM FOR VERTICAL HORIZONTAL ELLIPSIS is a compatibility character for use in mapping to the GB 18030 standard; it would not usually be used for an ellipsis except in systems that cannot handle the contextual choice of glyph variants for vertical rendering.

U+22EE VERTICAL ELLIPSIS and U+22EF MIDLINE HORIZONTAL ELLIPSIS are part of a set of special ellipsis characters used for row or column elision in matrix notation. Although their primary use is for a mathematical context, U+22EF MIDLINE HORIZONTAL ELLIPSIS has also become popular for the midline ellipsis in East Asian typography. When U+22EF is used this way, an East Asian font will typically contain a rotated glyph variant for use in vertical text layout. If an appropriate mechanism for glyph variant substitution (such as the “vert” GSUB feature in the Open Font Format) in vertically rendered text is not available, U+FE19 PRESENTATION FORM FOR VERTICAL HORIZONTAL ELLIPSIS is the preferred character substitution to represent a vertical ellipsis, instead of the mathematical U+22EE VERTICAL ELLIPSIS.

U+205D TRICOLON has a superficial resemblance to a vertical ellipsis, but is part of a set of dot delimiter punctuation marks for various manuscript traditions. As for the colon, the dots in the tricolon are always oriented vertically.

#Leader Dots. Leader dots are typically seen in contexts such as a table of contents or in indices, where they represent a kind of style line, guiding the eye from an entry in the table to its associated page number. Usually leader dots are generated automatically by page formatting software and do not require the use of encoded characters. However, there are occasional plain text contexts in which a string of leader dots is represented as a sequence of characters. U+2024 ONE DOT LEADER and U+2025 TWO DOT LEADER are intended for such usage. U+2026 HORIZONTAL ELLIPSIS can also serve as a three-dot version of leader dots. These leader dot characters can be used to control, to a certain extent, the spacing of leader dots based on font design, in contexts where a simple sequence of full stops will not suffice.

U+2024 ONE DOT LEADER also serves as a “semicolon” punctuation in Armenian, where it is distinguished from U+002E FULL STOP. See Section 7.6, Armenian.

#Other Basic Latin Punctuation Marks. The interword punctuation marks encoded in the Basic Latin block are used for a variety of other purposes. This can complicate the tasks of parsers trying to determine sentence boundaries. As noted later in this section, some can be used as numeric separators. Both period and U+003A “:COLON can be used to mark abbreviations as in “etc.” or as in the Swedish abbreviation “S:ta” for “Sankta”. U+0021 “!EXCLAMATION MARK is used as a mathematical operator (factorial). U+003F “?QUESTION MARK is often used as a substitution character when mapping Unicode characters to other character sets where they do not have a representation. This practice can lead to unexpected results when the converted data are file names from a file system that supports “?” as a wildcard character.

Several punctuation marks, such as colon, middle dot and solidus closely resemble mathematical operators, such as U+2236 RATIO, U+22C5 DOT OPERATOR and U+2215 DIVISION SLASH. The latter are the preferred characters, but the former are often substituted because they are more easily typed.

#Canonical Equivalence Issues for Greek Punctuation. Some commonly used Greek punctuation marks are encoded in the Greek and Coptic block, but are canonical equivalents to generic punctuation marks encoded in the C0 Controls and Basic Latin block, because they are indistinguishable in shape. Thus, U+037E “;GREEK QUESTION MARK is canonically equivalent to U+003B “;SEMICOLON, and U+0387 “·GREEK ANO TELEIA is canonically equivalent to U+00B7 “·MIDDLE DOT. In these cases, as for other canonical singletons, the preferred form is the character that the canonical singletons are mapped to, namely U+003B and U+00B7 respectively. Those are the characters that will appear in any normalized form of Unicode text, even when used in Greek text as Greek punctuation. Text segmentation algorithms need to be aware of this issue, as the kinds of text units delimited by a semicolon or a middle dot in Greek text will typically differ from those in Latin text.

