|3rd or 4th century AD to the present|
|ISO 15924||Arab, , Arabic|
The Arabic alphabet (Arabic: ?, al-abjad?yah l-?arab?yah or ?, al-?ur?f l-?arab?yah, IPA: [?al?abdadi:ja? l?arabi:ja?]), or Arabic abjad, is the Arabic script as it is codified for writing Arabic. It is written from right to left in a cursive style and includes 28 letters. Most letters have contextual letterforms.
The Arabic alphabet is considered an abjad, meaning it only uses consonants, but it is now considered an "impure abjad". As with other impure abjads, such as the Hebrew alphabet, scribes later devised means of indicating vowel sounds by separate vowel diacritics.
The Arabic alphabet is the third most used writing system in the world, after the Latin alphabet and Chinese characters. It is the most common script to write for the Arabic language, and it is also the most common script used to write for some other languages with large numbers of Muslim speakers such as Persian and Urdu, and it is one of the variant scripts used to write in other languages such as Punjabi, Malay, and Indonesian. All of these languages added some letters to represent phonemes absent from Arabic, but present in the aforementioned languages. It is also used as the base to write the Uighur Arabic alphabet, which uses many of the same letters as the Arabic alphabet, but differs in the fact that the Uighur Arabic alphabet added mandatory vowel letters, as opposed to vowel diacritics that are usually optional and/or omitted.
The basic Arabic alphabet contains 28 letters. Adaptations of the Arabic script for other languages added and removed some letters, as for Persian, Ottoman Turkish, Central Kurdish, Urdu, Sindhi, Malay, Pashto, Arwi and Arabi Malayalam, all of which have additional letters as shown below. There are no distinct upper and lower case letter forms.
Many letters look similar but are distinguished from one another by dots (?i?j?m) above or below their central part (rasm). These dots are an integral part of a letter, since they distinguish between letters that represent different sounds. For example, the Arabic letters ? (b), ? (t) and ? (th) have the same basic shape, but have one dot below, two dots above and three dots above, respectively. The letter ? (n) also has the same form in initial and medial forms, with one dot above, though it is somewhat different in isolated and final form.
Both printed and written Arabic are cursive, with most of the letters within a word directly connected to the adjacent letters.
There are two main collating sequences for the Arabic alphabet: abjad and hija.
The original ?abjad?y order (?), used for lettering, derives from the order of the Phoenician alphabet, and is therefore similar to the order of other Phoenician-derived alphabets, such as the Hebrew alphabet. In this order, letters are also used as numbers, Abjad numerals, and possess the same alphanumeric code/cipher as Hebrew gematria and Greek isopsephy.
The hij?'? () or alifb () order, used where lists of names and words are sorted, as in phonebooks, classroom lists, and dictionaries, groups letters by similarity of shape.
The ?abjad? order is not a simple historical continuation of the earlier north Semitic alphabetic order, since it has a position corresponding to the Aramaic letter samekh/semkat ?, yet no letter of the Arabic alphabet historically derives from that letter. Loss of same? was compensated for by the split of shin ? into two independent Arabic letters, ? (sh?n) and ? (s?n) which moved up to take the place of same?. The six other letters that do not correspond to any north Semitic letter are placed at the end.
This is commonly vocalized as follows:
Another vocalization is:
|The colours indicate which letters have different positions from the previous table|
This can be vocalized as:
Modern dictionaries and other reference books do not use the abjad? order to sort alphabetically; instead, the newer hij order is used wherein letters are partially grouped together by similarity of shape. The hij order is never used as numerals.
|The colours indicate which letters have different positions from the previous table|
The Arabic alphabet is always cursive and letters vary in shape depending on their position within a word. Letters can exhibit up to four distinct forms corresponding to an initial, medial (middle), final, or isolated position (IMFI). While some letters show considerable variations, others remain almost identical across all four positions. Generally, letters in the same word are linked together on both sides by short horizontal lines, but six letters (? ,? ,? ,? ,? ,?) can only be linked to their preceding letter. For example, (Ararat) has only isolated forms because each letter cannot be connected to its following one. In addition, some letter combinations are written as ligatures (special shapes), notably l?m-alif , which is the only mandatory ligature (the un-ligated combination is considered difficult to read).
name in Arabic script
|Value in Literary Arabic (IPA)||Closest English equivalent in pronunciation||Contextual forms||Isolated |
|1.||1.||1.||1.||?alif||? / ?
