Wikipedia:Manual of Style/Arabic

This page proposes a guideline regarding the use of Arabic words on the English Wikipedia.

On the English Wikipedia, Arabic is rendered into Latin script according to one of four methods in order of decreasing preference:


 * 1) Common English translation
 * 2) Common transcription
 * 3) Basic transcription
 * 4) Strict transliteration

The transliteration of Arabic used by Wikipedia is based on the ALA-LC romanization method, with a few simple changes that make it easier to read and manage in compliance with the main Manual of Style. The strict transliteration uses accents, underscores, and underdots, and is only used for etymology, usually alongside the original Arabic. All other cases of Arabic script romanization will use the same standard, but without accents, underscores, and underdots. Some exceptions to this rule may apply.

Arabic
In general, as specified on WP:English, a common English translation takes precedence over other methods to represent Arabic. This convention deals with the cases in which no common English translation is available. For the purposes of this convention, an Arabic word is defined as a name or phrase that is most commonly originally rendered in the Arabic script, and that in English is not usually translated into a common English word. These could be in any language that uses this script, such as Arabic, Persian, or Ottoman Turkish.

Examples of Arabic script rendered into Latin:
 * Abu Sayyaf (organization)
 * Jalal ad-Din Muhammad Rumi (poet)
 * Ta'iz (city)

Examples of titles not transliterated from Arabic script:
 * Yazdegerd III of Persia (ruled Persia before Arabic script became used)
 * Ahmet Necdet Sezer (ruled Turkey after the script stopped being used)
 * Egypt (does not derive from an Arabic word)
 * Algorism and Algorithm (common English words, no longer perceived as transcriptions of الخوارزمي al-Khwārizmī)

Common transcription
A word or name has a common transcription (anglicization) if a large majority of references in English use the same transcription or if a reliable source shows that an individual self-identifies with a particular transcription. Non-printable characters (including underscores) should be avoided.

Examples of references include the Oxford Dictionary, the FBI, the NY Times, CNN, the Washington Post, Al Jazeera, Encarta, Britannica, the Library of Congress, and other academic sources. Examples of self-identification include a driver's license or passport in which the individual personally chose a particular form of transcription.

Google searches can be useful in determining the most common usage, but should not be heavily relied upon. The content of large searches may not be relevant to the subject being discussed or may misrepresent the figures due to the use in languages other than English. For example, the ISO transliteration (ISO 233) of القائم is "al-Qāʾim", but the transcription "al-Qaim" receives five times as many hits. This word is used in the names of three historical Caliphs and a town in Iraq, and is also another name for the Mahdi in Shia Islam. Since Google searches do not discriminate between them, other sources must be used to determine if a common transcription exists for any particular usage. Google search counts are also biased toward syndicated news articles: a single syndicated reference may generate hundreds or thousands of hits, amplifying the weight of whatever spelling happens to be used by that one reference.

If there is no common transcription, a basic transcription is used (see below).

Examples:
 * There is no single most popular transcription for the name of the prophet of Islam. "Mohammed", "Mohammad", "Muhammad", and "Mohamed" are all commonly used. The basic transcription "Muhammad" is used.
 * The capital of Egypt is most widely known as Cairo. The basic transcription of "al-Qahira" is not used.
 * The common transcription of the leader of al-Qaeda (itself a common transcription of the strict transliteration al-Qāʿida) is "Osama bin Laden". The basic transcription of Usama ibn Ladin is not used.

Note: the Arabic word بن/ابن (son of) should be transcribed ibn unless a common transcription requires the colloquial bin.

Basic transcription
The basic transcription uses a systematic convention of rendering Arabic scripts. The basic transcription from Arabic to Roman letters is found below.

The basic transcription does not carry enough information to accurately write or pronounce the original Arabic script. For example, it does not differentiate between certain pairs of similar letters (e.g. س sīn vs. ص ṣād), or between long and short vowels. It does, however, increase the readability of the article to those not familiar with Arabic transliteration, and avoids characters that may be unreadable to browsers. This transcription method can be seen as a compromise between strict transliteration and Wikipedia conventions.

Strict transliteration
A strict transliteration is completely reversible, allowing the original writing to be faithfully restored. A strict transliteration need not be a 1:1 mapping of characters as long as there are clear rules for choosing one character over another. A source character may be mapped (1:n) into a sequence of several target characters without losing sequential reversibility.

A strict transliteration uses a system of accents, underscores, and underdots to render the original Arabic in a form that preserves all the information in the original Arabic.

ALA-LC romanization is most commonly used for this purpose; other common transliteration standards include ISO 233 and DIN 31635.

Note that several letters proposed in the strict transliteration system below do not render correctly for some widespread software configurations (e.g. ḥ, ṣ, ḍ, ṭ, ṛ, and ẓ). Using the transl template to enclose transliterations allows CSS classes to address these issues.

