Code page 955

JIS X 0208
Company type	Japanese Industrial Standard
Industry	Character encoding standard
Predecessor	JIS C 6226 (initial name)
Founded	1978
Headquarters	Japan
Products	Two-byte character set for Japanese text (6,879 graphical symbols)

Adapted from the English Wikipedia article Code page 955, available under CC BY-SA 4.0Updated 9 August 2025

JIS X 0208 represents a two-byte character set standardized under the Japanese Industrial Standard, encompassing 6,879 graphical symbols utilized for writing Japanese text, including place names and personal names. Its official designation at present is . Initially introduced as JIS C 6226 in 1978, this standard underwent revisions in the years 1983, 1990, and 1997. IBM identifies this character set as Code page 952, while the original 1978 edition is known as Code page 955.

Usage scope and interoperability

The primary purpose of JIS X 0208 is to facilitate data and text interchange among data processing systems and interconnected devices or communication networks.

Implementations that only partially support the standard are not deemed compatible. Early versions categorized characters into level 1 and level 2 groups, with some variant characters (異体字, itaiji) being reassigned between levels in subsequent updates. This indicates that Japanese computer systems focusing solely on non-kanji or level 1 characters were contemplated but never officially accepted as compliant—although early hardware such as the NEC PC-9801 did implement such subsets.

Although JIS X 0208:1997 addresses compatibility considerations, it is generally not interpreted as an official certification of compatibility nor as a manufacturing standard guaranteeing self-consistency. Therefore, no products genuinely labeled as JIS X 0208-"compatible" exist in practice. The terms and appear within the standard but are subject to diverse interpretations among users.

Lead byte

The initial byte in the encoding corresponds to the row or cell number plus an offset of 0x20 (decimal 32). For instance, a code beginning with 0x21 indicates row number one, with its first cell continuing at byte 0x21 (decimal 33).

For lead bytes representing non-kanji characters, links direct to relevant charts on this page displaying those symbols. In contrast, lead bytes associated with kanji characters link to the pertinent section of Wiktionary's kanji index.

Character Set 0x21 (Row 1, Special Symbols)

Different manufacturers may assign varying Unicode values to this character set. For instance, Microsoft maps kuten 1-29 (JIS 0x213D) to U+2015 (Horizontal Bar), whereas Apple associates it with U+2014 (Em Dash). Likewise, Microsoft designates kuten 1-61 (JIS 0x215D) as U+FF0D (fullwidth Hyphen-Minus), while Apple uses U+2212 (Minus Sign). The wave dash character's Unicode representation also differs among vendors.

When combined with JIS X 0208 in encoding systems such as Shift JIS, EUC-JP, or ISO 2022-JP, ASCII and JISCII punctuation marks (highlighted in this section) might alternatively be mapped to the Halfwidth and Fullwidth Forms Unicode block.

Template:Chset-cell1 |style=font-size:small;background:#DFD}}

Template:Chset-left1

Template:Chset-cell1 |style=font-size:small;background:#DFD}}

Template:Chset-left1

Template:Chset-cell1 |style=font-size:small;background:#DFD}}

Template:Chset-left1

Template:Chset-cell1 |style=font-size:small;background:#DFD}}

Template:Chset-left1

Template:Chset-cell1 |style=font-size:small;background:#DFD}}

Character Set 0x22 (Row 2, Special Symbols)

Most symbols within this collection were added in the 1983 revision, except for the range 0x2221–0x222E (kuten 2-1 through 2-14), which were included in the initial 1978 specification.


Template:Chset-left1	Template:Chset-cell1	Template:Chset-cell1	Template:Chset-cell1	Template:Chset-cell1	Template:Chset-cell1	Template:Chset-cell1	Template:Chset-cell1	Template:Chset-cell1	Template:Chset-cell1	Template:Chset-cell1	Template:Chset-cell1	Template:Chset-cell1	Template:Chset-cell1	Template:Chset-cell1	Template:Chset-cell1
Template:Chset-left1

}}

Row	Column	Unicode Code Point	Symbol	Description	Notes

{-}

Character set 0x23 (row 3, digits and Roman letters)

This character set comprises a portion of the ISO 646 invariant characters, which also form part of both ASCII and the JIS X 0201 Roman character collection, but excludes punctuation marks and symbols. It features western Arabic numerals along with uppercase and lowercase letters from the Basic Latin alphabet. When used alongside JIS X 0208 and ASCII or JIS X 0201 encodings such as EUC-JP, Shift JIS, or ISO 2022-JP, these characters may be represented using alternative Unicode code points found in the Halfwidth and Fullwidth Forms section.

This row is identical to row 3 in KPS 9566. By comparison, both row 3 of KS X 1001 and row 3 of GB 2312 incorporate their complete national ISO 646 variants rather than restricting to the alphanumeric subset.

... (rest of table omitted for brevity) ... }

Character set 0x24 (row 4, Hiragana)

This row consists of Japanese Hiragana syllabary characters.

For comparison, see row 4 of GB 2312, which corresponds exactly to this row. Additionally, rows in different positions but with identical layouts include row 10 of both KPS 9566 and KS X 1001.

