Arabic grammar

Arabic is a semitic language, classified as southern central Semitic together with Hebrew and Saramitan. See Arabic language for more information on the language in general.

Traditionally, rules of word formation and syntax are known collectively as an-naḥū (النحو, literally 'the orientation').

Phonology

Arabic has 28 consonantal phonemes, expressed by the 28 letters of the Arabic alphabet. In dialects, usually not all 28 phonemes are realized, so that for these speakers, some homophones are disambiguated only orthographically. Vowels are not usually represented in written language, although they may be indicated with diacritics. Arabic has three vowel phonemes; they appear as various allophones, depending on the preceding consonant. See Arabic alphabet.

Noun

State

The arabic noun can take one of three states of definiteness: definite, indefinite or construct state. The definite state is marked by the article al-. The indefinite state is marked by an ending -n (nunation). The construct state is unmarked and occurrs in the first member of a genitive construction.

the Article

The article (adātu-t-taʿrīf) al- is indeclinable and expresses definite state of a noun of any gender and number. The initial vowel is volatile in the sense that it dispappears in sandhi (hamzatu-l-waṣl), the article becoming mere -l- (although the alif is preserved in orthography in any case for clarity).

Also, the l is assimilated to a number of consonants (dentals and sibilants), so that in these cases, the article in pronunciation is expressed only by geminating the initial consonant of the noun (while in orthography, the writing alif lam is preserved, and the gemination may be expressed by putting šadda on the following letter).

The consonants causing assimilation (trivially including l) are: t, ṯ, d, ḏ, r, z, s, š, ṣ, ḍ, ṭ, ẓ, l , n. These 14 letters are called 'solar letters' (ḥuruf šamsiyyat), while the remaining 14 are called 'lunar letters' (ḥuruf qamariyyat).

Inflection

An arabic noun can take three cases: nominative, genitive and accusative, and three numbers: singular, dual and plural. Normally, nouns take the ending -u(n) in the nominative, -i(n) in the genitive and -a(n) in the accusative.

The plural of a noun is formed by a suffix in some cases (sound plurals), but frequently, the vowel structure of a word is changed to form the plural (broken plurals). There are a number of patterns of how this is done. Some words take several plurals. The phenomenon of the broken plurals is particular to Arabic and is not known from other semitic languages.

Gender

Arabic has two genders, expressed pronominal as well as verbal agreement. Agreement with numerals shows a peculiar 'polarity', c.f. the section on numerals.

The genders are usually referred to as masculine and feminine, but the situation is more complicated than that. The 'feminine' gender is also used to express 'singulatives'. The marker for the feminine gender is a -t suffix, but some nouns without this marker also take feminine agreement (e. g. umm 'mother', ard 'earth'). Already in classical arabic, -t marker was not pronounced in pausa. It is written with a special letter (ta marbuta) indicating that a t sound is to be pronounced in sandhi, but none in pausa.

Genitive Construction (Iḍāfa)

A noun may be defined more closely by a subsequent noun in the genitive. The relation is hierarchical; the first term (al-mudḍāf) governs the second term (al muḍāf ilayhi). E. g. baytu raǧulin 'house of a man'. The construction as a whole represents a nominal phrase, the state of which is inherited from the state of the second term. The first term must be in construct state, and thus cannot be marked definite or indefinite. Genitive constructions of multiple terms are possible. In this case, all but the final term are take construct state, and all but the first member take genitive case.

This construction is typical for a semitic language. In many cases the two members become a fixed coined phrase, the iḍāfa being used as the equivalent of nominal composition in indo-european languages (which does not exist in semitic). baitu-ṭ-ṭalabati thus may mean either 'house of the students' or 'the student hostel'.

Nisba

the Nisba (an-nisbatu) is a common suffix to form adjectives of relation or pertinence. The suffix is -iyy- for masculine and -iyyat- for feminine gender (in other words, it is -iyy- and is inserted before the gender marker). E. g. lubnānu 'Lebanon', lubnāniyyun 'lebanese'.