Article titles
Article titles should conform to WP:CRITERIA. Rules of thumb that will work in most cases: Choosing an article title that diverts from the above rules of thumb can only be done with a consensus that the alternative article title conforms better to WP:CRITERIA, and when all applicable redirects are in place.
 * 1) Use the translation or transcription that is most often used in English-language reliable sources (WP:COMMONNAME principle).
 * Example: Henna
 * 1) When there are several forms that occur often in English-language reliable sources, and for those that are used most often it is unclear which one outdoes the others in usage, choose among these the one that is closest to the basic transcription.
 * Example: Jinn (not Djinn nor Genies)
 * 1) In all other cases use the basic transcription.
 * Example: Jabir ibn Aflah
 * 1) Stay within the constraints of WP:TITLESPECIALCHARACTERS.
 * Example: Na'im ibn Musa (not Na‘im ibn Musa)
 * Example: Thābit ibn Qurra

Redirects
All frequently occurring name variants, including transcriptions and transliterations, should redirect to the article. There will often be many redirects, but this is intentional and does not represent a problem.

Lead paragraph
All articles with Arabic titles should have a lead paragraph which includes the article title, along with the original Arabic script and the strict transliteration in parentheses, preferably in the lead sentence. This is in accordance with the official Wikipedia policy at WP:ENGLISH. Many articles that are missing this information are listed at Category:Articles needing Arabic script or text. Arabic script is used in combination with the lang-ar or {{lang|ar}} template, while the strict transliteration is written using transl. A combination of the lang-ar and transl templates can also be represented by lang-ar:


 * will mark the text as Arabic. In some browsers, this may trigger a more legible font.
 * provides a mouseover note indicating that the inserted text is transliterated from Arabic. The transliteration has to be italicised manually.
 * provides a combination of a link to Arabic language, the original Arabic term, its transliteration and a literal translation.
 * Parameter 1: Arabic-script text
 * Parameter 2: strict transliteration
 * lit "literal translation"
 * link fill in  to unlink Arabic

The standard format, with, pursuant to transl, the transliteration system indicated, is given in the following examples:
 * Cairo (القاهرة) is ...
 * Al-Hakim bi-Amr Allah (الحاكم بأمر الله; 13 August 985 – 13 February 1021) was the sixth Fatimid caliph...

Some cases will require variations on this format. If the name is extremely long, the first appearance of the name is suitable to provide the strict transliteration. Likewise, if a strict transliteration appears overly repetitious, it should be in place of the page title in the lead paragraph.

Example:
 * Abū al-ʿAbbās ʿAbd Allāh ibn Muḥammad as-Saffāḥ (أبو العباس عبد الله بن محمد السفاح) was the first Abbasid caliph. Abu al-Abbas was the head of...

Main text and general usage
As with the convention for titles, common English translations should be used as much as possible. Likewise, if these are not available, one should first try a common transcription before resorting to the basic transcription. Strict transliterations in the main text should only be used out of necessity, e.g. explanations in linguistic texts or articles about transliterating.

Clash with wiki markup
Words ending with ayn or a hamza are transcribed with an apostrophe at the end. This can cause a problem if the word is at the end of an italicized or bold text. In order to prevent the final apostrophe from being interpreted as wiki markup  and ', use `.
 * Example: Karbala` for Karbala`.

Collation in alphabetical order

 * Index by family name in modern cases where there is one, otherwise by the first component in the commonly used name.
 * Example: Mu'awiya ibn Abi Sufyan should be indexed under "Muawiya".
 * For indexing of persons, the definite article "al-" and its variants (ash-, ad-, etc.) should be omitted when they form part of a modern family name.
 * Example: Hanadi Zakaria al-Hindi, Yahya El Hindi and Baba Ratan Hindi should all be indexed under "Hindi".
 * However, for organisational names, where a common transcription is established by usage, the al- or el- part is often treated as a full part of the word.
 * Example: Al-Qaeda should be indexed as "Al-Qaeda", not "Qaeda".
 * Include particles such as Abu, Abd, Abdel, Abdul, ben, bin and bint as part of the name. When found in modern surnames, such names are considered compound names and the particles are integral to the name.
 * Example: Osama bin Laden should be indexed as "Bin Laden, Osama".
 * For indexing, the apostrophe (representing hamza and ‘ayn) should be ignored, and letters with diacritics should be indexed as if they did not have their diacritics.
 * Example: Ibn Sa‘ūd should be indexed under "Saud".

Transliteration
The strict transliteration presented below is based on the ALA-LC Romanization method (1997), and standards from the United Nations Group of Experts on Geographical Names. It also includes some alternative symbols adopted in ISO 233 and DIN 31635, which are used by such sources as the Encyclopedia of Islam, and are available in the Arabic tab of the default Wikipedia editor.

The basic transcription is a simplified version.

