Open main menu


This template displays an inflection table for Arabic nouns. The parameters are largely the same as for the headword template {{ar-noun}}, and in general it is possible to copy such a template, change its name to {{ar-decl-noun}}, and get an appropriate declension table. (Ideally, delete extraneous stuff such as the gender and args with no equivalents below, such as |cons= and |f=.)

See also {{ar-decl-adj}}.

Base parametersEdit

The stem used in the singular, including vowels, of one of the following forms:
  1. most commonly, just STEM, i.e. an Arabic stem where the transliteration automatically determined and the inflection type inferred from the stem (this inference is very accurate and rarely makes mistakes);
  2. STEM/TR with manual transliteration supplied (use only when the automatic translit is wrong or can't be determined);
  3. STEM:TYPE with an explicitly specified inflection type (see below);
  4. STEM/TR:TYPE with both manual transliteration and explicitly specified inflection;
  5. ? to indicate an unknown singular (used more often the masculine plural of adjectives, and occasionally in nouns to specify an unknown dual);
  6. - for words without a singular (e.g. نَاس(nās, people)).
Note that manual transliteration can also be specified using separate parameters, as with headword templates like {{ar-noun}}.
|head2=, |head3=, ...
Additional singular stems.
|tr=, |tr2=, |tr3=, ...
Alternative way to specify manual transliteration for |1=, |head2=, |head3=, ..., respectively, instead of using the STEM/TR specification. Provided for compatibility with headword templates.
|pl=, |pl2=, ...
The stem(s) used in the plural, including vowels. Has the same forms as for the singular stem(s), with the addition that the value can be a raw inflection TYPE: e.g. sp to infer a strong plural from the singular and its gender; sfp, smp or awnp to construct a strong plural from the singular with a specified ending.
|pltr=, |pl2tr=, ...
Alternative way to specify manual transliteration for |pl=, |pl2=, ..., respectively, instead of using the STEM/TR specification. Provided for compatibility with headword templates. Similar parameters are available for duals.
|d=, |d2=, ...
The stem(s) used in the dual, including vowels. Same form as for plurals. Doesn't need to be present in most cases as it's inferred from the singular. Can be - for words without a dual when it would otherwise appear (which happens when the noun has a plural specified); can be ? to indicate that the dual exists but is unknown (e.g. it is currently used in بِيَانُو(biyānō, piano)).
|dtr=, |d2tr=, ...
See |pltr=, |pl2tr=, ...
Explicitly control which numbers appear. The value is a comma-separated list of one of more of sg, du, pl. By default, the singular appears unless it is specified as -; the dual appears when both the singular and plural appear; and the plural appears when at least one plural is given.
Explicitly control which states appear. The value is a comma-separated list of one of more of ind, def, con (standing for indefinite, definite, construct). By default, all states appear. (Alternatively, the value can be ind-def, for proper nouns that are syntactically definite but morphologically indefinite. This is equivalent to setting the value to def and adding |basestate=ind.) Currently, non-"appearing" states are displayed with a dash in place of the appropriate declension members instead of not appearing at all.
Override the morphological state. The value is one of ind, def or con. This can be used e.g. with proper nouns, which are syntactically definite but morphologically indefinite; but using |state=ind-def for this case is preferred.
If specified, omit the definite article that is otherwise automatically added in the definite state.
Specify a prefix to appear before the base word and before any definite article. This is typically a clitic such as وَ(wa-, and) or بِ(bi-, in, with, by), but can be a separate word such as فِي(, in), in which a space is needed after the word and the whole thing, space and all, needs to be enclosed in quotes since leading and trailing spaces are normally removed in named template arguments.
|nom_sg_ind=, |nom_sg_def=, |nom_sg_con=, ...
Override particular forms. The param name is a combination of case (nom/acc/gen/inf where inf is the informal form), number (sg/du/pl) and state (ind/def/con). The value can have semicolons or commas (or Arabic equivalents) to separate distinct stems and a vertical bar to separate alternative spellings for a given stem (typically different possible hamza seats). When overrides are specified, the page will be placed in a category indicating irregular declension such as Category:Arabic irregular nouns (or Category:Arabic irregular pronouns, etc. depending on the part of speech), unless |noirreg= is given a value.
Override the lemma (displayed at the head of the declension table).
If specified, don't place the page in an irregular-declension category when overrides are specified.
Override the part of speech (displayed at the head of the declension table).