The character properties for U+00B7 MIDDLE DOT are particularly problematical, in part because of identifier issues for that character. There is no guarantee that all of its properties align exactly with U+0387 GREEK ANO TELEIA, because the latter’s properties are based on the limited function of the middle dot in Greek as a delimiting punctuation mark.

#Bullets. U+2022 BULLET is the typical character for a bullet. Within the general punctuation, several alternative forms for bullets are separately encoded: U+2023 TRIANGULAR BULLET, U+204C BLACK LEFTWARDS BULLET, and so on. U+00B7 MIDDLE DOT also often functions as a small bullet. Bullets mark the head of specially formatted paragraphs, often occurring in lists, and may use arbitrary graphics or dingbat forms as well as more conventional bullet forms. U+261E WHITE RIGHT POINTING INDEX, for example, is often used to highlight a note in text, as a kind of gaudy bullet.

#Paragraph Marks. U+00A7 SECTION SIGN and U+00B6 PILCROW SIGN are often used as visible indications of sections or paragraphs of text, in editorial markup, to show format modes, and so on. Which character indicates sections and which character indicates paragraphs may vary by convention. U+204B REVERSED PILCROW SIGN is a fairly common alternate representation of the paragraph mark.

#Numeric Separators. Any of the characters U+002C COMMA, U+002E FULL STOP, and the Arabic characters U+060C, U+066B, or U+066C (and possibly others) can be used as numeric separator characters, depending on the locale and user customizations.

#Obelus. Originally a punctuation mark to denote questionable passages in manuscripts, U+00F7 ÷ DIVISION SIGN is now most commonly used as a symbol indicating division. However, modern use is not limited to that meaning. The character is sometimes used to indicate a range (similar to the en-dash) or as a form of minus sign. The former use is attested for Russian, Polish and Italian, and latter use is still widespread in Scandinavian countries in some contexts, but may occur elsewhere as well. (See also the following text on “Commercial Minus.”)

#Commercial Minus. U+2052 COMMERCIAL MINUS SIGN is used in commercial or tax-related forms or publications in several European countries, including Germany and Scandinavia. The string “./.” is used as a fallback representation for this character.

The symbol may also appear as a marginal note in letters, denoting enclosures. One variation replaces the top dot with a digit indicating the number of enclosures.

An additional usage of the sign appears in the Uralic Phonetic Alphabet (UPA), where it marks a structurally related borrowed element of different pronunciation. In Finland and a number of other European countries, the dingbats and are always used for “correct” and “incorrect,” respectively, in marking a student’s paper. This contrasts with American practice, for example, where and might be used for “correct” and “incorrect,” respectively, in the same context.

#At Sign. U+0040 COMMERCIAL AT has acquired a prominent modern use as part of the syntax for e-mail addresses. As a result, users in practically every language community suddenly needed to use and refer to this character. Consequently, many colorful names have been invented for this character. Some of these contain references to animals or even pastries. Table 6-9 gives a sample.

#Table 6-9. Names for the @
Language Name and Comments
Chinese = xiǎo lǎoshǔ (means “little mouse” in Mandarin Chinese)
= lǎoshǔ hào (means “mouse mark” in Mandarin Chinese)
Danish = snabel-a
Dutch = apenstaartje (common, humorous slang)
Finnish = ät, ät-merkki (Finnish standard)
= kissanhäntä, miukumauku (common, humorous slang)
French = arobase, arrobe, escargot, a crolle (common, humorous slang)
German = Klammeraffe
Hebrew = shtrudl (“Strudel”, modern Hebrew)
= krukhit (more formal Hebrew)
Hungarian = kukac (common, humorous slang)
Italian = chiocciola
Polish = atka, małpa, małpka (common, humorous slang)
Portuguese = arroba
Russian = sobachka (common, humorous slang)
Slovenian = afna (common, humorous slang)
Spanish = arroba
Swedish = snabel-a, kanelbulle (common, humorous slang)

#6.2.10 Archaic Punctuation and Editorial Marks

#Archaic Punctuation. Many archaic scripts use punctuation marks consisting of a set of multiple dots, such as U+2056 THREE DOT PUNCTUATION. The semantics of these marks can vary by script, and some of them are also used for special conventions, such as the use of U+205E VERTICAL FOUR DOTS in modern dictionaries. U+205B FOUR DOT MARK and U+205C DOTTED CROSS were used by scribes in the margin to highlight a piece of text. More of these multiple-dot archaic punctuation marks are encoded in the range U+2E2A..U+2E2D.

