Old Aramaic (c. 850 to c. 612 BCE)

1. Main Sources. Like Aramaic in general, Old Aramaic is generally sub-divided into western and eastern branches. The western dialect is represented by Standard Syrian/Old Western Aramaic, while the eastern dialect is represented by the Tell Fakhariya Aramaic-Akkadian bilingual and various Mesopotamian documents. Samalian is geographically western, but is a local dead-end dialect which has a mix of archaic and “Canaanite” features. Likewise, the Deir Alla text seems to be a local dialect with a mix of southern Canaanite and Aramaic features.¹

a. Standard Syrian/Old Western Aramaic (Mid 9th – end of 8th century BCE). BR-HDD, Zakkur, Sefire, Nerab, BR-RKB

b. Samalian (8th century BCE, from modern Zincirli). Hadad, Panamuwa.

c. Fakhariya (9th century BCE) – An Aramaic-Akkadian bilingual inscription written in an archaic script.

d. Mesopotamian. Several brief economic and legal texts inscribed on clay tablets, such as from Tell Shioukh Fawqani. Show large amount of Akkadian influence.

e. Deir Alla.

2. Phonology and Orthography. Old Aramaic preserves virtually the entire Proto-Semitic consonantal inventory, however the language uses the Phoenician alphabetic script which contains only 22 graphemes. This means that several characters are polyphonous: ש = /š/, /ś/, and /ṯ/; ז = /z/ and /ḏ/; צ = /ṣ/ and /ẓ/; ק = /q/ and /ḍ/; ח = /ḥ/ and /ḫ/; and ע = /ʕ/ and /ġ/. In the Tell Fakhariyah bi-lingual, /ṯ/ is also indicated by ס in the orthography.²

2.1. Old Aramaic Phonemic Inventory. The following table lists the Old Aramaic consonantal phonemic inventory along with the corresponding grapheme.³

	Emphatic	Voiced	Unvoiced
Laryngeals		[ʔ] א	[h] ה
Pharyngeals		[ʕ] ע	[ḥ] ח
Uvulars		[ġ] ע	[ḫ] ח
Velars	[q] ק	[g] ג	[k] כ
Palatovelars			[š] ש
Sibilants	[ṣ] צ	[z] ז	[s] ס
Interdentals	[ẓ] צ	[ḏ] ז	[ṯ] ש
Dentals	[ṭ] ט	[d] ד	[t] ת
Labials		[b] ב	[p] פ
Liquids	[l] ל and [r] ר
Nasals	[m] מ and [n] נ
Semivowels	[y] י and [w] ו

2.2. Vowel Letters. Along with using polyphonous consonants, Old Aramaic also developed a system to indicate long vowels with the set of consonants called glides: ו = /ū/ and diphthong /au/; י = /ī/ and diphthong /ay/ (and possibly /ē/ if the /*ay/ of the dual and plural had already monopthongized), ה = /ā/ and /ē/, and א may indicate /ā/.⁴

The western Old Aramaic inscriptions predominately marked final long vowels, but some medial vowels may have begun to be marked by י (/î/) and ו (/û/). For instance, תגלתפליסר in Bar Rakkab 1:3 (KAI 216) for Tiglath-pilesar (cf MT תִּגְלַ֣ת פִּלְאֶסֶר). The Tell Fakhariya bilingual (KAI 309) shows more regular marking of medial vowels. For instance גוגל in line 2 for Akkadian gugallu “irrigation controller” and לאלהינ “to the gods” (marking the m.p. -în ending) in line 4.⁵ Samalian, also seems to have marked medial vowels. For example קירת “cities” in Hadad 10 (KAI 214), אשור “Ashur” in Panamu 7 (KAI 215), and ביומי “in the days of” in Panamu 9, 10, 18.⁶

2.3. Final Vowels. The absolute and construct forms of feminine nouns (see below sec 3.2) may present indirect evidence that final unstressed vowels had dropped by this period (*malkatu > malkat > malkah).⁷

2.4. Dissimilation of Emphatics. Old Aramaic reflects the dissimilation of emphatics. This is similar to Geers’ Law in Akkadian, but it seems to be limited to roots with initial /q/.⁸ For instance, for קטל “to kill” we find קתל in the Samalian inscriptions and Sefire, and כטל is found in Nerab 1:11 (KAI 225).⁹ We also find כיצא “summer” for קיצא in BR RKB 1:19.