(also â )
including /a:/, ?[a]
(also ? )
(also ? )
(also ? )
(pharyngeal h, may be approximated as a whispered hat)
(also ?, ?, ? )
(also ? )
|20.||10.||20.||10.||r||?||r||Scottish English curd, Spanish rolled r as in perro||?|
|7.||11.||7.||11.||z?y / zayn||?||z||zebra||?|
(also ? )
(also ? )
(can be approximated with sauce, but with the throat constricted)
(also ? )
(can be approximated with dawn, but with the throat constricted)
(also ? )
(can be approximated with Taiwan, but with the throat constricted)
(also z? )
(can be approximated with xylophone, but with the throat constricted)
(similar to above, but voiced)
(also ?, ? )
(similar to caught, but pronounced further back in the mouth.)
|6.||27.||6.||27.||w?w||?||w / ? / ?||, , ?[b]||wet, pool||?|
|10.||28.||10.||28.||y||?||y / ?||, [b]||Yoshi, meet||?[d]|
(aka "glottal stop")
(used mainly in medial and final position, which is an unlinked letter)
|[g]||?alif maqrah||(end only)||
|o||Noto Sans Arabic|
|o||Noto Kufi Arabic|
|? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ?||abjad? sequence|
|o||Noto Nastaliq Urdu|
|o||Noto Naskh Arabic|
|o||Noto Sans Arabic|
|o||Noto Kufi Arabic|
|Context||Form||Value||Closest English Equivalent|
|With hamzah over
|With hamzah under
The following are not individual letters, but rather different contextual variants of some of the Arabic letters.
|Conditional forms||Name||Translit.||Phonemic Value (IPA)|
|/?a:/(aka "lengthening/ stressing 'alif")|
t / ?
|(aka "correlated t?'")
used in final position only and for denoting the feminine noun/word or to make the noun/word feminine; however, in rare irregular noun/word cases, it appears to denote the "masculine"; singular nouns: /a/,
plural nouns: ?t (a preceding letter followed by a fat?ah alif + t = ?)
|?||?alif maqrah ( ?)||? / á / ?||Two uses:|
1. The letter called ? alif maqrah or alif layyinah, pronounced /a:/ in Modern Standard Arabic. It is used only at the end of words in some special cases to denote the neuter/non-feminine aspect of the word (mainly verbs), where t?' marbah cannot be used.
|?||?||l?m + alif||laa||/la:/|
|?||?||?||?||y + m?m||?m||/i:m/|
|?||?||?||?||lam + m?m||lm||/lm/|
A more complex ligature that combines as many as seven distinct components is commonly used to represent the word All?h.
The only ligature within the primary range of Arabic script in Unicode (U+06xx) is l?m + alif. This is the only one compulsory for fonts and word-processing. Other ranges are for compatibility to older standards and contain other ligatures, which are optional.
Note: Unicode also has in its Presentation Form B FExx range a code for this ligature. If your browser and font are configured correctly for Arabic, the ligature displayed above should be identical to this one,
U+FEFB ARABIC LIGATURE LAM WITH ALEF ISOLATED FORM:
U+0640ARABIC TATWEEL + l?m + alif
Note: Unicode also has in its Presentation Form B U+FExx range a code for this ligature. If your browser and font are configured correctly for Arabic, the ligature displayed above should be identical to this one:
U+FEFCARABIC LIGATURE LAM WITH ALEF FINAL FORM
Another ligature in the Unicode Presentation Form A range U+FB50 to U+FDxx is the special code for glyph for the ligature All?h ("God"),
U+FDF2 ARABIC LIGATURE ALLAH ISOLATED FORM:
This is a work-around for the shortcomings of most text processors, which are incapable of displaying the correct vowel marks for the word All?h in Koran. Because Arabic script is used to write other texts rather than Koran only, rendering l?m + l?m + h?' as the previous ligature is considered faulty: If one of a number of fonts (Noto Naskh Arabic, mry_KacstQurn, KacstOne, DejaVu Sans, Harmattan, Scheherazade, Lateef, Iranian Sans) is installed on a computer (Iranian Sans is supported by Wikimedia web-fonts), the word will appear without diacritics.
U+0651ARABIC SHADDA +
U+0670ARABIC LETTER SUPERSCRIPT ALEF + h?'
