¹ See Bauer, H. and Leander, P., Historische Grammatik der hebräischen Sprache des Alten Testaments, 1922, 501.Google Scholar

² See Wright, W., A grammar of the Arabic language, 1933, Vol. I, 150.Google Scholar

³ ibid.

⁴ It is difficult to understand what Wright means and why he calls this phenomenon nomina relativa. Surely adjectiva relativa would be more appropriate in cases where the suffix -ī is appended to an adjective.

⁵ See Wright, loc. cit.

⁶ See Marwan ibn-Janaḥ, sēfer hāriqmāh, ed. Goldberg-Kircheim, 1856, ch. xxi.

⁷ See Gesenius-Kautzsch-Cowley, , Hebrew grammar, 1910, 240Google Scholar. It is interesting to note that Ibn-Janaḥ does not mention the function of the suffix -ī as an affix for the termination of the ordinals. Perhaps the reason for the omission lies in the fact that as he was writing his grammar in Arabic and as the suffix -ī does not apply to ordinals in that language he failed to make the observation concerning Hebrew.

⁸ The book of Zechariah (I.C.C.), a.l.

⁹ E. Ben-Yehuda, Thesaurius totius Hebraitatis, s.v.

¹⁰ M. B. Schneider, tōraṯ hallāšōn b‘hiṯpatḥuṯāh, 1923, 476. See also Livne, I., diqdūq hallāšon ha‘ibrīṯ, 1958, 248.Google Scholar

¹¹ It is difficult to understand what Schneider means. Does he mean it intensifies the meaning, i.e. 'aḵzār would mean “cruel” and 'aḵzāri “very cruel”, or does he mean it emphasizes the adjectival idea?

¹² ḥofši is used as an abstract noun in Exod. 21: 2, 27. Medieval commentators felt that the use of ḥofšī in these verses was unusual. On Exod. 21:2 Ibn-Ezra comments: “Rabbi Marinus said: the yod in laḥofšī is paragogic (nōsāf); Rabbi Judah the grammarian said that the lammed is paragogic and is equivalent to ‘and when thou sendest him free from thee’ I think that both the lammed and the yod are affixes one of which would have been sufficient”. But it is possible that the word ḥofšī, though an adjective, has become a noun by the addition of the determinative he included in la.

¹³ Gen. 25: 25; 1 Sam. 16:12, 17:42.

¹⁴ “Ruddy” is synonymous with “reddish”. The English suffix -ish seems to have striking semantic similarities with the Hebrew suffix -ī. It was first used to form adjectives of nationality from geographical names, e.g. English, Irish (compare Hebrew miṣrī, ‘iḇrī). Then it came to denote “belonging to, having the nature or the quality of” as in childish (cf. '^ewīlī, 'aḵzārī). Then the suffix developed to denote a diminutive, especially with adjectives of colour, e.g. whitish etc. (see OED under the suffix -ish). See also E. Ullendorff's remark on the word 'admōnī in Vetus Testamentum, 1956, 191–2.

¹⁵ A detailed analysis of the arguments can be found in the writer's University of London doctoral thesis, “The adjective in Hebrew: an analysis of its morphology and function”, 1969, pp. 18–64.

¹⁶ See Levita, E., sēfer pirkē Eliāhū, 1788, 26Google Scholar; who lists z^ekūkīṯ as a šēm tō'ar or šēm hammiṯgabbēr. Ibn-Janaḥ, op. cit., ch. xi, calls this phenomenon tō'ar goḇer (a translation of the Arabic term ṣifa ghāliba), which according to Lane, Arabic-English Lexicon, root wṣf, is “an epithet in which the substantive character predominates”. See also EI s.v. ṣifa.

¹⁷ See Lumsden, M., A grammar of the Arabic language, 1813, 266Google Scholar. He appears to be the first grammarian to have used the term to translate the Arabic ṣifa and to have opposed it to the adjective.

¹⁸ See Ewald, H., Ausführliches Lehrbuch der hebräischen Sprache, 1863, 425Google Scholar. The tendency to append the adjectival ending -ī to an already existing adjective has continued throughout the history of the Hebrew language. Thus, we find ḥīṣōnī, qīṣōnī, tīḵōnī, although ḥīṣōn, qīṣōn, tīḵōn are already adjectives. The adjective r^eṣīnī developed from the Arabic razīn, but the adjective in Hebrew is r^eṣīnī, not raṣīn. Mūḥāṡ is the hof'al participle of ḥūš, but the form preferred for the adjective is mūḥāšī. The phenomenon of affixing -ī to existing adjectives can also be seen in foreign adjectives, e.g. abstracti (abstract), agresiḇi (aggressive). The suffix -ī is very commonly appended to patterns qatlān and q^etaltān. H. Garland, who criticized this development (l^ešōnēnū, II, 229), failed to realize that the phenomenon had begun already in the Bible in order to give to the word an adjectival nature.

¹⁹ A comprehensive list can be found in the writer's doctoral thesis.

²⁰ Ezek. 1:4, 7, 14, 16, 26, 27; 3: 9. See also Gen. 3: 5; Lev. 26: 19; Nahum 2: 5; Joel 2: 4; Ps. 71: 7; 147: 16; Job 10: 9; Lam. 4: 8; Cant. 5: 15; Dan. 10: 6.

²¹ For the purpose of this article Rabbinic Hebrew consists of the Mishna, the Tosefta, the Talmudim, and the Midrasim.