... [table continues similarly with remaining Hiragana characters] ...

Character set 0x25 (row 5, Katakana)

This section features Japanese Katakana symbols.

It matches the layout found in row 5 of GB 2312, sharing an identical arrangement. Comparable patterns can be observed in row 11 of KPS 9566 and KS X 1001, although their positions vary. This differs from the unique Katakana configuration present in JIS X 0201.

... [table continues with consistent formatting, omitted for brevity] ...

Character set 0x26 (row 6, Greek)

This row encompasses fundamental characters of the contemporary Greek alphabet, excluding diacritical marks and the final sigma.

It corresponds to row 6 of GB 2312 and GB 12345, as well as row 6 of KPS 9566, which all share a similar Greek letter arrangement—though GB 12345 incorporates vertical presentation forms and the Korean standard includes Roman numerals. Conversely, row 5 of KS X 1001 shifts Greek letters to give precedence to Roman numerals.

Template:Chset-left1

Template:Chset-cell1

Character set 0x27 (Row 7, Cyrillic)

This row features the contemporary Russian alphabet, though it might not comprehensively cover all Cyrillic script variations.

It corresponds to row 7 of GB 2312 and exhibits a similar arrangement to row 12 of KS X 1001 and row 5 of KPS 9566, albeit in a different position.

... (rest of table unchanged)

Character set 0x28 (Row 8, Box Drawing)

All characters within this set were introduced in 1983, as they were not part of the initial 1978 specification.

Extension character set 0x2D (row 13, NEC-specific characters)

Rows 9 through 15 in the JIS X 0208 standard remain unassigned.

Nevertheless, the row 13 configuration introduced by NEC serves as a widely adopted extension (with slight modifications) employed by Windows-932 (which aligns with the WHATWG Encoding Standard for HTML5), the PostScript version of MacJapanese (starting from KanjiTalk version 7), and JIS X 0213, the successor to JIS X 0208. Notably, Windows-932/WHATWG and JIS X 0213 share identical characters in this row, enhancing compatibility during decoding.

... (rest of table omitted for brevity) ...

Code structure

To denote code points, single-byte codes utilize column and line numbers, whereas two-byte codes are identified by their kuten values. Character designations provide a code-independent reference.

Single byte codes

While most JIS X 0208 graphic characters are encoded using two bytes with at least seven bits each, control characters and the standard (space) character, excluding ideographic space, employ one-byte codes. The bit combination (ビット組合せ) for a single-byte code is represented by two decimal figures: a column number and a line number. These correspond to segments of the seven or eight-bit code and align with hexadecimal notation. For instance, the space character’s bit pattern is 010 0000 in 7-bit form or 0010 0000 in 8-bit form, which corresponds to column/line notation as 2/0, hexadecimal 0x20, or decimal value 32.

Code points and code numbers

Double-byte coding is structured into 94 groups known as , each group containing 94 individual codes called . This configuration results in a total of 8,836 (94 × 94) potential code points, arranged within a 94-by-94 matrix.

A kuten (区点) code point is defined by a combination of a row number and a cell number, both ranging from 1 to 94. The notation for a is given as "row-cell," separated by a hyphen; for instance, the character "" is located at row 16, cell 1, thus its code number is represented as "16-01."

In the context of 7-bit JIS X 0208 (utilized in JIS X 0202 / ISO-2022-JP), both bytes fall within the range of 0x21 (corresponding to row/cell 1) up to 0x7E (row/cell 94), aligning with printable ASCII characters except for the space character. To obtain encoded bytes, 0x20 (decimal 32) is added to each number. For example, the code "16-01" ("亜") translates to the byte sequence 0x30 0x21. The 8-bit EUC-JP encoding employs values between 0xA1 and 0xFE (with the high bit set), whereas Shift JIS uses more intricate transformations and incorporates extensions beyond row 94.

This organizational system is also applied in Mainland China's GB 2312 (区位 qūwèi) and South Korea's KS C 5601/KS X 1001 standards, where rows and cells are referred to as hang (행) and yol (열). The subsequent JIS X 0213 standard expands upon this framework by introducing multiple planes of rows, a design shared with CNS 11643 and related to CCCII.

Unassigned code points

Within the two-byte code space, rows numbered from 9 to 15 and from 85 to 94 are designated as unassigned code points (空き領域, aki ryōiki), indicating that no characters have been allocated there. Additionally, certain cells in other rows remain unassigned.

These regions are generally intended to be left unused. Characters (gaiji) intended for information exchange should not be assigned to these unallocated points without prior agreement. Even when new characters are allocated, it is imperative that none of the graphic characters defined by the standard are duplicated or assigned across multiple unassigned locations.

Special attention must be paid regarding Han unification of kanji glyphs. For example, row 25 cell 66 represents the kanji meaning "high" or "expensive," which merges both the common variant with the "mouth" radical (高) and the less frequent "ladder" radical form (髙) into a single code point. Assigning the latter form separately to an unassigned code point would contravene the standard's rules.