An attempt to show them on the faulty fonts without automatically adding the gemination mark and the superscript alif, although may not display as desired on all browsers, is by adding the
U+200d (Zero width joiner) after the first or second l?m
U+200dZERO WIDTH JOINER + h?'
Gemination is the doubling of a consonant. Instead of writing the letter twice, Arabic places a W-shaped sign called shaddah, above it. Note that if a vowel occurs between the two consonants the letter will simply be written twice. The diacritic only appears where the consonant at the end of one syllable is identical to the initial consonant of the following syllable. (The generic term for such diacritical signs is ?arak?t).
|General Unicode||Name||Name in Arabic script||Transliteration|
Nunation (Arabic: tanw?n) is the addition of a final -n to a noun or adjective. The vowel before it indicates grammatical case. In written Arabic nunation is indicated by doubling the vowel diacritic at the end of the word.
Users of Arabic usually write long vowels but omit short ones, so readers must utilize their knowledge of the language in order to supply the missing vowels. However, in the education system and particularly in classes on Arabic grammar these vowels are used since they are crucial to the grammar. An Arabic sentence can have a completely different meaning by a subtle change of the vowels. This is why in an important text such as the Qur'?n the three basic vowel signs (see below) are mandated, like the ?arak?t and all the other diacritics or other types of marks, for example the cantillation signs.
In the Arabic handwriting of everyday use, in general publications, and on street signs, short vowels are typically not written. On the other hand, copies of the Qur'?n cannot be endorsed by the religious institutes that review them unless the diacritics are included. Children's books, elementary school texts, and Arabic-language grammars in general will include diacritics to some degree. These are known as "vocalized" texts.
Short vowels may be written with diacritics placed above or below the consonant that precedes them in the syllable, called ?arak?t. All Arabic vowels, long and short, follow a consonant; in Arabic, words like "Ali" or "alif", for example, start with a consonant: 'Aliyy, alif.
(fully vocalized text)
|Code||Name||Name in Arabic script||Trans.||Value||Remarks|
|? ||064E||fat·?ah||? ||a||/a/||the fathah (half-short "a") sounds properly more like the English "E" (as in "hen", "man") in comparison to the letter alif (?) which is a full-whole "A" (= like the English letter "A" as in "ark", "up")
e.g. ? ?arakah is pronounced more like "?erekeh"; shams is pronounced more like "shems"
|? ||064F||?ammah||||u||/u/||English "OO" (as "boot" but shorter)|
|0650||kasrah||? ||i||/i/||English "I" (as in "pick")|
In the fully vocalized Arabic text found in texts such as Quran, a long ? following a consonant other than a hamzah is written with a short a sign (fat?ah) on the consonant plus an ?alif after it; long ? is written as a sign for short i (kasrah) plus a y; and long ? as a sign for short u (?ammah) plus a w?w. Briefly, ?a = ?; ?y = ?; and ?w = ?. Long ? following a hamzah may be represented by an ?alif maddah or by a free hamzah followed by an ?alif (two consecutive ?alifs are never allowed in Arabic).
The table below shows vowels placed above or below a dotted circle replacing a primary consonant letter or a shaddah sign. For clarity in the table, the primary letters on the left used to mark these long vowels are shown only in their isolated form. Most consonants do connect to the left with ?alif, w?w and y written then with their medial or final form. Additionally, the letter y in the last row may connect to the letter on its left, and then will use a medial or initial form. Use the table of primary letters to look at their actual glyph and joining types.
|Unicode||Letter with diacritic||Name||Trans.||Variants||Value|
|064E 0627||?||fat?ah ?alif||?||aa||/a:/|
|064E 0649||||fat?ah ?alif maqrah||?||aa|
|0650 0649||||kasrah ?alif maqrah||y||iy||/i:/|
|064F 0648||?||?ammah w?w||?||uw/ ou||/u:/|
|0650 064A||?||kasrah y||?||iy||/i:/|
In unvocalized text (one in which the short vowels are not marked), the long vowels are represented by the vowel in question: ?alif ?aw?lah/maqrah, w?w, or y. Long vowels written in the middle of a word of unvocalized text are treated like consonants with a suk?n (see below) in a text that has full diacritics. Here also, the table shows long vowel letters only in isolated form for clarity.