Modifier parametersEdit

Alternative manual translit params
|modtr=,|modtr2=, ...
Alternative manual translit params
|modpltr=,|modpl2tr=, ...
Alternative manual translit params
|moddtr=,|modd2tr=, ...
Nature of modifier, high-level (ʾiḍāfa or adjectival)
Nature of modifier, low-level
Specify another word that is declined separately as a modifier. Generally only |mod= is needed, but multiple heads can be given (|modhead2=, etc., which will be distributed over the base form(s)), as well as separate plural (|modpl=, etc.) and dual (|modd=, etc.) forms. All the forms for specifying a base stem, base plural stem, etc. are available in the corresponding modifier stems, and alternative manual translit params also exist. If the modifier has a fixed genitive case in an ʾiḍāfa construction, |idafa= should be specified, e.g. |idafa=def for a definite-state singular ʾiḍāfa modifier. See the list below of possible values for this parameter, and the numerous examples below. It is also possible to manually specify the case, state and/or number of the modifier using the lower-level parameters |modstate=, |modcase=, |modnumber=, and/or |modidafa= possibly in conjunction with |basestate=, but this is rarely needed. (Underlyingly, the various values of |idafa= are implemented by setting these low-level parameters. |idafa= can be combine with the low-level parameters, with the explicitly-specified values overriding the values induced by |idafa=.)
|mod2=,|mod2head2=,|mod2head3=,...; |mod2pl=,|mod2pl2=,...; |mod2d=,|mod2d2=,...; etc. Also |idafa2=. Also |mod2state=,|mod2tr=, etc.
Specify a second modifying word. All parameters specific to the first modifying word have corresponding parameters for the second and higher modifying words. The values for |idafa2= are the same as for |idafa=, but in addition it is necessary to specify e.g. |idafa2=adj-mod for an adjectival modifier that modifies a previous ʾiḍāfa modifier.
|mod3=,|mod3head2=,|mod3head3=,...; |mod3pl=,|mod3pl2=,...; |mod3d=,|mod3d2=,...; etc. Also |idafa3=. Also |mod3state=,|mod3tr=, etc.
Specify a third modifying word, etc. up to |mod9=.

Inflection typesEdit

Regular triptote.
Regular diptote.
Noun ending in -in, such as بَانٍ(bānin). This automatically selects the triptote or diptote variant according to the number (singular vs. plural) and the form of the noun.
Noun ending in -in, forcing the "triptote" variant (with indefinite accusative in -iyan).
Noun ending in -in, forcing the "diptote" variant (with indefinite accusative in -iya). Most but not all plural nouns use this variant. (Specifically, -iya is used with nouns whose pattern is analogous to a diptote pattern, such as أَيَادٍ(ʾayādin, hands) and لَيَالٍ(layālin, nights), which correspond to the diptote pattern CaCāCiC. The "singular" -iyan is used with plural nouns such as أَيْدٍ(ʾaydin, hands), which corresponds to the triptote pattern ʾaCCuC.)
Invariable declension, such as ذِكْرَى(ḏikrā, remembrance) or دُنْيَا(dunyā, world). Typically ends in alif or alif maqṣūra, but this is not required. Such words will have sound feminine plurals in -ayāt and duals in -ayān.
Invariable declension, loanword-type. Always ends in tall alif. Such words will have sound feminine plurals in -āt and duals in -atān, as if the alif was a tāʾ marbūṭa.
Noun ending in -an, such as مُسْتَشْفًى(mustašfan) or عَصًا(ʿaṣan).
Noun with a lengthened construct ending, such as أَبٌ(ʾabun) with construct أَبُو(ʾabū).
Sound masculine plural in -ūn(a)/-īn(a).
Sound feminine plural in -āt-.
Plural ending in -awn(a), esp. of adjectives, such as أَدْنَوْنَ(ʾadnawna) plural of أَدْنَى(ʾadnā, closer, lower).
Select the appropriate plural form if known, else a strong plural. Equivalent to sfp for feminines, awnp for masculines if the singular is of type an, smp for other masculines.
Select the appropriate dual form.
Indicate that the declension is unknown. All inflections will appear as question marks.