A construction noun + nisba-adjective is often equivalent to nominal composition in indo-european languages.

Pronoun

A pronomial paradigm consists of 12 forms: In singular and plural, the 2nd and 3rd persons differentiate gender, while the 1st person does not. In the dual, there is no 1st person, and only a single form for each 2nd and 3rd person. Traditionally, the pronouns are listed in order 3rd, 2nd, 1st.

Personal Pronouns

Person	Singular	Plural	Dual
3rd (m)	huwa	hum	humā
3rd (f)	hiya	hunna	-
2nd (m)	anta	antum	antumā
2nd (f)	anti	antunna	-
1st	ana	naḥnu	(n/a)

Numerals

Cardinal Numerals

Cardinal numerals from 1-10 (zero is ṣifr)

• 1 waḥidun
• 2 iṯnānu
• 3 ṯalāṯatu
• 4 arbaʿatu
• 5 ḫamsatu
• 6 sittatu
• 7 sabʿatu
• 8 ṯamāniyatu
• 9 tisʿatu
• 10 ʿašaratu

The numerals 1 and 2 are adjectives; 3-10 are diptotes (the ending -(t)u is dropped in oral usage).

Numerals 3-10 have a peculiar rule of agreement known as polarity: A feminine referrer agrees with a numeral in masculine gender and vice versa, e.g. ṯalāṯu fatayātin 'three girls'.

Numerals 11-19 are indeclinable, and they show gender agreement (not polarity). The noun counted takes accusative singular.

• 11 aḥada ʿašara
• 12 iṯnā ʿašara
• 13 ṯalāṯata ʿašara

The numerals 20-99 are followed by a noun in the accusative singular as well. There is agreement in gender with the numerals 1 and 2, and polarity for numerals 3-9.

• 20 ʿišrūna (dual of '10')
• 21 aḥadun wa ʿišrūna
• 22 iṯnāni wa ʿišrūna
• 23 ṯalāṯatu wa ʿišrūna
• 30 ṯalāṯūna
• 40 arbaʿūna

Whole hundreds, thousands etc. appear as first terms of genitive constructions, e.g. alfu laylati wa laylatu '1001 nights'.

• 100 miʾatu
• 1000 alfu

Verb

Like in many Semitic languages, the arabic word formation is based on a (usually) triconsonantal root, which is not a word in itself but contains the semantic core. The consonants k-t-b, for example, indicate 'write', q-r-ʾ indicate 'read', ʾ-k-l indicate 'eat' etc.; Words are formed by supplying the root with a vowel structure and with affixes.

Traditionally, arab grammarians have used the root f-ʿ-l 'do' as a template to discuss word formation.

Perfect

The simplest form of the verb is the perfect, third person masculine singular: kataba 'he wrote', qaraʾa 'he read'. From this the other persons and numbers form:

• katabtu I wrote
• katabta you (masc.) wrote
• katabti you (fem.) wrote
• katabat she wrote
• katabna we wrote
• katabū they (masc.) wrote
• etc., there also being dual forms

Imperfect

The imperfect has a different shape and different affixes:

• jaktubu he is writing
• taktubu she is writing; you (masc.) are writing
• taktubīna you (fem.) are writing
• ʾaktubu I am writing, etc.

Stem Formation

Derived verbs are variations on the shape of the primary kataba stem, such as kattaba, kātaba, inkataba, takattaba, etc., with senses such as intensive, reflexive, and causative, though the exact meaning varies from verb to verb and needs to be recorded in the lexicon.

Syntax

In Arabic, a word is classified as either a noun, a verb, a pronoun or a preposition. Adverbials are expressed with nominal forms, for example, to say in Arabic the sentence "the man ran slowly", one would say in Arabic a sentence whose word-for-word translation is "the man ran with slowness", or "the man ran (with) a slowness". Similarly, the Arabic word for "quick" would be treated as a noun, thus it is more properly translated not as the English word "quick", but as the English phrase "quick one".

External links

• Xerox Arabic Morphological Analysis and Generation (uses Java)