In actual usage, vendor-specific Shift JIS variants such as Windows-932 and MacJapanese encode their extensions within these unallocated rows. Most of the codes left unused in JIS X 0208 have since been assigned by the more recent JIS X 0213 standard.

Character names

Every character in JIS X 0208 possesses a name that facilitates identification independently of its code value. These names correspond with other standards, particularly the Universal Coded Character Set (UCS/Unicode), which aids in mapping characters to Unicode.

For example, both ISO/IEC 646 (US-ASCII) column 4 line 1 and JIS X 0208 row 3 cell 33 carry the name "LATIN CAPITAL LETTER A," signifying their equivalence despite differing encodings. Conversely, certain ASCII characters like quotation mark, apostrophe, hyphen-minus, and tilde are not included in this standard.

Names for non-kanji characters consist of uppercase Roman letters, spaces, and hyphens and also have commonly used Japanese names (日本語通用名称). Kanji names are generated algorithmically from their UCS/Unicode hexadecimal code points by prefixing "CJK UNIFIED IDEOGRAPH-". For instance, row 16 cell 1 (亜) corresponds to U+4E9C and is named "CJK UNIFIED IDEOGRAPH-4E9C." Kanji do not receive Japanese common names.

Overview

JIS X 0208 specifies a collection of 6879 graphic characters encoded in two-byte sequences, collectively referred to as the kanji set (漢字集合, kanji shūgō). This set comprises 6355 kanji characters alongside 524 non-kanji (非漢字, hikanji) symbols, which include Latin letters and kana.

Special characters: Rows 1 and 2 feature a total of 147 characters, such as 18 descriptor symbols including the ideographic space and Japanese punctuation marks; eight diacritical marks like dakuten; 10 kana- or kanji-related characters; 22 bracket symbols; 45 mathematical symbols; and 32 unit indicators encompassing currency and postal signs.
Numerals: A segment of row 3 contains the digits from "0" through "9".
Latin letters: Another portion of row 3 holds the full set of 26 uppercase and lowercase Latin letters (totaling 52).
Hiragana: Row 4 includes 83 kana characters covering unvoiced, voiced, semi-voiced, and small kana forms.
Katakana: Row 5 presents 86 characters, consisting of katakana counterparts to hiragana plus small ka/ke and vu kana.
Greek letters: Row 6 incorporates 48 Greek letters in both uppercase and lowercase forms (excluding the final sigma).
Cyrillic letters: Row 7 contains 66 uppercase and lowercase Russian alphabet letters.
Box-drawing characters: Row 8 consists of 32 segments in thin, thick, and mixed styles.
Kanji: Rows 16 through 84 encompass levels 1 and 2, totaling 6355 kanji.

Special characters, numerals, and Latin characters

Certain ASCII characters from the International Reference Version (IRV) of ISO/IEC 646:1991—specifically QUOTATION MARK, APOSTROPHE, HYPHEN-MINUS, and TILDE—are either omitted or divided into multiple code points within JIS X 0208. Notably, the TILDE lacks a direct counterpart but is approximated by the WAVE DASH character.

The organization of these characters in the kanji set departs from ISO/IEC 646 conventions. While it shares ninety special symbols, numerals, and Latin letters with IRV, their placement differs across rows. For instance, row 3 aligns with ISO for letters and digits; however, overall JIS X 0208 is not upward-compatible with ISO/IEC 646.

These full-width alphanumeric glyphs (全角英数字, zenkaku eisūji) reflect design decisions that create incompatibilities with IRV. From the initial standard edition onward, composite symbols such as encircled numbers, unit ligatures, and Roman numerals have been representable through composition rather than assigned distinct code points. Vendors commonly provide these as proprietary additions instead of seeking their inclusion in the official standard.

The fourth edition (1997) clarified that such composite symbols are spacing characters that advance the cursor position and forbade their representation by composition. Consequently, Latin letters with diacritics are generally prohibited except for limited exceptions like the ångström symbol (Å).

Hiragana and katakana

Unlike JIS X 0201, the hiragana and katakana sets in JIS X 0208 incorporate dakuten and handakuten as standalone characters. Additionally, obsolete katakana such as and , along with the small , which are missing from JIS X 0201, are included here.

The ordering of kana in JIS X 0208 differs from that in JIS X 0201. While JIS X 0201 sequences katakana beginning with , followed by small kana then full-size kana arranged in gojūon order, JIS X 0208 organizes kana primarily by gojūon sequence then groups them by "small kana, full-size kana, kana with dakuten, and kana with handakuten," placing base kana alongside their derived forms. This arrangement supports more effective dictionary sorting based on kana. However, this departure from JIS X 0201’s katakana sequence contributed to the creation of "half-width kana" as a compatibility measure—a recognized limitation inherent to this standard.

Kanji

The process of selecting, dividing into Level 1 and Level 2, and organizing kanji within this standard is thoroughly explained in the fourth standard (1997). The original set of 6,349 kanji from the first standard (1978) was derived from four principal kanji compilations:

Created in 1971 by the Information Processing Society of Japan's kanji code committee, this list comprised roughly 6,086 characters.