ʾIdāfa parameter valuesEdit

sg, du, pl
An ʾidāfa modifier that can exist in multiple states and is of the specified number, e.g. indefinite-state غُرْفَة نَوْم(ḡurfat nawm, bedroom), with corresponding definite state غُرْفَة ال-نَوْم (ḡurfat an-nawm).
Same as sg.
An ʾidāfa modifier that exists in a fixed state (def, ind or ind-def) and number (sg, du or pl) .
Same as STATE-sg.
adj or adj-base
An adjectival modifier that modifies the base word. Unnecessary to specify; this is the default.
adj-mod, adj-mod2, etc.
An adjectival modifier that modifies a previous modifier word, which must be an ʾidāfa modifier.


A simple exampleEdit

Normally, it is sufficient to copy the headword declaration for a noun and make minor changes, in particular changing the template name from {{ar-noun|...}} to {{ar-decl-noun|...}} and removing the gender. For example, for the noun مَدْرَسَة(madrasa, school), the headword is declared as follows:


and a declension template can easily be declared by the above minor changes, yielding the following:


which produces the following:

Note that the plural مَدَارِس(madāris, schools) is a diptote, and this is automatically determined by its shape. Generally, diptotes have a shape different from triptotes, and the code is very good at auto-detecting the declension.

A more complicated exampleEdit

The same trick for generating declensions from headword templates usually works with more complex examples, as well, e.g. with قَفًا(qafan, nape). For this noun, the headword is declared as follows:


Removing the gender and changing the template name yields this:


which produces a correct declension table, even though there is an irregular dual and five broken plurals of three different declensions:

In general, the code is very good at figuring out the correct declension from the form of the stem, and almost always gets it right (but see the following section).

An example where the declension must be manually specifiedEdit

The code is very good at guessing the declension of a noun or adjective based on its form, but sometimes it needs help. An example is off-shape diptotes, e.g. بَبَّغَاء(babbaḡāʾ, parrot) (also بَبْغَاء(babḡāʾ) and بَبَغَاء(babaḡāʾ)), which needs an annotation like :di:


which produces

Nisba nounsEdit

Nisba nouns use special headword templates. A basic nisba noun headword template like this, for مِصْرِيّ(miṣriyy, Egyptian person):


can be converted to a declension template as follows:


This produces

Note that it is necessary to specify the plural; otherwise the declension table will have only singular. However, it is unnecessary to write out the plural; it is enough to just specify the type (in this case sp for sound plural), and the actual form will be inferred from the singular.

When the plural is overridden in the headword, conversion is even easier. An example is عَرَبِيّ(ʿarabiyy, Arab). The nisba noun headword template is:


This converts easily to a declension template:



Definite proper nounsEdit

Definite proper nouns like الإِسْلَام(al-ʾislām, Islam) exist only in the definite state. This can be specified using |state=def. The stem should be specified without the definite article, e.g.


which produces

Indefinite proper nounsEdit

Proper nouns like مِصْر(miṣr) or أَحْمَد(ʾaḥmad) or مُحَمَّد(muḥammad) are morphologically declined only in the indefinite state, but are syntactically definite, and are often diptote. This can be specified using |state=ind-def and adding a ـُ(-u) diacritic after the stem if needed, e.g.


which produces

Note how the noun is declined as if it is indefinite, but is listed as definite. This would matter, for example, in that a modifying adjective would be in the definite state. (See below for the example of كَانُون الثَانِي(kānūn aṯ-ṯānī).)

Irregular nounsEdit

Irregular nouns can be specified using overrides.

Irregular example 1Edit

One complex example is the word ذُو(ḏū, owner (of)), which exists only in the construct state and is somewhere between a noun and a demonstrative pronoun.