These kinds of punctuation marks occur in ancient scripts and are also common in medieval manuscripts. Their specific functions may be different in each script or manuscript tradition. However, encoding only a single set in the Unicode Standard simplifies the task of deciding which character to use for a given mark.

There are some exceptions to this general rule. Archaic scripts with script-specific punctuation include Runic, Aegean Numbers, and Cuneiform. In particular, the appearance of punctuation written in the Cuneiform style is sufficiently different that no unification was attempted.

#Double Oblique Hyphen. U+2E17 “DOUBLE OBLIQUE HYPHEN is used in ancient Near Eastern linguistics to indicate certain morphological boundaries while continuing to use the ordinary hyphen to indicate other boundaries. This symbol is also semantically distinct from U+003D “=EQUALS SIGN. Fraktur fonts use an oblique glyph of similar appearance for the hyphen, but that is merely a font variation of U+002D HYPHEN-MINUS or U+2010 HYPHEN, not the distinctly encoded DOUBLE OBLIQUE HYPHEN.

#Editorial Marks. In addition to common-use editorial marks such as U+2041 CARET INSERTION POINT encoded in the General Punctuation block, there are a number of editorial marks encoded in the Supplemental Punctuation block (U+2E00..U+2E7F). Editorial marks differ from ordinary punctuation marks, in that their primary purpose is to allow editors to mark up a scholarly publication of a text to show the location and contents of insertions, omissions, variant readings, and other such information about the text.

The half brackets encoded in the range U+2E22..U+2E25 are widely used as editorial marks in critical editions of ancient and medieval texts. They appear, for example, in editions of transliterated Cuneiform and ancient Egyptian texts. U+2E26 LEFT SIDEWAYS U BRACKET and U+2E27 RIGHT SIDEWAYS U BRACKET are a specialized bracket pair used in some traditions, and should be distinguished from mathematical set symbols of similar appearance. The double parentheses are employed by Latinists.

#New Testament Editorial Marks. The Greek text of the New Testament exists in a large number of manuscripts with many textual variants. The most widely used critical edition of the New Testament, the Nestle-Aland edition published by the United Bible Societies (UBS), introduced a set of editorial characters that are regularly used in a number of journals and other publications. As a result, these editorial marks have become the recognized method of annotating the New Testament.

U+2E00 RIGHT ANGLE SUBSTITUTION MARKER is placed at the start of a single word when that word is replaced by one or more different words in some manuscripts. These alternative readings are given in the apparatus criticus. If there is a second alternative reading in one verse, U+2E01 RIGHT ANGLE DOTTED SUBSTITUTION MARKER is used instead.

U+2E02 LEFT SUBSTITUTION BRACKET is placed at the start of a sequence of words where an alternative reading is given in the apparatus criticus. This bracket is used together with the U+2E03 RIGHT SUBSTITUTION BRACKET. If there is a second alternative reading in one verse, the dotted forms at U+2E04 and U+2E05 are used instead.

U+2E06 RAISED INTERPOLATION MARKER is placed at a point in the text where another version has additional text. This additional text is given in the apparatus criticus. If there is a second piece of interpolated text in one verse, the dotted form U+2E07 RAISED DOTTED INTERPOLATION MARKER is used instead.

U+2E08 DOTTED TRANSPOSITION MARKER is placed at the start of a word or verse that has been transposed. The transposition is explained in the apparatus criticus. When the words are preserved in different order in some manuscripts, U+2E09 LEFT TRANSPOSITION BRACKET is used. The end of such a sequence of words is marked by U+2E0A RIGHT TRANSPOSITION BRACKET.