Formulated by Japan's Administrative Management Agency in 1975, it contained 2,817 characters. The selection was informed by comparative studies analyzing various kanji lists.

A now-obsolete list featuring 3,044 characters that contributed data to the "Correspondence Analysis Results" used during standardization.

Including 3,251 characters, this list represented kanji found in administrative geographic names gathered by the Japan Geographic Data Center.

Later updates—the second and third standards—introduced six additional characters to Level 2 and adjusted certain character forms and level assignments to enhance consistency and practical application.

Level partitioning

Level 1 comprises 2,965 kanji located in rows 16 through 47, while Level 2 contains 3,390 kanji positioned in rows 48 to 84.

Kanji in Level 1 were chosen based on their widespread usage and frequency across multiple glyph lists, referencing sources such as the tōyō kanji, its revision draft, and the jinmeiyō kanji. Codes like JIS C 6260 (prefecture identifiers) and JIS C 6261 (municipality codes) influenced the inclusion of kanji representing nearly all Japanese prefectures, cities, and towns. Expert recommendations were also incorporated.

Level 2 includes kanji present in these major lists but excluded from Level 1, often due to ambiguous pronunciations or lower usage frequency at the time of selection. Consequently, Level 1 generally contains more commonly used kanji; however, usage patterns have evolved—some Level 2 kanji such as "to soar" (翔) and "to glitter" (煌) have gained popularity, whereas certain Level 1 kanji like those for "centimeter" (糎) and "millimeter" (粍) have seen diminished use. Among the current jōyō kanji, thirty belong to Level 2 with three not included at all; for jinmeiyō kanji, 192 are classified in Level 2 while 105 remain outside this standard.

Arrangement

Kanji in Level 1 are organized according to a designated "representative reading," which may be an on (Sino-Japanese) or kun (native Japanese) reading, arranged following the gojūon order. Typically, the dominant on reading is selected; if none exists or is uncommon, a kun reading is applied—verbs utilize the ren'yōkei (continuative) form. For example, row 16 cells 1–41 begin with readings starting with a, including some based on kun readings like 葵 ("aoi") and 粟 ("awa").

When multiple kanji share the same representative reading, those with on readings precede those with kun. If readings coincide exactly, sorting is determined by primary radical followed by stroke count.

Variant forms (itaiji) are placed immediately after their standard counterparts in both levels. For instance, in Level 2 following cell 49-88 (劍), three variant forms (劔, 劒, 剱) appear consecutively despite not following stroke count order. The main variant (剣) is listed separately in Level 1 at row 23 cell 85; another variant (釼) is grouped under the "gold" radical within Level 2.

Level 2 kanji are primarily organized by radical and stroke count; ties are broken according to reading order.

Kanji with Unverified Origins

Kanji in JIS X 0208:1997 Appendix 7 with unclear or unidentified sources
Kuten	Character	Source Status
52-55	墸	Unknown origin
52-63	壥	Unknown origin
54-12	妛	Source uncertain
55-27	彁	Unidentifiable source
57-43	挧	Source uncertain
58-83	暃	Source uncertain
59-91	椦	Source uncertain
60-57	槞	Source uncertain
74-12	蟐	Source uncertain
74-57	袮	Source uncertain
79-64	閠	Source uncertain
81-50	駲	Source uncertain

Certain kanji included in this collection do not appear in comprehensive dictionaries, and their provenance remains unknown. For instance, Tajima (1979) highlighted 63 kanji absent from major references such as Shinjigen and Dai Kan-Wa jiten, noting these characters did not correspond to recognized abbreviations. These particular kanji came to be referred to as or .

The committee responsible for the fourth edition undertook an investigation into the original references used by earlier committees. They discovered that the initial committee primarily depended on the "Correspondence Analysis Results," which incorporated kanji from sources like "Japanese Personality Registration Name Kanji" and "Kanji for National Administrative District Listing." However, no original documents exist for the former source, and the latter was examined through preliminary drafts, classical texts, and personal name compilations such as NTT telephone directories.

As a result of this scrutiny, the count of kanji with unverifiable origins was reduced to twelve (as listed above). Some characters seem to have arisen from transcription mistakes; for example, 妛 is believed to have originated from a printing error when combining 山 and 女 to form 𡚴, inadvertently adding an extra stroke.

Consolidation of Kanji Variants

In the fourth edition standard (1997), the term unification (包摂, hōsetsu) denotes assigning identical code points to characters that differ in form. This concept resembles but is distinct from Unicode's Han unification. The standard restricts permissible glyph variations and explicitly outlines how specific allographic glyphs merge into a single grapheme code point.

A represents an abstract graphical concept of a character, whereas a refers to its actual visual appearance, such as handwritten or printed forms. Differences between various renditions of a single glyph are termed .

Unification hinges on both a code point’s and the applied . Glyphs conforming to these conditions share one code point. For instance, at position 33-46 (僧), the character combines radical 9 (亻) with the kanji 曽. Under criterion 101, three variants—common Japanese 曽, traditional 曾, and an inverted radical version 曾—are all unified under one code point.