Combinations and are always pronounced w? and y respectively. The exception is the suffix ? in verb endings where ?alif is silent, resulting in ? or aw.
|(implied fat?ah) ?alif||?||/a:/|
|(implied fat?ah) ?alif maqrah||? / y|
|(implied ?ammah) w?w||?||/u:/|
|(implied kasrah) y||?||/i:/|
In addition, when transliterating names and loanwords, Arabic language speakers write out most or all the vowels as long (? with ? ?alif, ? and ? with ? ya?, and ? and ? with ? w?w), meaning it approaches a true alphabet.
The diphthongs /aj/ and /aw/ are represented in vocalized text as follows:
(fully vocalized text)
An Arabic syllable can be open (ending with a vowel) or closed (ending with a consonant):
A normal text is composed only of a series of consonants plus vowel-lengthening letters; thus, the word qalb, "heart", is written qlb, and the word qalaba "he turned around", is also written qlb.
To write qalaba without this ambiguity, we could indicate that the l is followed by a short a by writing a fat?ah above it.
To write qalb, we would instead indicate that the l is followed by no vowel by marking it with a diacritic called suk?n ( ?), like this: ?.
This is one step down from full vocalization, where the vowel after the q would also be indicated by a fat?ah: .
The Qurn is traditionally written in full vocalization.
The long i sound in some editions of the Qur'?n is written with a kasrah followed by a diacritic-less y, and long u by a ?ammah followed by a bare w. In others, these y and w carry a suk?n. Outside of the Qur'?n, the latter convention is extremely rare, to the point that y with suk?n will be unambiguously read as the diphthong /aj/, and w with suk?n will be read /aw/.
For example, the letters m-y-l can be read like English meel or mail, or (theoretically) also like mayyal or mayil. But if a suk?n is added on the y then the m cannot have a suk?n (because two letters in a row cannot be suk?nated), cannot have a ?ammah (because there is never an uy sound in Arabic unless there is another vowel after the y), and cannot have a kasrah (because kasrah before suk?nated y is never found outside the Qur'?n), so it must have a fat?ah and the only possible pronunciation is /majl/ (meaning mile, or even e-mail). By the same token, m-y-t with a suk?n over the y can be mayt but not mayyit or meet, and m-w-t with a suk?n on the w can only be mawt, not moot (iw is impossible when the w closes the syllable).
Vowel marks are always written as if the i'r?b vowels were in fact pronounced, even when they must be skipped in actual pronunciation. So, when writing the name A?mad, it is optional to place a suk?n on the ?, but a suk?n is forbidden on the d, because it would carry a ?ammah if any other word followed, as in A?madu zawj? "Ahmad is my husband".
Another example: the sentence that in correct literary Arabic must be pronounced A?madu zawjun shirr?r "Ahmad is a wicked husband", is usually mispronounced (due to influence from vernacular Arabic varieties) as A?mad zawj shirr?r. Yet, for the purposes of Arabic grammar and orthography, is treated as if it were not mispronounced and as if yet another word followed it, i.e., if adding any vowel marks, they must be added as if the pronunciation were A?madu zawjun sharr?run with a tanw?n 'un' at the end. So, it is correct to add an un tanw?n sign on the final r, but actually pronouncing it would be a hypercorrection. Also, it is never correct to write a suk?n on that r, even though in actual pronunciation it is (and in correct Arabic MUST be) suk?ned.
Of course, if the correct i'r?b is a suk?n, it may be optionally written.
|General Unicode||Name||Name in Arabic script||Translit.||Phonemic Value (IPA)|
|0652||? ||suk?n||(no vowel with this consonant letter or
diphthong with this long vowel letter)
|0670||? ||alif khanjariyyah [dagger 'alif - smaller 'alif written above consonant]||?||/a:/|
The suk?n is also used for transliterating words into the Arabic script. The Persian word ? (mâsk, from the English word "mask"), for example, might be written with a suk?n above the ? to signify that there is no vowel sound between that letter and the ?.