The characters U+2E0B RAISED SQUARE and U+2E0C LEFT RAISED OMISSION BRACKET are conventionally used in pairs to bracket text, with RAISED SQUARE marking the start of a passage of omitted text and LEFT RAISED OMISSION BRACKET marking its end. In other editorial traditions, U+2E0C LEFT RAISED OMISSION BRACKET may be paired with U+2E0D RIGHT RAISED OMISSION BRACKET. Depending on the conventions used, either may act as the starting or ending bracket.

Two other bracket characters, U+2E1C LEFT LOW PARAPHRASE BRACKET and U+2E1D RIGHT LOW PARAPHRASE BRACKET, have particular usage in the N’Ko script, but also may be used for general editorial punctuation.

#Ancient Greek Editorial Marks. Ancient Greek scribes generally wrote in continuous uppercase letters without separating letters into words. On occasion, the scribe added punctuation to indicate the end of a sentence or a change of speaker or to separate words. Editorial and punctuation characters appear abundantly in surviving papyri and have been rendered in modern typography when possible, often exhibiting considerable glyphic variation. A number of these editorial marks are encoded in the range U+2E0E..U+2E16.

The punctuation used in Greek manuscripts can be divided into two categories: marginal or semi-marginal characters that mark the end of a section of text (for example, coronis, paragraphos), and characters that are mixed in with the text to mark pauses, end of sense, or separation between words (for example, stigme, hypodiastole). The hypodiastole is used in contrast with comma and is not a glyph variant of it.

A number of editorial characters are attributed to and named after Aristarchos of Samothrace (circa 216–144 BCE), fifth head of the Library at Alexandria. Aristarchos provided a major edition of the works of Homer, which forms the basis for modern editions.

A variety of Ancient Greek editorial marks are shown in the text of Figure 6-5, including the editorial coronis and upwards ancora on the left. On the right are illustrated the dotted obelos, capital dotted lunate sigma symbol, capital reversed lunate sigma symbol, and a glyph variant of the downards ancora. The numbers on the left indicate text lines. A paragraphos appears below the start of line 12. The opening brackets “[” indicate fragments, where text is illegible or missing in the original. These examples are slightly adapted and embellished from editions of the Oxyrhynchus Papyri and Homer’s Iliad.

#Figure 6-5. Examples of Ancient Greek Editorial Marks

U+2E0F PARAGRAPHOS is placed at the beginning of the line but may refer to a break in the text at any point in the line. The paragraphos should be a horizontal line, generally stretching under the first few letters of the line it refers to, and possibly extending into the margin. It should be given a no-space line of its own and does not itself constitute a line or paragraph break point for the rest of the text. Examples of the paragraphos, forked paragraphos,and reversed forked paragraphos are illustrated in Figure 6-6.

#Figure 6-6. Use of Greek Paragraphos

#6.2.11 Indic Punctuation

#Dandas. Dandas are phrase-ending punctuation common to the scripts of South and South East Asia. The Devanagari danda and double danda characters are intended for generic use across the scripts of India. They are also occasionally used in Latin transliteration of traditional texts from Indic scripts.

There are minor visual differences in the appearance of the dandas, which may require script-specific fonts or a font that can provide glyph alternates based on script environment. For the four Philippine scripts, the analogues to the dandas are encoded once in Hanunóo and shared across all four scripts. The other Brahmi-derived scripts have separately encoded equivalents for the danda and double danda. In some scripts, as for Tibetan, multiple, differently ornamented versions of dandas may occur. The dandas encoded in the Unicode Standard are listed in Table 6-10.