The fourth standard specifies 186 unification criteria. When example glyphs comprise multiple components, criteria apply separately to each; once a criterion affects a part, no subsequent criteria alter it. Additionally, criteria cannot unify glyphs identical to other code points.

Example glyphs serve solely as references for code points and are not official design endorsements. Unification criteria mainly guide assignment of frequently used kanji; generating unused characters based on these guidelines is discouraged.

Kanji selection within this set does not always adhere strictly to unification criteria. For example, entry 41-7 includes both crossing (彥) and non-crossing (彦) stroke variants according to criterion 72, whereas entries 20-73 and 80-90 correspond exclusively to one form each (顔 and 顏 respectively).

The terms "unification," "unification criteria," and "example glyph" were introduced in this fourth edition. Previous standards categorized kanji relationships as "independent" (独立), "compatible" (対応), and "equivalent" (同値), explaining that equivalent characters consolidate into one code point despite minor stylistic variations.

From the initial standard onward, specific design details of character forms were excluded from scope; emphasis remained on defining characters and their encoding rather than graphical representation. The fourth edition reiterates this principle by regulating graphic characters and bit patterns while excluding usage and detailed design aspects (JIS X 0208:1997, item 1).

Criteria for Unification to Ensure Compatibility

The fourth criterion outlines unification standards aimed at preserving backward compatibility (過去の規格との互換性を維持するための包摂規準). These standards pertain to 29 code points exhibiting notable glyph variations between JIS C 6226-1983 and earlier JIS C 6226-1978. For these characters, the glyphs from the 1983 edition are designated as "A," while those from the 1978 edition are marked as "B." Both variants are permitted; however, any compatibility assertion must clearly specify which glyph form is applied for each relevant code point.

Encoding Formats Defined by JIS X 0208

Article 7 and Appendices 1 and 2 of JIS X 0208:1997 specify eight distinct encoding formats.

The regions "CL" (control left), "GL" (graphic left), "CR" (control right), and "GR" (graphic right) correspond to column/row ranges of 0/0–1/15, 2/1–7/14, 8/0–9/15, and 10/1–15/14 respectively. The graphic character "SPACE" occupies code 2/0, while the control character "DELETE" resides at 7/15. The C0 control characters defined in JIS X 0211 (aligned with ISO/IEC 6429) are assigned to the CL region.

7-bit kanji encoding: Specified within the standard; the JIS X 0208 double-byte set is mapped to the GL area.
8-bit kanji encoding: Similar to the 7-bit scheme but uses full 8-bit bytes; the CR area may be unused or contain C1 control characters from JIS X 0211; GR remains unassigned.
International Reference Version plus 7-bit kanji encoding: The IRV (ISO/IEC 646:1991, synonymous with US-ASCII) is allocated to GL via shift-in, with shift-out switching to the JIS X 0208 double-byte set.
Latin letters plus 7-bit kanji encoding: Functions like the previous method but substitutes IRV with ISO/IEC 646:JP (the Roman set of JIS X 0201).
International Reference Version plus 8-bit kanji encoding: Places ISO/IEC 646:IRV in GL and JIS X 0208 in GR; essentially a subset of EUC-JP, excluding half-width katakana (JIS X 0201) and additional kanji (JIS X 0212).
Latin letters plus 8-bit kanji encoding: Mirrors IRV+8-bit but replaces IRV with ISO/IEC 646:JP.
Shift-coded character set: Outlined in Appendix 1 as Shift-Coded Representation (シフト符号化表現); serves as the authoritative definition of Shift JIS.
RFC 1468-coded character set: Defined in Appendix 2; resembles ISO-2022-JP but employs eight-bit bytes instead of seven.

Among these schemes, only the Shift-coded character set holds registration with the IANA. The others bear close relation to IANA-registered encodings such as EUC-JP and ISO-2022-JP.

Escape Sequences for JIS X 0202 / ISO 2022

JIS X 0208 is utilized within ISO 2022/JIS X 0202 frameworks (including ISO-2022-JP). The following escape sequences designate JIS X 0208 for each of the four ISO 2022 code sets, where "ESC" denotes the control character "Escape" (0x1B).

ISO 2022 escape sequences for selecting JIS C 6226 and JIS X 0208
Standard	G0	G1	G2	G3
78	ESC 2/4 4/0	ESC 2/4 2/9 4/0	ESC 2/4 2/10 4/0	ESC 2/4 2/11 4/0
83	ESC 2/4 4/2	ESC 2/4 2/9 4/2	ESC 2/4 2/10 4/2	ESC 2/4 2/11 4/2
90 onward	ESC 2/6 4/0 ESC 2/4 4/2	ESC 2/6 4/0 ESC 2/4 2/9 4/2	ESC 2/6 4/0 ESC 2/4 2/10 4/2	ESC 2/6 4/0 ESC 2/4 2/11 4/2

The ESC 2/4 sequence indicates selection of a multibyte character set, while ESC 2/6 signals a revision update. The identifier byte for JIS C 6226:1978 is 4/0 (@), whereas both JIS C 6226:1983 and JIS X 0208:1983 use byte 4/2 (B). For JIS X 0208:1990, byte 4/2 remains but includes an additional revision identifier at byte 4/0.

