Skills on AI

Active theme: Light

Arabic Language

Use when asked to write, translate into, or check text in Arabic — a right-to-left abjad script that normally omits short vowels, root-and-pattern morphology, and a major gap between Modern Standard Arabic and regional spoken dialects; official across the Middle East and North Africa with hundreds of millions of speakers.

Arabic (العربية, al-ʿArabiyyah) is a Semitic language written in the Arabic script and spoken across the Middle East and North Africa, where it is an official language in over twenty countries. It is also a major liturgical and literary language well beyond that region.

Writing system

Arabic script is written right-to-left, and letters change shape depending on their position in a word (initial, medial, final, or isolated) — a single letter can look quite different across those forms, which matters for font rendering and text layout as much as for reading. Arabic is an abjad: it primarily represents consonants, and short vowels are normally omitted entirely in ordinary text. Short vowels can optionally be marked with diacritics (tashkeel/harakat), but this is mainly done in the Quran, children's books, and materials for learners — everyday writing (news, business documents, most literature) leaves them out. This means a fluent reader infers the missing vowels from context, grammar, and vocabulary knowledge, unlike English where nearly all vowels are written; an unvocalized word can be genuinely ambiguous out of context, so surrounding text matters more than it would in English.

Modern Standard Arabic vs. regional dialects

There is a significant, practically important gap between Modern Standard Arabic (MSA, الفصحى) and the spoken vernaculars used day to day. MSA is the shared written standard used in formal writing, news media, education, and official documents across the entire Arab world, and it is what a formal document should generally be written in unless a dialect is explicitly requested. Regional spoken dialects — Egyptian, Levantine (Syria, Lebanon, Jordan, Palestine), Gulf, Maghrebi (Morocco, Algeria, Tunisia), and others — differ from MSA and from each other in vocabulary, pronunciation, and grammar enough that they are not simply casual accents of the same thing; speakers of distant dialects can struggle to understand one another in speech even though they share MSA as a common written reference. When a task doesn't specify which variety is wanted, ask or default to MSA for anything formal, and flag explicitly if dialect-specific text (social media, dialogue, marketing aimed at one country) seems to be what's actually needed.

Grammatical gender

Arabic nouns are grammatically masculine or feminine, and this gender governs agreement across verbs, adjectives, numerals, and pronouns referring to that noun. Feminine nouns are frequently but not always marked with a final tāʾ marbūṭah (ة); many feminine nouns (body parts that come in pairs, some place names) carry no overt marker at all, so gender must be learned per word rather than always inferred from spelling. Verb agreement in particular changes form for gender in ways that don't exist in English.

Root-and-pattern morphology

Most Arabic words are built from a root of typically three consonants that carries a core, abstract meaning, combined with a pattern (a template of vowels, prefixes, and infixes) that shapes that root into a specific noun, verb, or adjective. For example, the root k-t-b (relating to writing) underlies kitāb (book), kātib (writer), maktab (office/desk), and maktaba (library) — related meanings produced by different patterns applied to the same root. This is structurally different from English derivation (which mostly adds prefixes/suffixes to a fixed stem) and means dictionary lookup, spell-checking, and morphological analysis for Arabic often work root-first rather than word-first.

Common pitfalls

  • Writing informal, dialect-specific vocabulary or grammar into a document that calls for Modern Standard Arabic, or vice versa.
  • Adding full vowel diacritics to ordinary formal or business text where native readers would find them unnecessary or childish, or omitting them where a task specifically needs disambiguation (e.g. teaching materials, religious text).
  • Getting verb/adjective/pronoun gender agreement wrong, especially for feminine nouns that lack the typical tāʾ marbūṭah ending.
  • Treating dialects as mutually interchangeable, or assuming that content in one dialect (e.g. Egyptian) reads naturally to speakers of a distant one (e.g. Moroccan) without adaptation.
  • Rendering or copying text without preserving right-to-left direction and correct positional letter forms, which can garble text in systems not built for bidirectional script.
  • Assuming a root's core meaning fully predicts every derived word's actual, current usage — real usage can drift from the literal root-pattern combination.

Learn more

  • Cross Cultural Communication for navigating formality and audience expectations across the Arab world's diverse regions.
  • Locale for regional formatting conventions (dates, numbers, currency) that vary across Arabic-speaking locales.
  • Hindi Language for another language where a closely related spoken form (Urdu) diverges sharply in script and register.

View arabic-language/SKILL.md on GitHub