Arabic grammar

See Arabic pronunciation for my guide to pronunciation of Classical (Quranic) Arabic, and my system of transliteration of the script.

Arabic grammar, like that of all Semitic languages, hinges on a 'root' of three consonants (occasionally two or four). The root is not pronounced: it is not a word. It carries the basic meaning, and words are formed by filling in the consonant root with vowels and adding affixes in characteristic patterns.

The consonants k-t-b denote writing, j-l-s denotes sitting, d-r-s denotes studying, sh-r-b drinking.

The verb

As simple verbs the roots are filled and inflected for person as follows. For now I stick to the singular of the perfect.

    katabtu    I wrote
    katabta    you (masc.) wrote
    katabti    you (fem.) wrote
    kataba     he wrote
    katabat    she wrote

Note that a gender distinction is made not only in the third person as in most European languages, but also in the second person. The simplest form of the verb is the 'he' form. We name our verbs with the infinitive, 'to write'; Arabs speak of the verb kataba 'he wrote'.

Similarly jalasa 'he sat', darasa 'he studied', shariba 'he drank'. Most verbs have an a-a-a pattern in the perfect, some have a-i-a like shariba, and a few have a-u-a.

Arabic is an aspectual language. It doesn't have tenses as such, it has two aspects, perfect (or perfective) and imperfect(ive). The imperfect often covers present and future times. The vowel pattern is different and so are the person affixes:

    ?aktubu    I am (was) writing
    taktubu    you (masc.) are writing
    taktubiina you (fem.) are writing
    yaktubu    he is writing
    taktubu    she is writing

The she form and the you-masculine form are the same here.

The future is formed by prefixing sa-: sataktabu 'she will write'.

Only the singular forms were given above. There are plurals formed along similar lines (katabnaa 'we wrote', naktubu 'we are writing'), and in the second and third persons there is also a dual. Gender distinction extends into the plural.

Where the perfect has a-i-a pattern the imperfect has a-a-u, as in sharibnaa 'we drank', nashrabu 'we are drinking'.

The noun

Other parts of speech can also be formed from these consonantal roots, e.g. maktuub 'writer', maktabah 'office', kitaab 'book', majlis 'assembly, session, parliament', madrasah 'school'.

'Patterns' of vowels and affixes are pervasive in the language. Some nouns have plurals that are made simply by adding something to the singular. The 'weak' masculine ending is -uun:

    t!aalib    student
    t!aalibuun students

but the majority change the vowel pattern internally and often add affixes. Although some rules exist, in general there is no comprehensive way of predicting these. You just have to learn what plural class a singular belongs to. This is called the 'broken plural'. Some words can have both kinds of plural.

    t!aalib    student     t!ullaab   students
    kitaab     book        kutub      books
    bayt       house       buyuut     houses
    manzil     house (sic) manaazil   houses
    walad      boy         ?awlaad    boys
    jawab      answer      ?ajwibah   answers
    su?aal     question    ?as?ilah   questions

The feminine ending is -ah in the singular; it forms its plural by changing to -aat in the plural: malikah 'queen', malikaat 'queens'.

There is a dual, used for two things, so the plural is only for three or more. The dual ending is -aan, with feminine -ataan.

Case endings

Arabic has three cases, nominative, accusative, and genitive. In the singular these have the respective endings -u, -a, -i. They are also used on broken plurals. When the word is said in isolation these endings are not pronounced, so I have omitted them in the above. They are also not pronounced at the end of a phrase. This may be called the pausal form.

The weak masculine plural ending is actually -uuna, with accusative and genitive -iina. The dual ending is actually -aani with accusative/genitive -ayni. The final vowels of these are also omitted in pausal position, giving -uun, -iin, -aan, -ayn.

The feminine -ah is pausal; when it's followed by a case ending it's pronounced -at-. This alternation of consonants is ancient in Semitic, paralleled in Hebrew, and is called the ta marbuta.

The adjective

Adjectives follow nouns and agree in gender and number. However, broken plurals may be treated as feminine singular.

    kitaabu kabiir          big book
    kutubu kabiirah         big books
    madrasatu kabiirah      big school
    madrasaatu kabiiraat    big schools

The article

The definite article is a prefix al-, familiar from words like alcohol and algebra: al-kitaab 'the book', al-malikaat 'the queens'.

Agreeing adjectives also take it: al-baytu al-kabiir 'the big house'. The vowel of the article drops off after another one, so that's pronounced al-baytu-l-kabiir.

This is classical (Quranic) Arabic; in modern Arabic you often see el-.

The L of the article assimilates phonetically to a lot of consonants. These are called 'sun letters' because the L assimilates to the SH of shams 'sun', and where there is no change they are called 'moon letters' because of qamar 'moon'.

    ash-shams    the sun
    as-su?aal    the question
    at!-t!aalib  the student
    ad-dars      the lesson
    an-nahr      the river

The indefinite article is -n on the case ending: baytun kabiirun 'a big house', pronounced baytun kabiir because the second ending is pausal. It also occurs on some proper names, e.g. Muh!ammad-un.

The a of the definite article is also lost after prepositions, such as: fii 'in', li 'to', bi 'with, by', and min 'from'. Long vowels shorten, and min grows its own linking vowel. They take the genitive. So fii al-kitaabi 'in the book' is pronounced fi-l-kitaab, and 'from the sun' is mina-sh-shams.

The article is used after the demonstratives: 'this, these' is m.sg. haadhaa, f.sg. haadhihii, m.dual haadhaani, f.dual haataani, pl. haa?ulaa?i, again with shortening of the final long vowel: haadha-l-bayt 'this house'. The corresponding words for 'that' are m.sg. dhaalika, f.sg. tilka, m.dual dhaanika, f.dual taanika, pl. ?ulaa?ika.

The sentence

There is no verb 'to be' in the present. A nominal sentence is one where the adjective is not definite:
al-baytu kabiir 'the house is big'. In the past the verb kaana is used, future yakuunu.

Where a verb is used it occurs initially:
qara?a Muh!ammadun al-kitaab 'Muhammad read the book'.

Numerals

The numerals are
  1. ?ah!ad masc., ?ih!daa fem.
  2. ithnaan masc., ithnataan fem.
  3. thalaath(ah)
  4. ?arba&(ah)
  5. khams(ah)
  6. sitt(ah)
  7. sab&(ah)
  8. thamaaniy(ah)
  9. tis&(ah)
  10. &ashar(ah)
The numerals from three onward have the peculiar property (shared with other Semitic languages) that they have opposite gender to the noun they qualify.

    thalaathatu ?awlaad  three boys
    thalaathu malikaat   three queens

Derived verbs

From the simple verb a number of other verb stems can be derived by internal change and affixation. For example, kasara 'he broke' gives the intensive kassara 'he smashed' by consonant doubling. These derived stems typically have a characteristic meaning, but often they have departed a great deal in semantic latitude from the original meaning, and just have to be learnt. No verb exhibits all the possible derived stems, but using kasara 'break' as the example, the other stems are kassara, kaasara, ?aksara, takassara, takaasara, inkasara, iktasara, istaksara. All of these have their own personal inflections, imperfects, passives, and derived nouns. (Arab grammarians use the verb fa&ala 'do, make' as the paradigm.)
Disclaimer. Needless to say this is a very simplified sketch and omits anything that looks like an exception. But that said, /msg me if you notice any corrections that need to be made.