3. Morphology. There are several morphological features which distinguish the Old Aramaic dialects from other dialects within the Northwest Semitic group.¹⁰

3.1. Pronouns. The set of independent pronouns are as follows (unattested forms are marked with * and reconstructed based on later dialects):

Person	Singular	Plural
1c	אנה ˀanā	*ˀanaḥnā
2m	את ˀattā or ˀatt (< *ˀantā)	אתם ˀattūma
2f	*ˀattī	*ǎttina
3m	הא hūˀa (> hū(ˀ))	הם (and המו Zkr A 9) humu
3f	*hīˀa (> *hī(ˀ))	*hina

The suffixed pronouns are as follows:

Person	Singular	Plural
1c	י- -ī (nominal); ני- -anī (verbal)	*-anā
2m	ך- -āk	כם- -kuma
2f	*-ek, -ekī	*kina
3m	ה- -eh (-hī after vowels)	הם huma
3f	ה- -ah (-h(ā) after vowels)	הן- -hina

The interrogatives are מן man “who?” and מה mā “what?”. The relative pronoun is זי ḏī (which is later spelled די). It is not yet used for the genitive construction in the western dialects, but is so used in the Fakhariyah inscription, for example דמותא זי הדיסעי “The image of Had-yiṯ’i”.

The singular demonstratives are זנה ḏənā (m.sg.) and זא ḏāˀ (f.sg) “this”, and the plural are אל ˀillay and אלן ˀillayn “these”.¹¹ Note that the f.sg. זאת, as in Hebrew, occurs in Tell Fakhariyah line 15.

3.2. Noun Morphology. There are two distinctive features of Old Aramaic noun morphology. The feature that is probably most commonly associated with Aramaic is the suffixed definite article -ā(ˀ), written א-. This probably derives from the deictic particle hāˀ “here” as does the Phoenician and Hebrew definite article ha-.¹² However, note that Samalian shows no evidence of a definite article, nor does one appear in the Deir Alla text.

A second feature that distinguishes Aramaic is the use of final -n to mark plural nouns (compare Hebrew -m for mp and -t for fp). Again, note that Samalian differs by using -t for fp. Also, the -t ending seems to be retained for fp adjectives, מלן לחית “evil words (Sefire III:2)”.¹³

The form with the suffixed article is referred to as the determined or emphatic state and gives Aramaic the appearance of three states in the noun paradigm: absolute, construct, and emphatic.

	Absolute	Construct	Emphatic
m.sg.	מלך	מלך	מלכא
m.pl.	מלכן (-īn)	מלכי (-ay)	מלכיא (-ayyāˀ)
f.sg.	מלכה (-a(h))	מלכת (-at)	מלכתא (-atāˀ)
f.pl.	מלכן (-ān)	מלכת (-āt)	מלכתא (-ātāˀ)

An interesting feature of Samalian is that it lacks nunation on the plural nouns, instead showing case distinction by marking the nominative with ו- and genitive with י-. For example קמו·עמי·אלהו “The gods stood with me” (KAI 214 1:2), but אשא[ל·מ]ן·אלהי “I asked from the gods” (KAI 214 1:4, also 1:12).

3.3. Verb Morphology. The Old Aramaic verbal system is similar to Biblical Hebrew. The finite verbs include perfect, imperfect, and imperative. There is also a participle, infinitive construct, and a distinct infinitive absolute. A distinct jussive form exists, and the so-called waw-consecutive is attested similar to the wayyiqtol, but not the weqatal.

3.3.1. Tense-Aspect. As in Biblical Hebrew, the finite verbs center on the contrast between a suffixed perfect conjugation and a prefixed imperfect conjugation.

The paradigm of the perfect is as follows:

Person	Singular	Plural
1c	כתבת (-t, -tu/i)	כתבן (-na)
2m	כתבת (-t(a))	כתבתם (-tum)
3m	כתב	כתבו (-u)
3f	כתבת (-at)	כתבן (-u/a/in)

The paradigm of the imperfect is as follows:

Person	Singular	Plural
1c	אכתב	נכתב
2m	תכתב	תכתבן (-ūn)
3m	יכתב	יכתבן (-ūn)
3f	תכתב	תכתבן (-ūn/-ān)

3.3.2. Mood. A separate jussive form also exists which is distinguished (a) in the plural forms by lack of nunation, (b) in the final weak verbs by the ending י- (probably -ī) instead of ה- (probably ē), and (c) when taking a pronominal suffix by inserting the so-called “energic” nun in imperfect forms.¹⁴ Thus (a) the jussive יסחו in Nerab 1:9-10: שהר ושמש ונכל ונשך יסחו 10 שמך ואשרך מן חין, “May Sahar and Shamash and Nikkal and Nusk remove your name from life”. Also (b) the negative ואל תהרי “May she not conceive” (Sefire A1:21). Finally (c) the imperfect ויקתלנה is used for the future-conditional ויקתלנה וישלח ידה ויקח מן ארקי או מן מקני, “and he should kill him (my son) and stretch out his hand and take from my land or my possessions…” (Sefire A2:26).