Duplicate encodings in ASCII and JIS X 0201

When integrating the kanji repertoire with ISO/IEC 646:1991 IRV (commonly known as ASCII) or the JIS X 0201 Latin character set (JIS-Roman), overlapping characters may result in duplicate code assignments due to differing code point allocations.

To resolve this, JIS X 0208:1997 generally prohibits assigning kanji set code points to characters present in both collections, considering identically named characters as equivalent. For example, the character "LATIN CAPITAL LETTER A" exists at bit pattern 4/1 in ASCII and at row 3 cell 33 within the kanji set; the latter encoding is disallowed to avoid redundancy.

Nonetheless, for legacy support, kanji set code points can be interpreted as full-width variants distinct from ASCII or JIS-Roman counterparts. Therefore, the bit pattern for row 3 cell 33 may serve to represent a full-width "A".

Strict adherence to the standard does not always ensure unique encoding. The character "HYPHEN-MINUS," for instance, might be encoded either as bit pattern 2/13 or as kanji set row 1 cell 30 ("HYPHEN"), with no explicit preference indicated. Similar uncertainties apply to minus signs and quotation marks.

Moreover, the simultaneous presence of the full-width IDEOGRAPHIC SPACE at row 1 cell 1 and the half-width space at bit pattern 2/0 lacks clear differentiation within the standard, complicating efforts to achieve unambiguous character representation.

History

Japanese Industrial Standards are subject to reaffirmation, revision, or withdrawal within five years following their establishment or latest update. Since its original release, this standard has undergone three revisions; the currently effective version is the fourth iteration.

Initial standard

The inaugural standard, JIS C 6226-1978, was promulgated by Japan’s Minister of International Trade and Industry on January 1, 1978. Known as 78JIS, it was formulated by a JIPDEC committee led by Moriguchi Shigeichi. This edition encompassed 453 non-kanji characters—including Hiragana, Katakana, Roman, Greek, Cyrillic alphabets, and punctuation—and 6349 kanji (2965 in level 1 and 3384 in level 2), totaling 6802 characters. Notably, box-drawing characters were absent at this stage. The standard utilized Ishii Mincho typeface developed by Shaken Co., Ltd.

Second standard

The subsequent revision, JIS C 6226-1983, updated the original on September 1, 1983, and is often referred to as 83JIS. Created by a JIPDEC committee chaired by Motooka Tōru under commission from AIST, this update took into account developments such as the publication of the jōyō kanji, enforcement of jinmeiyō kanji, and standardization of Japanese-language Teletex by the Ministry of Posts and Telecommunications. It also harmonized with JIS C 6234-1983 (24-pixel matrix printer character forms; currently JIS X 9052).

Additions and revisions: 39 special characters were incorporated, selected from standards like JIS Z 8201-1981 (mathematical symbols) and JIS Z 8202-1982 (units, quantities, chemical symbols) where previous composition was insufficient.; 32 box-drawing characters were introduced for the first time.; Code points for 22 variant kanji pairs were exchanged between levels 1 and 2 to enhance structural organization.; Four kanji received new code assignments in row 84 to accommodate variant forms.; Approximately 300 kanji glyphs were modified—primarily simplifying level 1 characters from Kangxi-style variants toward more contemporary shapes (for example, row 18 cell 10 underwent a transformation). Some glyphs had strokes added or removed to clarify distinctions.

These adjustments aimed to unify character forms based on criteria later formalized in the fourth standard. Simplifications predominantly affected level 1 kanji rather than level 2. Nevertheless, for around 29 code points where conflicting forms persisted into the fourth standard, special unification rules were applied to preserve backward compatibility.

Following the introduction of the "X" classification for information-related Japanese Industrial Standards, this second edition was re-designated as JIS X 0208-1983 on March 1, 1987.

Third standard

The third iteration, JIS X 0208-1990, commonly referred to as 90JIS, updated the second standard on 1 September 1990. The draft was developed by a committee from the Japanese Standards Association, led by Tajima Kazuo and commissioned by AIST.

This update involved modifications to 225 kanji glyphs and the addition of two characters to level 2 (positions 84-05 "" and 84-06 ""). It also differentiated itaiji variants that had previously been unified (such as 49-59 "" and 63-70 ""). Some revisions aligned with the 118 jinmeiyō kanji introduced in March 1990. The standard adopted the Heisei Mincho font style.

Fourth standard

Released on 20 January 1997, the fourth edition JIS X 0208:1997, or 97JIS, superseded the third standard. This revision was prepared by a JSA committee chaired by Shibano Kōji under AIST’s direction, aiming to enhance clarity and usability without altering the character set or glyph designs.