#Table 6-10. Unicode Danda Characters
Code Name
U+0964 DEVANAGARI DANDA
U+0965 DEVANAGARI DOUBLE DANDA
U+0E5A THAI CHARACTER ANGKHANKHU
U+0F08 TIBETAN MARK SBRUL SHAD
U+0F0D TIBETAN MARK SHAD
U+0F0E TIBETAN MARK NYIS SHAD
U+0F0F TIBETAN MARK TSHEG SHAD
U+0F10 TIBETAN MARK NYIS TSHEG SHAD
U+0F11 TIBETAN MARK RIN CHEN SPUNGS SHAD
U+0F12 TIBETAN MARK RGYA GRAM SHAD
U+104A MYANMAR SIGN LITTLE SECTION
U+104B MYANMAR SIGN SECTION
U+1735 PHILIPPINE SINGLE PUNCTUATION
U+1736 PHILIPPINE DOUBLE PUNCTUATION
U+17D4 KHMER SIGN KHAN
U+17D5 KHMER SIGN BARIYOOSAN
U+1AA8 TAI THAM SIGN KAAN
U+1AA9 TAI THAM SIGN KAANKUU
U+1B5E BALINESE CARIK SIKI
U+1B5F BALINESE CARIK PAREREN
U+1C3B LEPCHA PUNCTUATION TA-ROL
U+1C3C LEPCHA PUNCTUATION NYET THYOOM TA-ROL
U+1C7E OL CHIKI PUNCTUATION MUCAAD
U+1C7F OL CHIKI PUNCTUATION DOUBLE MUCAAD
U+A876 PHAGS-PA MARK SHAD
U+A877 PHAGS-PA MARK DOUBLE SHAD
U+A8CE SAURASHTRA DANDA
U+A8CF SAURASHTRA DOUBLE DANDA
U+A92F KAYAH LI SIGN SHYA
U+A9C8 JAVANESE PADA LINGSA
U+A9C9 JAVANESE PADA LUNGSI
U+AA5D CHAM PUNCTUATION DANDA
U+AA5E CHAM PUNCTUATION DOUBLE DANDA
U+AA5F CHAM PUNCTUATION TRIPLE DANDA
U+AAF0 MEETEI MAYEK CHEIKHAN
U+ABEB MEETEI MAYEK CHEIKHEI
U+10A56 KHAROSHTHI PUNCTUATION DANDA
U+10A57 KHAROSHTHI PUNCTUATION DOUBLE DANDA
U+11047 BRAHMI DANDA
U+11048 BRAHMI DOUBLE DANDA
U+110C0 KAITHI DANDA
U+110C1 KAITHI DOUBLE DANDA
U+11141 CHAKMA DANDA
U+11142 CHAKMA DOUBLE DANDA
U+111C5 SHARADA DANDA
U+111C6 SHARADA DOUBLE DANDA
U+11238 KHOJKI DANDA
U+11239 KHOJKI DOUBLE DANDA
U+113D4 TULU-TIGALARI DANDA
U+113D5 TULU-TIGALARI DOUBLE DANDA
U+1144B NEWA DANDA
U+1144C NEWA DOUBLE DANDA
U+115C2 SIDDHAM DANDA
U+115C3 SIDDHAM DOUBLE DANDA
U+11641 MODI DANDA
U+11642 MODI DOUBLE DANDA
U+1173C AHOM SIGN SMALL SECTION
U+1173D AHOM SIGN SECTION
U+11944 DIVES AKURU DOUBLE DANDA
U+11A42 ZANABAZAR SQUARE MARK SHAD
U+11A43 ZANABAZAR SQUARE MARK DOUBLE SHAD
U+11A9B SOYOMBO MARK SHAD
U+11A9C SOYOMBO MARK DOUBLE SHAD
U+11C41 BHAIKSUKI DANDA
U+11C42 BHAIKSUKI DOUBLE DANDA
U+11C71 MARCHEN MARK SHAD
U+11F43 KAWI DANDA
U+11F44 KAWI DOUBLE DANDA
U+16A6E MRO DANDA
U+16A6F MRO DOUBLE DANDA
U+16D6E KIRAT RAI DANDA
U+16D6F KIRAT RAI DOUBLE DANDA

The Bidirectional Class of the dandas matches that for the scripts they are intended for. Kharoshthi, which is written from right to left, has Bidirectional Class R for U+10A56 KHAROSHTHI PUNCTUATION DANDA. For more on bidirectional classes, see Unicode Standard Annex #9, “Unicode Bidirectional Algorithm.”