Note that examples (a) and (c) are relevant to the use of the so-called paragogic and energic nun in Biblical Hebrew. Based on the Aramaic evidence, both of these are remnants of the nunation which originally distinguished the long imperfect from the short jussive/preterite, but which has generally dropped in Biblical Hebrew. Thus the -ennu suffix marks the long imperfect while the -ehu suffix marks the short jussive/preterite.¹⁵

In the Tell Fekhariya and Mesopotamian dialects, as well as Samalian, the 3ms jussive may take a -ל prefix instead of the normal -י, presumably related to Akkadian liprus = lū + the preterite/jussive iprus. For instance, ומנ אחר כנ 11 יבל לכננה חדס ושמימ לשמ בה “And whoever (comes) after, if it (the statue) becomes worn, should erect it anew and my name alone he should set upon it…” (Fakh 9-10). (Also note the spelling חדס for חדת and the apparent enclitic mem on שמימ “my name alone”.) This -ל may be a precursor for the later use of -ל as the normal prefix (along with -נ) for the 3ms imperfect in the eastern Aramaic dialects.¹⁶

3.3.3. Waw-consecutive. Corresponding to the existence of a short jussive form alongside the long imperfect is the existence of the so-called waw-consecutive narrative tense. For example, …ו֯אשא·ידי·אל·בעלש[מי]ן·ויענני·בעלשמי[ן “And I lifted my hands to Baalshamayn and Baalshamayn answered me…” (Zakkur A:11 [KAI 202]). In contrast, there is no waw-consecutive perfect, as only Hebrew seems to have innovated such a form.¹⁷ This suggests that there is nothing “converting” about the waw, but rather it is the preservation of an archaic preterite that gives the verbal form a past-tense meaning.

3.3.4. Infinitives. Also like Biblical Hebrew, Old Aramaic has distinct infinitive construct and absolute forms, and the infinitive absolute is used in a cognate construction similar to Hebrew. For instance, הסכר תהסכרהם בידי “Indeed you must hand them over to me (Sefire C:2)”. The infinitive construct of the derived stem seems to have a feminine ending, for example לחזיה “to see” (Sefire A1:13) from חזה with the final ה- and להמתתי ולהמתת ברי ועקרי “to kill me and to kill my son and my offspring” (Sefire C:11) with the feminine construct ת-. Further, again like Hebrew, in the basic stem the infinitive does not have a mem-prefix as in later dialects. In contrast, in the Fakhariyah inscription, the infinitive of the basic stem does have the mem-prefix, but not the feminine ending in the derived stem , ולמארכ יומוה “and to lengthen his days (line 7).”

3.3.5. Stems. Old Aramaic has three stems: peˤal (G-stem), paˤel (D-stem/factitive), and hapˤel (causative). Note that the causative prefix is still -ה rather than -א as in later dialects. The nipˤal only appears in the Deir Alla text and possibly in Samalian. Instead, passives seem to be expressed by internal vowel modification, presumably u-a as in Hebrew. The middle/reflexive t-stems are rare. Where they exist they are tG forms with a prefixed-t (יתחזה “it can be seen” [Sefire I A 28], יתאחז “it is closed” [Nerab 2:4]) rather then Gt with infixed-t as in Byblian and Ugaritic. However, a Gt is found at Tell Fakhariyah (יגתזר “it may be cut off”, line 23).¹⁸

3.4. Morphemic Miscellany. Besides the conjunction waw, in Old Aramaic פ occurs a few times, corresponding to Ugaritic p and Arabic fa as a specific marker of consecution. For example פלאכהל לאשלח י[ד בך “Then I will not be able to raise a hand against you… (Sefire B2:6)”. Also of note, the negative lā is affixed to the verb and written as ל (see לאשלח in preceding example) rather than לא as in Hebrew. In contrast, the preposition מן is never affixed to the following word nor its nun assimilated.

4. Syntax. In general, the syntax is very similar to Biblical Hebrew. However, the eastern texts from Tell Fakhariyah and Mesopotamia show the influence of Akkadian.