Its declension can be specified as follows:



Note that this declension uses a number of advanced features:

  1. The singular behaves as a long-construct noun, and is thus specified as ذ:lc, with an annotation indicating that it is a long-construct noun.
  2. The declension is restricted to be construct-state-only using |state=con.
  3. The plural أُولُو(ʾulū) is the construct state of a sound masculine plural noun, and hence it is given in the form it would occur if it existed in other states. Furthermore, the pronunciation is irregular, with the first vowel spelled long but pronounced short; as a result, the transliteration is explicitly given with a short vowel. Note that this is automatically declined as nominative أُولُو(ʾulū), accusative/genitive أُولِِي(ʾulī), with the ending changed appropriately while keeping the stem of both the Arabic and explicit transliteration. (Note, the written-out transliteration of the hamza ʾ appears backwards due to display issues. Underlyingly it is written correctly, as can be seen by editing this page to examine its source code.)

Irregular example 2Edit

An even more complex irregular noun is اِمْرُؤ(imruʾ, man), which has assimilation between stem and ending in the singular and a different stem المَرْء(al-marʾ) in the definite state (both singular and dual). The declension is specified as follows:


Here we have to override the entire singular declension and the definite state of the dual declension in order to get the right forms. We specify the stem as اِمْرَأ(imraʾ) so that the non-overridden part of the dual is correct, but that means we need to override the lemma to get the indefinite nominative singular as the lemma. We also have to explicitly specify the numbers we want to appear; otherwise only singular would appear.

Note that the dual اِمْرَآنِ(imraʾāni), with ʾalif madda, is automatically constructed correctly from the stem اِمْرَأ(imraʾ). See below.

Hamza seatsEdit

It was noted above that the dual of the stem اِمْرَأ(imraʾ, man) was correctly constructed as اِمْرَآنِ(imraʾāni), with an ʾalif madda hamza seat. In general, the code has extensive knowledge of the rules regarding the correct hamza seat in various circumstances and should essentially always get this correct. An example of this knowledge at work is the declension of the words شَيْء(šayʾ, thing) and مُبْتَدَأ(mubtadaʾ, (grammatical) subject), declared as follows:

  • {{ar-decl-noun|شَيْء|pl=أَشْيَاء}}
  • {{ar-decl-noun|مُبْتَدَأ|pl=sp}}

and appearing as follows:

In the former case, the hamza-on-the-line ء‎ automatically changes into ئ‎ when the accusative and dual endings are added. In the latter case, the hamza at the end of the stem مُبْتَدَأ(mubtadaʾ) changes in various ways depending on the ending, e.g. becoming ئ‎ before ـِين(-īn) and becoming آ‎ before ـَان(-ān). Before ـُون(-ūn), there are two possible seats, yielding either مُبْتَدَأُون(mubtadaʾūn) or مُبْتَدَؤُون(mubtadaʾūn). Both forms are displayed in the declension table, separated by the second-level separator (a slash). (The first-level separator, a semicolon, separates distinct stems, especially in the plural.)


ʾiḍāfa construction #1Edit

Some phrases appear as lemmas in Wiktionary. An example is the ʾiḍāfa construction بَيْت المُقَدَّس(bayt al-muqaddas) or بَيْت المَقْدِس(bayt al-maqdis), both meaning Jerusalem (literally house of the holy place). This is composed of بَيْت(bayt, house) in the construct state, declined according to case, plus مُقَدَّس(muqaddas) or مَقْدِس(maqdis) in the definite state, genitive case. The declension for this phrase can be expressed as follows:


which produces

ʾiḍāfa construction #2Edit

The previous example had a both head noun and modifier in the singular. However, they can disagree, in which case |idafa= should reflect the number of the modifier. An example is يَوْم الاِثْنَيْن(yawm al-iṯnayn, Monday) (literally "day of the two"), where اِثْنَيْن(iṯnayn, two) appears in the dual. The declension for this phrase can be expressed as follows:


which produces

Note that we express اِثْنَان(iṯnān, two) in the nominative case and without the definite article; the use of |idafa=def-du will cause the correct case and state to be generated. We also need to code the modifier using |modd= to express the fact that it is a dual (modd is mod + d for "dual").