Note that the name of the danda in Hindi is viram, while the different Unicode character named virama is called halant in Hindi. If this distinction is not kept in mind, it can lead to confusion as to which character is meant.

#6.2.12 CJK Punctuation

CJK Punctuation comprises punctuation marks and symbols used by writing systems that employ Han ideographs. Most of these characters are found in East Asian standards. Typical for many of these wide punctuation characters is that the actual image occupies only the left or the right half of the normal square character cell. The extra whitespace is frequently removed in a kerning step during layout, as shown in Figure 6-7. Unlike ordinary kerning, which uses tables supplied by the font, the character space adjustment of wide punctuation characters is based on their character code.

#Figure 6-7. CJK Parentheses

U+3000 IDEOGRAPHIC SPACE is provided for compatibility with legacy character sets. It is a fixed-width wide space appropriate for use with an ideographic font. For more information about wide characters, see Unicode Standard Annex #11, “East Asian Width.”

U+3030 WAVY DASH is a special form of a dash found in East Asian character standards. (For a list of other space and dash characters in the Unicode Standard, see Table 6-2 and Table 6-3.)

U+3037 IDEOGRAPHIC TELEGRAPH LINE FEED SEPARATOR SYMBOL is a visible indicator of the line feed separator symbol used in the Chinese telegraphic code. It is comparable to the pictures of control codes found in the Control Pictures block.

U+3005 IDEOGRAPHIC ITERATION MARK is used to stand for the second of a pair of identical ideographs occurring in adjacent positions within a document.

U+3006 IDEOGRAPHIC CLOSING MARK is used frequently on signs to indicate that a store or booth is closed for business. The Japanese pronunciation is shime, most often encountered in the compound shime-kiri.

The U+3008 and U+3009 angle brackets are unambiguously wide, as are other bracket characters in this block, such as double angle brackets, tortoise shell brackets, and white square brackets. Where mathematical and other non-CJK contexts use brackets of similar shape, the Unicode Standard encodes them separately.

U+3012 POSTAL MARK is used in Japanese addresses immediately preceding the numerical postal code. It is also used on forms and applications to indicate the blank space in which a postal code is to be entered. U+3020 POSTAL MARK FACE and U+3036 CIRCLED POSTAL MARK are properly glyphic variants of U+3012 and are included for compatibility.

U+3031 VERTICAL KANA REPEAT MARK and U+3032 VERTICAL KANA REPEAT WITH VOICED SOUND MARK are used only in vertically written Japanese to repeat pairs of kana characters occurring immediately prior in a document. The voiced variety U+3032 is used in cases where the repeated kana are to be voiced. For instance, a repetitive phrase like toki-doki could be expressed as <U+3068, U+304D, U+3032> in vertical writing. Both of these characters are intended to be represented by “double-height” glyphs requiring two ideographic “cells” to print; this intention also explains the existence in source standards of the characters representing the top and bottom halves of these characters (that is, the characters U+3033, U+3034, and U+3035). In horizontal writing, similar characters are used, and they are separately encoded. In Hiragana, the equivalent repeat marks are encoded at U+309D and U+309E; in Katakana, they are U+30FD and U+30FE.

#Wave Dash. U+301C WAVE DASH is a compatibility character that was originally encoded to represent the character in the JIS C 6226-1978 standard and all subsequent revisions and extensions with the kuten code: 1-33 (0x8160 in Shift-JIS encoding). The mapping of this character has been problematical. Some major implementations originally mapped, and continue to map for compatibility purposes, that JIS character to U+FF5E FULLWIDTH TILDE, instead. The mapping issue has been documented in the Unicode Standard since Version 3.0.

From Version 2.0 through Version 7.0 of the Unicode Standard, U+301C was shown in the code charts with a representative glyph that had a wide reversed tilde shape. Starting with Version 8.0, however, the representative glyph has been corrected to a wide tilde shape, to reflect predominant practice in commercial fonts. For most purposes, U+301C WAVE DASH should be treated simply as a duplicate representation of U+FF5E FULLWIDTH TILDE.

