In this classic article, Moran summarizes some of the contributions of other West Semitic languages to our understanding of the history and development of Hebrew before the biblical period. He primarily concentrates on Ugaritic and the Amarna letters from Byblos, as well as West Semitic personal names.
1. Sources for pre-Biblical Hebrew
1.1. Personal Names. For the period 1900-1700 BCE, most evidence comes from personal names. Theo Bauer collected all West Semitic personal and geographic names found in Old Babylonian documents. From various Egyptian documents, we also have around 150 names from Syria and Palestine during the period.
Until its destruction by Hammurapi, Mari was ruled by a dynasty with a West Semitic dialect. There are some 500 personal and geographic names in the Mari texts which are relevant to reconstruction of early Northwest Semitic. Alalakh supplies about 100 names from the period contemporary with the First Dynasty of Babylon. There are also names from this period at Chagar Bazar.
1.2. Peripheral Akkadian. The Amarna letters (14th Century BCE) were written in Akkadian by Canaanite scribes. They include many Canaanite glosses to Akkadian words as well as forms and idioms which betray the speech of their authors. To a lesser extent, the Mari tablets also contain reflections of the local West Semitic dialect. However, scholars must be careful since not every non-Babylonian feature is necessarily Canaanite or West Semitic.
1.3. Ugaritic. The discovery of alphabetic texts at Ugarit on the northern Syrian coast has affected all areas of Biblical Studies. These texts are in a Northwest Semitic dialect, though there was (and continues to be) disagreement over whether it should be deemed a “Canaanite” language.
2.1. Consonants. Proto-Hebrew and other early Northwest Semitic dialects possessed about 25 to 27 consonants. It seems that c. 1400 is the terminus post quem for the developments which led to the 22 consonant Hebrew alphabet, following the Ugaritic ABC tablet which reflects a 27 consonant alphabet (presumably borrowed from Canaanite speakers to the south as it follows the same order as the later Phoenician and Hebrew alphabets, with three uniquely Ugaritic graphemes added to the end).
2.2. Vowels. In most Canaanite groups to the south of Ugarit, /ā/ became /ō/ during the period 1700-1375 BCE. After the Amarna period, short final vowels were generally lost, including the case endings of the noun (-u, -i, -a). Diphthongs were contracted before the Amarna period (au > ō, ai > ē) from Ugarit up to Jerusalem, which retained the diphthongs. This reflects a divergence from the northern dialect.
3. Morphology and Syntax
3.1. Remnants of Case Endings. The existence of forms such as šmmh “heavenward’ in Ugaritic suggests that the so-called he-locale is not a remnant of the earlier accusative ending -a, but represents an adverbial particle -h. Similarly, the hireq compaginis which occurs chiefly in poetry (as in Ex 15:6a, יְמִֽינְךָ֣ יְהוָ֔ה נֶאְדָּרִ֖י בַּכֹּ֑חַ “your right hand, O Lord, is fearful (ne’dārî) in strength”) was thought to be a remnant of the genitive case -i. However, it may be better seen as an archaic infinitive absolute with an -i ending (ne’dōrî), a form found in both the Jerusalem and Byblos Amarna letters.
3.2. Particles. Ugaritic, along with Amarna and Amorite, has clarified several Hebrew particles. Most important is enclitic mem. Compare Dt 33:11 מָתְנַ֧יִם קָמָ֛יו “the loins of his adversaries” to Ugaritic ṯkmm hmt “the top of the wall” where enclitic mem interrupts the construct chain. Other particles are asseverative l-, kî as “indeed”, and hm(h) as a deictic rather than conditional particle.
3.3. Pronouns. Ugaritic has demonstrated an indefinite interrogative mn, the use of אשר as a relative pronoun, and the archaic use of the demonstrative as a relative pronoun similar to the Biblical Hebrew expression ze Sînai “the one of Sinai”, paralleled by the Ugaritic d p’id “the one of Mercy” as an epithet of El.
3.4. Verb. The verb is one of the most debated areas in Hebrew grammar. There are several areas where comparative studies have helped clarify the debate.
3.4.1. Infinitive Absolute. As mentioned above, Moran argues that in Ex 15:6 the so-called hireq compaginis should probably rather be taken as an infinitive absolute which is being used in place of a finite verb, the -i being an archaic ending of the infinitive which is found in both Ugaritic and Amarna. This use of the infinitive in place of a finite verb is also paralleled by the Phoenician inscriptions where the infinitive absolute is used as a narrative tense.
The more common paronomastic use of the infinitive absolute is found also in Amarna. Combined with the -i ending, this may explain Genesis 30:8, נַפְתּוּלֵ֨י אֱלֹהִ֧ים ׀ נִפְתַּ֛לְתִּי. The first word is pointed as a noun, but a naqtûl noun pattern is otherwise unknown in Biblical Hebrew. Re-pointing the first word not as a plural noun, but an infinitive absolute, niptôlî, may make more sense, “Greatly, O God, have I contended…”
3.4.2. Prefixes of Piel and Causative. In Ugaritic the prefix is ya– for both the Piel and causative, in contrast to yu– for the Piel in Arabic and Akkadian. The Amorite names also reflect a ya– prefix, for example Ia-ki-in and Ia-ri-im. Amorite also shows evidence of the antiquity of the mē- prefix for causative participles of hollow verbs, for example Me-ki-in from Alalakh and Me-ki-nu-um from Mari, compare Hebrew mēkîn.
3.4.3. taqtulû(na) 3mp form. Alongside yaqtulû(na), a 3mp form taqtulû(na) is found in 14th century Canaanite. It is difficult to tell if the form is used in Biblical Hebrew, as most instances could also be explained as a 3fs form with plural subject taken as a collective. However, Moran suggests that it is highly probable that the form occurs in archaizing texts.
3.4.4. Indicative yaqtulu. In the Amarna texts from Byblos, two primary uses of the indicative yaqtulu can be discerned: a present-future and a past iterative. There is no reason why Byblian usage should not be comparable to the Hebrew of the time. Thus, by the 14th century this usage was already well established in the verbal system.
3.4.5. Cohortative. The Amarna letters contain a subjunctive yaqtula, corresponding to the Arabic subjunctive form, which seems to explain the origin of the Hebrew cohortative ending in -ā.
3.4.6. weqatal. In the Byblos letters, there are 33 occasions where the perfect is used with future time reference. In 24 of these cases it is preceded by the conjunction u, and is therefore comparable to the Hebrew so-called ‘waw-conversive’ with the perfect. Of the remaining cases, eight occur in the protasis of a conditional clause, and the ninth is a temporal clause. On the other hand, there are also cases where the perfect preceded by the conjunction u refers to past time and is therefore “unconverted”. Moran concludes that the data from Byblos reflects an early period of development for the Hebrew waw-conversive. The use of perfects in conditional sentences also seems to corroborate Ginsberg’s insight that the weqatal developed out of the earlier optative or precative function of the perfect.