²² A list of Rabbinic gentilitial adjectives can be found in the writer's doctoral thesis, p. 124.

²³ We include in this period the literature that is commonly known as scientific, e.g. philosophy, medicine, etc., and we exclude the literature that is written in Rabbinic style.

²⁴ See e.g. the translation of Ibn-Sinna's qānūn.

²⁵ For the terminology see Sweet, H., A new English grammar, 1891, pt. 1, 67Google Scholar. Almost all Biblical adjectives and most Rabbinic adjectives express a pure qualification. Biblical exceptions belong mainly to the participial adjectives, e.g. nāḥūš, which is used as a concrete adjective in 2 Sam. 22: 35, but also as an attribute-adjective in Job 6: 12. Ṣarūa‘ may be considered a concrete adjective derived from ṣara‘aṯ. Similarly m^eṣōra‘, maqrīn, sa‘īr, and mūfāz. But these are few. To express concrete adjectives the Bible has recourse to the construct state. Sometimes a pure adjective is also expressed by a construct state, especially qōdeš, which very frequently serves as a periphrasis for the adjective qādōš, and ṣedeq, whose corresponding adjective ṣaddiq is always confined to beings and not to objects. Most of the Rabbinic adjectives express pure qualification. Some can be considered concrete adjectives, e.g. māliaḥ, ‘āfīṣ, qāmīaḥ, zāfūṯ, ṭ^elūlāh, mālūaḥ, etc., but they can be used as attribute-adjectives as well. It is interesting to note that the phenomenon of the construct state expressing concrete adjectives is frequent in English but rare in the Romance languages. Hebrew beiṯ'eben is equivalent to the English “stone house”. In French we would say une maison de pierre and in Spanish una casa de piedra. In modern Hebrew the construct state is sometimes used to express qualification, e.g. ‘ōr mešī (“a velvet skin”). In order to differentiate between the construct state expressing qualification and the construct state expressing composition the language sometimes has recourse to the particle šel. Thus, ‘ōr šel mešī is illogical because “skin” can never be made of silk. Similarly in English we would say “a velvet(y) skin” and not “a skin of velvet”. See J. Zajicek, “Essais de stylistique comparée et raisonnée, anglais-français et Langues germaniques modernes”, Thèse, Université de Paris, 1951; who calls the first “genitive zero” and the second “genitive”.

²⁶ s^efāṯēnū, II, 1923, 54.Google Scholar

²⁷ l^ešōnēnū la‘am, IX, 1958, 190.Google Scholar

²⁸ l^ešōnēnū la‘am, X, 1959, 23.Google Scholar

²⁹ yad hallāšōn, 1965, 603.

³⁰ l^ešōnēnū la‘am, II, 1951, 16Google Scholar. See also Etan, Y., hattōren, 1922, 64Google Scholar, and Perez, Y., ‘ibrīṯ kah^alaḵāh, 1961, 158Google Scholar, where slightly different criteria are used.

³¹ It should be noted that the particle šel is not always equivalent to the construct state, kos yayin means a cup of wine, a cup containing wine; kos šel yayin means a glass used to serve wine, which might be empty. Compare “a wine-glass” and “a glass of wine”.

³² op. cit., 602.

³³ op. cit., 476.

³⁴ op. cit.

³⁵ In the Babli and the Y^erušalmī ḥēfānī is found. See b. M^eg. 24b; j. B^er. 4d.

³⁶ In the same manner s^emā'lī is derived from s^emō'l, where we would have expected *s^emō'li but have s^emā'lī on the analogy of y^emānī. The latter is derived from yāmīn and should have been *y^emīnī, but the form y^emānī is preferred probably in order to differentiate between y^emīnī “Benjamite” and y^emānī “right (side)”. Similarly in the Mishna (Pe. I:1), Ben Azzai differentiates between š^elīšī “third” and š^elāšī “three years old”; so it is too with r^eḇī‘ī and r^eḇā’ī. Ibn-Janaḥ, Riqmāh, ch. x, uses ḥ^amāšī to designate a quinqueliteral root in distinction to h^amišši “fifth”.

³⁷ Biblical and modern Hebrew would use yāḥid for “unique”.

³⁸ Modern Hebrew uses 'āfōr of the pattern qātōl which produces adjectives of colour in the Bible and in the post-Biblical period.

³⁸ Modern Hebrew uses m^eguyyāḏ for “tendinous”.

⁴⁰ s.v. Modern Hebrew uses y^eraqraq for the diminutive.

⁴¹ see Judg. 7: 3 and Isa. 66: 5.

⁴² Hebrew seems to dislike the ending אד. Thus we have qāruy replacing qārū'. It is a well-known fact that the א״ל and the ה״ל verbs are often interchanged. In the Bible we find tālū' and tālūy; in the Mishna we find qārīṯī and qārā'ṯī. E. Ullendorff, “Modern Hebrew as a subject of linguistic investigation”, J. Sem. Stud., 1957, points out that modern Hebrew is inconsistent in this respect as it has nāsūy (ה״ל) and n^esū'īm (א״ל). It seems that there is a difference in meaning between nāsū' and nāsūy, the former meaning “predicate” and the latter “married”.

⁴³ 'aḇqī has fallen into disuse; for “dust-like” modern Hebrew uses the Biblical k^e'āḇāq.