Some letters take a traditionally different form in specific regions:
|?||||||||A traditional form to denotate the s?n ? letter, rarely used in areas influenced by Persian script and former Ottoman script.|
|?||||||||A traditional Maghrebi variant (except for Libya and Algeria) of f?' ?.|
|||||A traditional Maghrebi variant (except for Libya and Algeria) of q?f ?. Generally dotless in isolated and final positions and dotted in the initial and medial forms.|
|?||||||||An alternative version of k?f ? used especially in Maghrebi under the influence of the Ottoman script or in Gulf script under the influence of the Persian script.|
|?||||||||The traditional style to write or print the letter, and remains so in the Nile Valley region (Egypt, Sudan, South Sudan... etc.) and sometimes Maghreb; y?' ? is dotless in the isolated and final position. Visually identical to alif maqrah ?; resembling the Perso-Arabic letter ? which was also used in Ottoman Turkish.|
Some modified letters are used to represent non-native sounds of Modern Standard Arabic. These letters are used in transliterated names, loanwords and dialectal words.
|?||It is a Persian letter. Sometimes used in Arabic as well when transliterating foreign names and loanwords. Can be substituted with b?' ? and pronounced as such.|
|?||Used in loanwords and dialectal words instead of f?' ?. Not to be confused with ?.|
|?||Used in Tunisia, Algeria and Morocco.|
|?||1||It is a Persian letter. Sometimes used in Arabic as well when transliterating foreign names and loanwords and in the Gulf and Arabic dialects. The sequence t?'-sh?n is usually preferred (e.g. ? for "Chad").|
|2||Used in Egypt and can be a reduction of , where ? is pronounced .|
|3||Used in Israel, for example on road signs.|
|?||It is a Persian letter. Sometimes used in Arabic as well.|
|?||Used in Tunisia and in Algeria for loanwords and for the dialectal pronunciation of q?f ? in some words. Not to be confused with ?.|
|?||Used in Morocco.|
There are two main kinds of numerals used along with Arabic text; Western Arabic numerals and Eastern Arabic numerals. In most of present-day North Africa, the usual Western Arabic numerals are used. Like Western Arabic numerals, in Eastern Arabic numerals, the units are always right-most, and the highest value left-most. Eastern Arabic numbers are written from left to right.
In addition, the Arabic alphabet can be used to represent numbers (Abjad numerals). This usage is based on the ?abjad? order of the alphabet. ? ?alif is 1, ? b is 2, ? j?m is 3, and so on until ? y = 10, ? k?f = 20, ? l?m = 30, ..., ? r = 200, ..., ? ghayn = 1000. This is sometimes used to produce chronograms.
The Arabic alphabet can be traced back to the Nabataean alphabet used to write Nabataean. The first known text in the Arabic alphabet is a late 4th-century inscription from Jabal Ramm (50 km east of 'Aqabah) in Jordan, but the first dated one is a trilingual inscription at Zebed in Syria from 512. However, the epigraphic record is extremely sparse, with only five certainly pre-Islamic Arabic inscriptions surviving, though some others may be pre-Islamic. Later, dots were added above and below the letters to differentiate them. (The Aramaic language had fewer phonemes than the Arabic, and some originally distinct Aramaic letters had become indistinguishable in shape, so that in the early writings 15 distinct letter-shapes had to do duty for 28 sounds; cf. the similarly ambiguous Pahlavi alphabet.) The first surviving document that definitely uses these dots is also the first surviving Arabic papyrus (PERF 558), dated April 643, although they did not become obligatory until much later. Important texts were and still are frequently memorized, especially in Qur?an memorization.
Later still, vowel marks and the hamzah were introduced, beginning some time in the latter half of the 7th century, preceding the first invention of Syriac and Hebrew vocalization. Initially, this was done by a system of red dots, said to have been commissioned in the Umayyad era by Abu al-Aswad al-Du'ali a dot above = a, a dot below = i, a dot on the line = u, and doubled dots indicated nunation. However, this was cumbersome and easily confusable with the letter-distinguishing dots, so about 100 years later, the modern system was adopted. The system was finalized around 786 by al-Far?h?d?.
Although Napoleon Bonaparte generally receives credit for introducing the printing press to Egypt during his invasion of that country in 1798, and though he did indeed bring printing presses and Arabic script presses to print the French occupation's official newspaper Al-Tanbiyyah ("The Courier"), printing in the Arabic language started several centuries earlier.
In 1514, following Gutenberg's invention of the printing press in 1450, Gregorio de Gregorii, a Venetian, published an entire prayer-book in Arabic script; it was entitled Kitab Salat al-Sawa'i and was intended for eastern Christian communities.
Between 1580 and 1586, type designer Robert Granjon designed Arabic typefaces for Cardinal Ferdinando de' Medici, and the Medici press published many Christian prayer and scholarly Arabic texts in the late 16th century.