Notable improvements comprised:

Encoding schemes: While earlier versions defined encoding solely through JIS X 0202 code extensions, this edition introduced additional methods that do not require escape sequences.
Unassigned code points: It explicitly prohibited the use of unassigned code points but allowed limited usage under specified conditions.
Removal of duplicated encodings: Each character was assigned a unique "character name" consistent with other standards, ensuring only one code point is used in conjunction with ISO/IEC 646 or JIS X 0201, thereby eliminating duplicates.
Kanji source verification: Characters missing from prominent references such as the Kangxi Dictionary and Dai Kanwa Jiten were identified, and the reasoning behind their initial inclusion was reevaluated.
Kanji unification guidelines: The scope of glyph unification per code point was reinstated and clearly defined based on original drafting documents from the first standard.
Recognition of prevalent encoding practices: Widely adopted encoding formats like Shift JIS and ISO-2022-JP were formally acknowledged as "Shift-Coded Representation" and "RFC 1468-Coded Representation," reflecting their extensive use in personal computing and e-mail communications.

Successors

JIS X 0213 (extended kanji) was developed to offer a more comprehensive character repertoire suitable for contemporary Japanese text, expanding upon the kanji set established in JIS X 0208. The standard’s authors advocate transitioning to JIS X 0213 due to its alignment with the Hyōgai Kanji Glyph List and newer jinmeiyō kanji characters.

Despite its introduction in 2000, adoption has been gradual. According to a committee report from 2004, most information systems continue to rely predominantly on JIS X 0208. Support for JIS X 0213 has been incorporated in Microsoft Windows starting with Windows Vista (2006), included in Mac OS X from version 10.1 (2001), and is optionally available on many Unix-like platforms such as Linux. Compatibility improvements on personal computers are anticipated to facilitate broader acceptance over time.

Several experts predict an extended period during which JIS X 0208 and JIS X 0213 will coexist due to several challenges:

Character sets used by Japanese mobile phone devices still depend on JIS X 0208, with no announced plans for migration, limiting widespread adoption given mobile phones’ central role in communication and internet access.
JIS X 0213 lacks full upward compatibility with JIS X 0208’s unification principles, complicating data conversion for extensive archives (for example, bibliographic databases or Aozora Bunko) that require preservation of textual accuracy.
Numerous systems allocate unassigned code points within JIS X 0208 for additional characters or emoji, which conflict with codes defined in JIS X 0213, creating obstacles for migration. Although there are intentions to transition through Unicode, concerns about implementation stability have delayed progress.
Since JIS X 0213 primarily introduces less commonly used glyphs—almost doubling the total number of glyphs—it offers limited practical benefits, rendering investment less appealing where resources are limited.

Implementations

Because JIS X 0208 / JIS C 6226 functions mainly as a character set rather than a precisely defined character encoding, multiple corporations have created their own encoding schemes derived from it:

Apple: MacJapanese (based on Shift_JIS)
Fujitsu: JEF kanji code (utilizing EBCDIC)
Hitachi: KEIS (EBCDIC-based)
IBM: including IBM-932 and IBM-942 (both derived from Shift_JIS)
Microsoft: Windows-932 (Shift_JIS variant)
NEC: JIPS

Some of these encodings introduce vendor-specific characters by occupying previously unassigned code areas, for example Windows-932, MacJapanese, and NEC’s PC98 encoding. IBM-932 and IBM-942 also incorporate proprietary character assignments, but these lie outside the JIS X 0208 designated range.

ISO/IEC 646 IRV and ASCII

The kanji collection is not backward compatible with ISO/IEC 646:1991 IRV (ASCII). Nevertheless, JIS X 0208 prescribes combining the kanji set with the IRV graphic characters using either 7-bit or 8-bit coding schemes, a method also supported by EUC-JP.

JIS X 0201

JIS X 0208 omits three Latin symbols that exist in the graphic set of JIS X 0201: QUOTATION MARK, APOSTROPHE, and HYPHEN-MINUS, yet it includes all katakana characters defined in JIS X 0201.

The kanji collection can be combined with Latin letters according to JIS X 0208’s specifications, employing either 7-bit or 8-bit codes. It also allows for the integration of kanji, Latin, and katakana characters within the shift-coded format (Shift JIS). Moreover, the kanji and katakana sets are jointly usable in EUC-JP.

JIS X 0212

JIS X 0212 specifies an auxiliary kanji set occupying an additional 94-by-94 code grid for characters absent from JIS X 0208. This extension caters to information processing applications requiring a broader character repertoire.

In EUC-JP, JIS X 0212 is employed alongside JIS X 0208. Both standards serve as foundational sources for UCS/Unicode’s Han unification, enabling kanji from these sets to appear in Unicode documents.

Importantly, certain code points modified in the second edition of JIS X 0208 are mirrored in the character forms of JIS X 0212. For example, JIS X 0212 reassigns the "closure mark" from a non-kanji role in JIS X 0208 to a kanji character. Apart from this reassignment, these two sets do not share overlapping characters, which restricts JIS X 0212’s use as an independent set.

The fourth revision of JIS X 0208 removes any formal linkage to JIS X 0212, reflecting committee concerns regarding its character selection process and ambiguous identification. The standard explicitly states that combining usage with JIS X 0212 is impractical and therefore not defined.