ʾiḍāfa construction #3Edit

The previous examples had a head noun appearing only in the singular, and only in the definite state. However, many phrases using ʾiḍāfa can appear in all states (definite, indefinite and construct), according to the state of the modifying noun. For example, "bedroom" is غُرْفَة نَوْم(ḡurfat nawm, literally room of sleep), while "the bedroom" is غُرْفَة النَوْم (ḡurfat an-nawm) and "her bedroom" is غُرْفَة نَوْمِهَا (ḡurfat nawmihā). It also exists in the plural (غُرَف نَوْم (ḡuraf nawm)) and dual ((غُرْفَتَان نَوْم (ḡurfatān nawm)). This can be expressed using |idafa=sg, which does not constrain the state, as follows:


which produces

ʾiḍāfa construction #4Edit

An example similar to the previous one but which uses a singulative noun is دُودَة فَرَاشَة(dūdat farāša, caterpillar, literally worm of butterfly). دُودَة(dūda, worm) is a singulative noun, so we use a singulative noun template. فَرَاشَة(farāša, butterfly) is in the singular; in this case, we have to use {para|idafa|sing}} since that's the closest we can get to a singular number. (For collective nouns we would use |idafa=coll to indicate the singular.) The declension for this phrase can be expressed as follows:


which produces

As with the previous example, the value of |idafa= does not constrain the state; thus it will exist in the indefinite ﺩُﻭﺩَﺓ ﻑَﺭَﺎﺷَﺓ(dūdat farāša), the definite ﺩُﻭﺩَﺓ الﻑَﺭَﺎﺷَﺓ (dūdat al-farāša) and the construct state ﺩُﻭﺩَﺓ ﻑَﺭَﺎﺷَة أَحْمَد (dūdat farāšat ʾaḥmad).

Adjectival modifier #1Edit

Things are different for adjectival modifiers, which do not need to have their case, state or number fixed. An example is the declension for زَوْبَعَة شَدِيدَة(zawbaʿa šadīda, hurricane), literally "strong storm". Both noun and adjective need to be declined for case, state and number. This can be expressed as follows:


which produces

Note the need to express the plural of both noun and modifier. The modifier happens to be the same in the plural as the singular, because inanimate plural nouns are treated as feminine singular; but the dual will be different. In this case there is no need for |idafa= because an adjectival modifier of the base word is the default.

Adjectival modifier #2Edit

Another example is the declension for الْوِلَايَات الْمُتَّحِدَة(al-wilāyāt al-muttaḥida, United States). This noun appears only in the definite state and in the plural. This can be expressed as follows:


which produces

Note the use of a - to indicate no singular, with both the main noun and modifier expressed as plurals, so that the declension table labels them as plural rather than singular.

Adjectival construction #3Edit

A tricky example is كَانُون الثَانِي(kānūn aṯ-ṯānī, January, literally the second Kānūn). Despite appearances this is an adjectival construction, not an ʾiḍāfa construction. كَانُون(kānūn) is actually a diptote proper noun, morphologically indefinite but syntactically definite, i.e. its declension is indefinite but it is treated grammatically as definite for the purposes of modifiers. This can be expressed using the lower-level parameters to override parts of what is normally generated by |idafa= (including in its default value of |idafa=adj). The declension for this phrase can be expressed as follows:


which produces

In this case, we use |, which sets the overall state to definite, but declines the base noun as indefinite. (This is equivalent to using |state=def|basestate=ind.) This is consistent with the fact that the expression is grammatically definite and its modifier will appear with the definite article. Note also that we need to write the modifier in its indefinite form ثَانٍ(ṯānin); the appropriate definite form is automatically generated.

Other similar examples are أَمْرِيكَا الشَمَالِيَّة(ʾamrīkā š-šamāliyya, North America), كُورِيَا الجَنُوبِيَّة(kōriyā l-janūbiyya, South Korea) and فِرْجِينِيَا الغَرْبِيَّة(firjīniyā l-ḡarbiyya, West Virginia).

Prefix example #1Edit

An example where |prefix= (actually |modprefix=) comes in handy is the phrase التَوْرَاة وَالإِنْجِيل(at-tawrāh wa-l-ʾinjīl). The declension for this phrase can be expressed as follows:


which produces

Note the use of a prefix to handle the conjunction وَ(wa-, and), and the automatic implementation of elision in the transliteration.

Another prefix example is found in the documentation for {{ar-decl-numeral}}.