#Sesame Dots. U+FE45 SESAME DOT and U+FE46 WHITE SESAME DOT are used in vertical text, where a series of sesame dots may appear beside the main text, as a sidelining to provide visual emphasis. In this respect, their usage is similar to such characters as U+FE34 PRESENTATION FORM FOR VERTICAL WAVY LOW LINE, which are also used for sidelining vertical text for emphasis. Despite being encoded in the block for CJK compatibility forms, the sesame dots are not compatibility characters. They are in general typographic use and are found in the Japanese standard, JIS X 0213.

U+FE45 SESAME DOT is historically related to U+3001 IDEOGRAPHIC COMMA, but is not simply a vertical form variant of it. The function of an ideographic comma in connected text is distinct from that of a sesame dot.

#6.2.13 Unknown or Unavailable Ideographs

U+3013 GETA MARK is used to indicate the presence of, or to hold a place for, an ideograph that is not available when a document is printed. It has no other use. Its name comes from its resemblance to the mark left by traditional Japanese sandals (geta). A variety of light and heavy glyphic variants occur.

U+303E IDEOGRAPHIC VARIATION INDICATOR is a graphic character that is to be rendered visibly. It alerts the user that the intended character is similar to, but not equal to, the character that follows. Its use is similar to the existing character U+3013 GETA MARK. A GETA MARK substitutes for the unknown or unavailable character, but does not identify it. The IDEOGRAPHIC VARIATION INDICATOR is the head of a two-character sequence that gives some indication about the intended glyph or intended character. Ultimately, the IDEOGRAPHIC VARIATION INDICATOR and the character following it are intended to be replaced by the correct character, once it has been identified or a font resource or input resource has been provided for it.

U+303F IDEOGRAPHIC HALF FILL SPACE is a visible indicator of a display cell filler used when ideographic characters have been split during display on systems using a double-byte character encoding. It is included in the Unicode Standard for compatibility.

See also “Ideographic Description Sequences” in Section 18.2, Ideographic Description Characters.

#6.2.14 CJK Compatibility Forms

#Vertical Forms. CJK vertical forms are compatibility characters encoded for compatibility with legacy implementations that encode these characters explicitly when Chinese text is being set in vertical rather than horizontal lines. The preferred Unicode approach to representation of such text is to simply use the nominal characters that correspond to these vertical variants. Then, at display time, the appropriate glyph is selected according to the line orientation.

The Unicode Standard contains two blocks devoted primarily to these CJK vertical forms. The CJK Vertical Forms block, U+FE10..U+FE1F, contains compatibility characters needed for round-trip mapping to the Chinese standard, GB 18030. The CJK Compatibility Forms block, U+FE30..U+FE4F, contains forms found in the Chinese standard, CNS 11643.

#Styled Overscores and Underscores. The CJK Compatibility Forms block also contains a number of compatibility characters from CNS 11643, which consist of different styles of overscores or underscores. They were intended, in the Chinese standard, for the representation of various types of overlining or underlining, for emphasis of text when laid out horizontally. Except for round-trip mapping with legacy character encodings, the use of these characters is discouraged; use of styles is the preferred way to handle such effects in modern text rendering.

#Small Form Variants. CNS 11643 also contains a number of small variants of ASCII punctuation characters. The Unicode Standard encodes those variants as compatibility characters in the Small Form Variants block, U+FE50..U+FE6F. Those characters, while construed as fullwidth characters, are nevertheless depicted using small forms that are set in a fullwidth display cell. (See the discussion in Section 18.4, Hiragana and Katakana.) These characters are provided for compatibility with legacy implementations.

Two small form variants from CNS 11643/plane 1 were unified with other characters outside the ASCII block: 213116 was unified with U+00B7 MIDDLE DOT, and 226116 was unified with U+2215 DIVISION SLASH.

#Fullwidth and Halfwidth Variants. For compatibility with East Asian legacy character sets, the Unicode Standard encodes fullwidth variants of ASCII punctuation and halfwidth variants of CJK punctuation. See Section 18.5, Halfwidth and Fullwidth Forms, for more information.