Two kinds of declension are, according to the Arabic gramma­rians, to be found in the language; the one that proper to the noun, the other that proper to the verb. The noun is declined with the tanwîn, the ḍamm, the fetḥ, and the kesr; the verb with the ḍamm, the fetḥ, and the jezm. Thus the tanwîn and the kesr are proper to the noun, and the jezm is proper to the verb. If then a noun be declined without the tanwîn and the kesr, but with the ḍamm and fetḥ only, its declension is assimilated to that of the verb. Such is the case with the class of nouns which we are considering. If the noun be fully declinable, that is with the tanwîn and kesr, it is called , because it possesses its full nominality, , in form as well as meaning. If it be not fully declinable it is called , because it does not possess all the characteristics of a noun in form, but is assimilated to the character of a verb. Now, according to all grammarians, the noun is fundamental, or primitive, and the verb is derivative, or secondary. This of the verb is two-fold; firstly, logical, , inasmuch as the noun gives a perfect sense in itself, while the verb requires a noun to be joined to it; secondly, , formal or etymological. The schools of Basra and Kufa agree as to the former cause of the verb’s secondariness, but differ as to the latter. The Basrians say that the verb is derived from the noun, that is from the maṣdar, supporting their opinion partly on logical grounds, since they say that the notion of running or eating must precede the notion he ran, or he ate, and partly on actual linguistic experience, since there are numbers of verbs which have evidently been formed from nouns expressing even material objects, as in English we say “to steam,” “to sail,” “to post a letter.” This they give as the cause of the verb’s formal secondariness. The Kufians, on the other hand, deny the derivation of the verb from the noun, and make the third person of the preterite the root. But they equally admit the formal secondariness of the verb, which they say consists in teṛkîb or composition, since any part of the verb expresses a notion compounded of the idea of the verb and a subject, while the noun expresses only a simple notion, and as the compounded is secondary to the simple, there­fore the verb is formally secondary to the noun. It is somewhat difficult to distinguish this formal secondariness from the logical secondariness. But the distinction will be perceived if we consider the logical secondariness to consist in the dependence of the abstract verb on a noun (or pronoun), for a complete meaning, and the formal secondariness to consist in the actual presence, real or implied, of a , or pronoun, in every part of the verb. It must be remembered that in each tense and person the is supposed to exist, either , apparent, or , hidden. Thus, in , and the ḍamîr is hidden; so, also, it is hidden in and , the tâ in both cases being the mark of the feminine: in some other cases it is expressed. Hence the Kufians might argue that a verb actually uttered in language inevitably contained a compounded signi­fication. By the consent of all, then, the verb has two kinds of secondariness, the one logical, the other formal or in regard of utterance.

Now the principle of the grammarians is that the nouns which have the imperfect declension have it because they correspond to the verb in these two respects; that is, they combine the two defects of logical and formal secondariness. Thus a noun to be imperfectly declined must have two defects, one of the logical and one of the formal kind; for if both be logical, or both be formal, it is not imperfectly declined. These defects are nine in number: , the quality of description or of being an adjective; , determination as a proper name; , deviation of form in a word, while its meaning remains the same, as when is turned into , composition; , the quality of being a foreign word; , the plural number; , the being feminine; , the form of the verb; and , that is, the addition of to the root, as in , drunken. Some writers add five more, making up the number to fourteen; but these last are useless refinements. These nine qualities give the noun a character of secondariness, since the adjective is secondary to the substantive, the determinate noun to the indeterminate, the deviate to the original form, the compounded to the simple, the foreign to the Arab, the plural to the singular, the feminine to the masculine, the form of the verb to the form of the noun, the augmented word to the unaugmented. Now, it has been said that the noun, to be imperfectly declined, must have two of these defects, and, moreover, one of them must be logical and the other formal. The logical defects are the first two—the quality of an adjective, and the quality of a proper name; the other defects are formal, that is, affect the form or utterance of the words themselves. Hence one of the first two defects must be united with one of the last seven to make the noun imperfectly declined. This, according to the gram­marians, is consistent with reason, inasmuch as the first, namely, the quality of description, represents the adjective, and the second, the proper name, represents the substantive; and so, between them, all nouns are represented. These two are, therefore, the “pillars” to which the others are united. Each one of the two may be combined with the deviation, or the form of the verb, or the augmentation. But the proper name alone can be combined with the foreign word, or the compounded, or the tâ of the feminine, so as to produce imperfect declinability. In this brief space only principles can be enunciated; the full meaning of these distinctions can only be understood from the explanations and examples of Arabic writers.

There are two cases, however, in which one defect is counted as two. The first concerns the defect of femininity, in the case of a word ending with the elif of the feminine. The class of words which are thus made imperfectly declinable will be found by a reference to any grammar. The reason for the double power of femininity in this case is that the elif of the feminine adheres continually to such words as if it were part of the root, which is not the case with the tâ of the feminine, which only attaches inseparably to proper names, except in some rare cases. Thus, in the case of these nouns it is held that the elif of the feminine gives them the formal defect of femininity, while the feminine essentiality which they derive from the presence of this elif, as a radical, gives them a logical femininity. They have, therefore, both formal and logical secondariness, and are, consequently, imperfectly declined.

The second concerns the defect of the quality of the plural, and affects those plurals which belong to what are called the last forms of the broken plurals, which are diverse from the form of any singular. These are plurals of the form , or , or , where the first letter is meftûḥ, and the third letter is elif, followed by a letter with teshdîd, or by two moved letters, either juxtaposed or divided by a quiescent letter. There are no singulars of these forms in the Arabic language, and the double secondariness consists in conveying the notion of plurality, which is a logical defect, and in diverging from the form of the sin­gular, which is a formal defect. The two defects, one logical and the other formal, being combined, such plurals are imper­fectly declined.

Ḥarîri does not mention in his tefsîr that the first letter of such a plural should be meftûḥ, and the same condition is omitted by other grammarians, since it always co-exists with the others mentioned; but it is logically necessary to state it, since there is a form of the singular with the first letter maḍmûm, as the word . Al Ashmûni; Bulak edition, Part II. p. 461.

Certain grammarians hold that plurals of these forms are imperfectly declined because they are really or by supposition plurals of plurals; for the repeated plurals have these forms. Thus, makes and , and makes and ; and plurals of the form and are held to correspond to these. According to this theory the logical secondariness would consist in the quality of plurality, and the formal secondariness would consist in the assumption of the form of the mukarrar or repeated plural; real in the first in­stance, and by supposition in the second. Other explanations are given of less plausibility. The opinion that one of the causes of imperfect declinability is the divergence from the forms of the singular is the teaching of Abû ‘Ali, and has been generally adopted. The answers to the objections to it are given by Al Ashmûni, Part II. p. 463. On this subject see also Alfîyeh, and the grammar of Naṣîf; as well as Anthol. Gram. Arabe, p. 95 Ar. text. The sense in which the words , and , as applied to this declension, are to be taken will be sufficiently understood from the above.