Consonants

 * Notes from the ALA-LC specifications:
 * 1) Rule 8(a): "In initial position, whether at the beginning of a word, following a prefixed preposition or conjunction, or following the definite article, ء is not represented in romanization. When medial or final, ء is romanized." In basic transcriptions, the same applies to ‘ayn and consonantal alif.
 * 2) Rule 8(a): "In initial position, whether at the beginning of a word, following a prefixed preposition or conjunction, or following the definite article, ء is not represented in romanization. When medial or final, ء is romanized." In basic transcriptions, the same applies to ‘ayn and consonantal alif.
 * 3) Rule 7: "When the word ending in ة is in the construct state, ة is romanized t. ... When the word ending in ة is used adverbially, ة (vocalized ةً) is romanized tan."
 * 4) Rule 11(b)(2): "Final &rlm;◌ِيّ is romanized ī."

Definite article
Romanizing the Arabic definite article is usually preferred unassimilated.

Arabic has only one definite article (الـ al-). However, if it is followed by a solar letter (listed in the table right), the "L" is assimilated in pronunciation with this solar letter and the solar letter is doubled.
 * Examples:
 * تقي الدين is pronounced as and is either transliterated as Taqi al-Din (preferably) or Taqi ad-Din.

Both the non-assimilated (al-) or the assimilated (ad-) form appear in various standards of transliteration. Choose one and use it consistently throughout the article.

"Al-" and its variants (ash-, ad-, ar-, etc.) are always written in lower case, also when forming part of proper nouns, except when beginning a sentence. It is always separated from the following word (which takes the upper case when it is a proper noun) by a hyphen.
 * Examples:
 * "He was a member of al-Qaeda."
 * "Al-Qaeda has been designated as a terrorist group."

Dynastic Al
Some people, especially in the region of Arabia, when they are descended from a famous ancestor, start their last name with آل āl, a noun meaning "family" or "clan", like the dynasty Al Saud (family of Saud) or Al ash-Sheikh (family of the Sheikh). آل āl is distinct from the definite article ال al-.

Capitalization
Rules for the capitalization of English should be followed, except for the definite article, as explained above.

Names
The basic transcription of Arabic names comprises a variation on the following structure:
 * the given name (ism)
 * multiple patronymics (nasab), as appropriate, each preceded by the particle ibn (son) or bint (daughter).
 * Note: the Arabic particle بن (son of) should be transcribed ibn unless a common transcription requires the colloquial form bin (e.g. Osama bin Laden)


 * multiple descriptive nicknames (laqab) or family names (nisba), as appropriate.

If Abū is preceded by ibn, the correct grammatical format is ibn Abī, not ibn Abū.

Persian
When the Arabic script was adopted for the Persian language, there were letters pronounced in Persian which did not have a representation in the Arabic alphabet, and vice versa. The Persian alphabet adds letters to the Arabic alphabet, and changes the pronunciation of some Arabic letters. In addition, Persian does not use a definite article (al-).



Urdu
Urdu adds additional letters, and some existing letters are transliterated differently. The strict transliteration is based on the ALA-LC Romanization method for Urdu (2012). The basic transcription is the same for the additional letters, but without accents, underscores and underdots. All letters in common with Arabic should likewise follow the Arabic transcription and/or translation conventions.

Ottoman Turkish
The Ottoman Turkish language differs from the above languages in that, since 1928, words that were once written with a Persian-influenced version of the Arabic abjad have been written using the Latin alphabet. As such, there is a long established set of standards for writing the language in a basic transcription; however, in a strict transliteration, the language adheres closely to the standards for strict transliteration described above.

Guidelines for writing Ottoman Turkish words according to the basic transcription can be found at the website of the Turkish Language Association (Türk Dil Kurumu): here for the majority of words, and here for names of people.

In the following table, only those letters which differ in either their strict transliteration or their basic transcription from the Arabic-oriented table above are shown; all others are transliterated according to that table.

Definite article
In words that use the Arabic definite article ال, the article always follows the assimilation of solar letters. However, the vowel ا can be transliterated in a number of ways.


 * 1) For a definite article in initial position, the definite article is written as el- in both the basic and the strict renderings; e.g. الوهاب el-Vehhāb, الرمضان er-Ramażān.
 * 2) For a definite article in medial position, such as is found in many names of Arabic origin, the vowel in the strict transliteration can be written in a variety of ways; e.g. u’l, ü’l, i’l, ’l, etc. In such cases, the diacritic representing the hamza or ‘ayin (i.e. ’ or ‘) is always used, and the choice of vowel should follow modern Turkish orthography; e.g. عبد الله ‘Abdu’llah', عبد العزيز ‘Abdü’l-‘Azīz, بالخاصه bi’l-ḫaṣṣa.
 * 3) For a definite article in medial position in the basic transcription, ’ is not used, and the choice of vowel and spelling should follow modern Turkish orthography; e.g. عبد الله Abdullah, عبد العزيز Abdülâziz, بالخاصه bilhassa.