JIS X 0213

JIS X 0213 broadens the scope of the kanji repertoire introduced by JIS X 0208, targeting a more comprehensive encoding suitable for contemporary Japanese text. It defines a total of 11,233 characters: 1,183 non-kanji and 10,050 kanji arranged across two separate 94-by-94 planes. The first plane includes non-kanji along with level 1 through level 3 kanji based on JIS X 0208. The second plane contains level 4 kanji, utilizing previously unallocated rows from JIS X 0212 to support compatibility with EUC-JP. Additionally, it specifies Shift_JISx0213, a Shift_JIS variant that encodes the entire JIS X 0213 character set.

Although plane 1 of JIS X 0213 largely replaces JIS X 0208, some code points differ due to revised unification criteria. Certain kanji that were unified under a single code point in JIS X 0208 receive separate assignments in plane 1 of JIS X 0213. For instance, the glyph located at row 33 cell 46 in JIS X 0208 is divided into two distinct code points in JIS X 0213 plane 1 (row 33 cell 46 and row 14 cell 41). This division complicates automatic mapping processes and limits upward compatibility—an issue acknowledged by the drafting committee.

Nevertheless, most code points correspond directly between JIS X 0208 and plane 1 of JIS X 0213. This alignment reduces practical confusion since fonts and users generally adhere to glyph examples established by JIS X 0208.

ISO/IEC 10646 and Unicode

The kanji characters defined in JIS X 0208 form a primary basis for the Han unification process utilized in ISO/IEC 10646 (UCS) and Unicode. Every kanji from JIS X 0208 is assigned a distinct code point within the UCS/Unicode Basic Multilingual Plane (BMP).

Characters other than kanji included in JIS X 0208 are also aligned with BMP code points. Nevertheless, certain implementations adopt alternative mappings for specific special characters, differing from those in UCS/Unicode, which adhere to the character names established in JIS X 0208:1997.

Reference footnotes

References

When citing, these Japanese names are presented in Western order if Romanized, while Eastern order is maintained otherwise.

Nishimura, Hirohiko [], 1978. The Kanji JIS []. Standardization Journal [], 171: 3–8.
Nomura, Masaaki [], 1984. Revision of JIS C 6226: Kanji codes for information interchange []. Standardization Journal [], 14 (3): 4–9.
Ogata, Katsuhiro [], 2006a. Things not unified in 97JIS among glyphs changed in JIS C 6226-1983 (83JIS) (accessed 29 January 2007).
Ogata, Katsuhiro [], 2006b. Glyphs within unification scope changed in JIS C 6226-1983 (83JIS) (accessed 29 January 2007).
Satō, Takayuki [], 2004. Revision of JIS X 0213 (extended Kanji sets) []. Standardization Journal [], 34 (4): 8–12.
Shibano, Kōji [], various years. Multiple articles on revisions and establishment of JIS kanji standards.
Tajima, Kazuo [], 1979. Design and handling of kanji in processing systems. Journal of Information Processing Society of Japan [], 21 (10): 753–761.
Uchida, Tomio [], 1990. Establishment of JIS X 0212 supplemental kanji codes. Standardization Journal [], 20 (11): 6–11.
Yasuoka, Kōichi [], multiple publications on Japanese character codes from early to mid-2000s.
Yasuoka, Kōichi & Motoko Yasuoka, 2006. The History of Character Codes: Europe, America, and Japan. Tokyo: Kyōritsu Shuppan (ISBN4-32012102-3).

External links

The International Register managed by IPSJ/ITSCJ.
Japanese Industrial Standards Committee database (access to the most recent standards).
Japanese Standards Association database: option to acquire current standards.
Provisions on unification in JIS X 0208 and 0213 standards
Cyber Librarian – comprehensive JIS kanji directory

Usage scope and interoperability

Lead byte

Character Set 0x21 (Row 1, Special Symbols)

Character Set 0x22 (Row 2, Special Symbols)

Character set 0x23 (row 3, digits and Roman letters)

Character set 0x24 (row 4, Hiragana)

Character set 0x25 (row 5, Katakana)

Character set 0x26 (row 6, Greek)

Character set 0x27 (Row 7, Cyrillic)

Character set 0x28 (Row 8, Box Drawing)

Extension character set 0x2D (row 13, NEC-specific characters)

Code structure

Single byte codes

Code points and code numbers

Unassigned code points

Character names

Overview

Special characters, numerals, and Latin characters

Hiragana and katakana

Kanji

Level partitioning

Arrangement

Kanji with Unverified Origins

Consolidation of Kanji Variants

Criteria for Unification to Ensure Compatibility

Encoding Formats Defined by JIS X 0208

Escape Sequences for JIS X 0202 / ISO 2022

Duplicate encodings in ASCII and JIS X 0201

History

Initial standard

Second standard

Third standard

Fourth standard

Successors

Implementations

ISO/IEC 646 IRV and ASCII

JIS X 0201

JIS X 0212

JIS X 0213

ISO/IEC 10646 and Unicode

Reference footnotes

See also

References

External links

Related profiles