Maronite monks at the Maar Quzhayy Monastery in Mount Lebanon published the first Arabic books to use movable type in the Middle East. The monks transliterated the Arabic language using Syriac script.
A goldsmith (like Gutenberg) designed and implemented an Arabic-script movable-type printing-press in the Middle East. The Greek Orthodox monk Abd Allah Zakhir set up an Arabic printing press using movable type at the monastery of Saint John at the town of Dhour El Shuwayr in Mount Lebanon, the first homemade press in Lebanon using Arabic script. He personally cut the type molds and did the founding of the typeface. The first book came off his press in 1734; this press continued in use until 1899.
The Arabic alphabet can be encoded using several character sets, including ISO-8859-6, Windows-1256 and Unicode (see links in Infobox above), latter thanks to the "Arabic segment", entries U+0600 to U+06FF. However, none of the sets indicates the form that each character should take in context. It is left to the rendering engine to select the proper glyph to display for each character.
Each letter has a position-independent encoding in Unicode, and the rendering software can infer the correct glyph form (initial, medial, final or isolated) from its joining context. That is the current recommendation. However, for compatibility with previous standards, the initial, medial, final and isolated forms can also be encoded separately.
The basic Arabic range encodes the standard letters and diacritics but does not encode contextual forms (U+0621-U+0652 being directly based on ISO 8859-6). It also includes the most common diacritics and Arabic-Indic digits. U+06D6 to U+06ED encode Qur'anic annotation signs such as "end of ayah" and "start of rub el hizb" ?. The Arabic supplement range encodes letter variants mostly used for writing African (non-Arabic) languages. The Arabic Extended-A range encodes additional Qur'anic annotations and letter variants used for various non-Arabic languages.
The Arabic Presentation Forms-A range encodes contextual forms and ligatures of letter variants needed for Persian, Urdu, Sindhi and Central Asian languages. The Arabic Presentation Forms-B range encodes spacing forms of Arabic diacritics, and more contextual letter forms. The Arabic Mathematical Alphabetical Symbols block encodes characters used in Arabic mathematical expressions.
See also the notes of the section on modified letters.
Keyboards designed for different nations have different layouts so proficiency in one style of keyboard, such as Iraq's, does not transfer to proficiency in another, such as Saudi Arabia's. Differences can include the location of non-alphabetic characters.
All Arabic keyboards allow typing Roman characters, e.g., for the URL in a web browser. Thus, each Arabic keyboard has both Arabic and Roman characters marked on the keys. Usually the Roman characters of an Arabic keyboard conform to the QWERTY layout, but in North Africa, where French is the most common language typed using the Roman characters, the Arabic keyboards are AZERTY.
To encode a particular written form of a character, there are extra code points provided in Unicode which can be used to express the exact written form desired. The range Arabic presentation forms A (U+FB50 to U+FDFF) contain ligatures while the range Arabic presentation forms B (U+FE70 to U+FEFF) contains the positional variants. These effects are better achieved in Unicode by using the zero-width joiner and non-joiner, as these presentation forms are deprecated in Unicode, and should generally only be used within the internals of text-rendering software, when using Unicode as an intermediate form for conversion between character encodings, or for backwards compatibility with implementations that rely on the hard-coding of glyph forms.
Finally, the Unicode encoding of Arabic is in logical order, that is, the characters are entered, and stored in computer memory, in the order that they are written and pronounced without worrying about the direction in which they will be displayed on paper or on the screen. Again, it is left to the rendering engine to present the characters in the correct direction, using Unicode's bi-directional text features. In this regard, if the Arabic words on this page are written left to right, it is an indication that the Unicode rendering engine used to display them is out of date.
There are competing online tools, e.g. Yamli editor, which allow entry of Arabic letters without having Arabic support installed on a PC, and without knowledge of the layout of the Arabic keyboard.
The first software program of its kind in the world that identifies Arabic handwriting in real time was developed by researchers at Ben-Gurion University (BGU).
The prototype enables the user to write Arabic words by hand on an electronic screen, which then analyzes the text and translates it into printed Arabic letters in a thousandth of a second. The error rate is less than three percent, according to Dr. Jihad El-Sana, from BGU's department of computer sciences, who developed the system along with master's degree student Fadi Biadsy.
This article contains major sections of text from the very detailed article Arabic alphabet from the French Wikipedia, which has been partially translated into English. Further translation of that page, and its incorporation into the text here, are welcomed.