Prefix example #2Edit

A similar example occurs with numbers, such as تِسْرَة وَعِشْرُون(tisʿa wa-ʿišrūn, twenty-nine). The declension can be expressed as follows:


which produces

Multiple modifiers #1Edit

Things get more complicated when multiple modifiers are involved, because there are various possibilities: Each of the modifiers can independently be adjectival or ʾidāfa, and a second or further-on adjectival modifier can modify an earlier ʾidāfa modifier rather than the base.

Perhaps the simplest case is when all modifiers are adjectival. An example is غُدَّة بَصَلِيَّة إِحْلِيلِيَّة(ḡudda baṣaliyya ʾiḥlīliyya, bulbourethral gland, literally bulbous urethral gland). This can appear in multiple states and numbers, e.g. definite plural الغُدَد البَصَلِيَّة الإِحْلِيلِيَّة(al-ḡudad al-baṣaliyya l-ʾiḥlīliyya). This can be expressed as follows:


which produces

Note that the non-personal plural of the feminine adjective بَصَلِيَّة(baṣaliyya, bulbous) happens to be the same as the singular since non-personal nouns take feminine singular agreement, but the |modpl= parameter still needs to be expressed; likewise for إِحْلِيلِيَّة(ʾiḥlīliyya, urethral).

Multiple modifiers #2Edit

Another example with entirely adjectival modifiers is البَحْر الأَبْيَض المُتَوَسِّط(al-baḥr al-ʾabyaḍ al-mutawassiṭ, Mediterranean Sea), this time only in the definite singular. This can be expressed with a |state=def parameter, as follows:


which produces

Multiple modifiers #3Edit

Consider مُنَظَّمَة التَحْرِير الفِلَسْطِينِيَّة(munaẓẓamat at-taḥrīr al-filasṭīniyya, Palestine Liberation Organization). مُنَظَّمَة(munaẓẓama, organization) has an ʾidāfa modifier تَحْرِير(taḥrīr, liberation) and a second adjectival modifier فِلَسْطِينِيّ(filasṭīniyy, Palestinian) (the fact that this modifies مُنَظَّمَة(munaẓẓama) and not تَحْرِير(taḥrīr) is clear from the feminine gender of the adjective). The first, ʾidāfa modifier appears only in the definite state, which can be expressed by |idafa=def, and the fact that the second adjectival modifier modifies the base noun could be expressed by |idafa2=adj, but this is the default and so can be omitted, as follows:


which produces

Multiple modifiers #4Edit

Another example is جَنُوب القَارَّة الأَفْرِيقِيَّة(janūb al-qārra l-ʾafrīqiyya, Southern Africa, literally the South of the African continent). In this case قَارَّة(qārra, continent) is an ʾidāfa modifier of جَنُوب(janūb, south), and أَفْرِيقِيّ(ʾafrīqiyy, African) modifies قَارَّة(qārra), meaning it will agree in case, number and state (in this case, fixed to genitive singular definite). This can be expressed as follows, with a parameter |idafa2=adj-mod indicating that the second modifier is an adjectival modifier of the first modifier:


which produces

Multiple modifiers #5Edit

A similar example requiring |idafa=ind-def is جُمْهُورِيَّة غِينِيَا الاِسْتِوَائِيَّة(jumhūriyyat ḡīniyā l-istiwāʾiyya, Republic of Equatorial Guinea). In this case, غِينِيَا(ḡīniyā, Guinea) is a proper noun, grammatically definite but morphologically indefinite (similar to مِصْر(miṣr, Egypt), above), and اِسْتِوَائِيّ(istiwāʾiyy, equatorial) agrees with it, requiring the definite state. This is signalled using |idafa=ind-def:


which produces

Multiple modifiers #6Edit

An example with two ʾidāfa modifiers which can appear in various states and numbers is وَاجِهَة بَرمَجَة تَطْبِيقَات(wājihat barmajat taṭbīqāt, application programming interface), declined as follows:


which produces

Multiple modifiers #7Edit

A complex example with three modifiers is نِصْف الكُرَة الأَرْضِيَّة الشَمَالِيّ(niṣf al-kura l-ʾarḍiyya š-šamāliyy, Northern Hemisphere, literally the northern half of the earthly ball). The first modifier is an ʾidāfa; the second is adjectival, modifying the first, and the third is adjectival, modifying the base. This would be specified as follows:


which produces