4.1. Word Order. The Old Aramaic texts generally follow the V(erb)S(ubject)O(bject) word order that is standard in Semitic, though allowing for fronting of an element to raise its level of saliency in the discourse.¹⁹ In the Tell Fakhariya inscription, however, the verb sometimes appears in final position as in Akkadian, for example the verb שם “to place” comes in the final position in the following sentence: קדם הדד יסב סכנ מרא חבור צלמה שמ “Before Hadad who dwells in Sikan and the lord of Habur he set his image (Line 15-16)”. This verb-final order will become distinctive of Imperial Aramaic.

4.2. Subordinate Clauses. The imperfect seems to be preferred for general subordinate clauses. This may be a remnant of the proto-West Semitic subjunctive yaqtula.²⁰ However, the construction ל + infinitive is used for telic clauses. Thus פקחו עיניכם לחציה “Open your eyes in order to see… (Sefire A1:13)”, but compare the common phrase in Sefire לאכהל לאשלח ידי “I am not able to raise my hand (against you)”, literally “I am not able, I cannot raise my hand.”

4.3. Conditional Statements. The conditional particle is הן “if…” (also הנו hinnu in Samalian). In the Old Aramaic inscriptions, conditional statements generally occur in the treaty stipulations and are real future conditions (“If s/o does A, then B will be the result…”). The protasis uses the imperfect or a verbless clause while the apodosis can use either the perfect or imperfect (or jussive). For instance והן יאתה חד מלכן ויסבני יאתה ח[ילך “If one of the kings comes and surrounds me, then your army must come (to help me)” with the imperfect, but the perfect is used with the stative שקר “to be false”, הן להן שקרתם “If this is not so, then you will be a liar”.

Notes
———————————
1. Note that traditionally Northwest Semitic has tended to be divided into Aramaic and Canaanite (Hebrew, Phoenician, etc) branches, hence the confusion over whether Deir Alla should be labeled Aramaic or Canaanite. However, following the approach of Randall Garr (Dialect Geography of Syria-Palestine, 1000-586 B.C.E. Philadelphia: University of Pennsylvania Press, 1985) it is more proper to see a continuum of dialects on which Phoenician and Aramaic are the poles. Thus it should not be strange to find languages, like the Deir Alla text, that sit somewhere in the middle of the continuum. Nor should it be surprising that Hebrew, supposedly Canaanite, shares much in common with Aramaic against Phoenician.
2. See Stephen A. Kaufman, “Aramaic” in ABD (ed D.N. Freedman. New York: Doubleday, 1992), 176, and Stanislav Segert, “Old Aramaic Phonology,” in Phonologies of Asia and Africa (ed Alan S Kaye and Peter T Daniels, Winona Lake: Eisenbrauns, 1997), 119.
3. Adapted from Segert, “Old Aramaic Phonology”, 119.
4. For א as a vowel letter see Segert, though Cross and Freedman argued based on comparison to Phoenician that א is always consonantal, or at least morphemic when marking the emphatic state, see Cross and Freedman, Early Hebrew Orthography, New Haven: American Oriental Society, 1952, and the discussion in Fitzmayer, “The Phases of the Aramaic Language,” in A Wandering Aramean: Collected Aramaic Essays. Missoula: Scholars Press, 1979: 63-65.
5. See Anderson and Freedmen “The Orthography of the Aramaic Portion of the Tell Fakhariyah Bilingual” in Text and Context: Old Testament and Semitic Studies for FC Fensham (ed W Classen. Sheffield: JSOT, 1988), 9-49.
6. See Dion, Paul-E., “The language Spoken in Ancient Sam’al,” JNES 37/2, Colloquium on Aramaic Studies (April 1978), 115-118.
7. Stephen A. Kaufman, “Aramaic” in The Semitic Languages, (ed Robert Hetzron. New York; London: Routledge, 1997), 120.
8. In fact, it was assumed by earlier scholars that the dissimilation of emphatics in Aramaic was due to the influence of Akkadian, however this view has been rejected by Stephen A. Kaufman, The Akkadian Influences on Aramaic. Assyriological studies 19. (Chicago: University of Chicago Press, 1974), 122, where he argues in part that in Akkadian the dissimilation is a function of root formation alone.
9. Greenfield, Jonas, “The Dialects of Early Aramaic,” JNES 37/2, Colloquium on Aramaic Studies (Apr, 1978): 95. See also Garr, 44-45 and 72 note 168 where he suggests that the original form was probably קתל, but the tav assimilated to the qoph as in Biblical Hebrew and Imperial Aramaic, and then the qoph dissimilated from the ṭet: *qtl > *qṭl > kṭl.
10. For a detailed comparison of the 1st millennium NWS dialects see Garr’s Dialect Geography of Syria-Palestine mentioned above.
11. For this vocalization, see Kaufman, “Aramaic” in The Semitic Languages, 122. See also discussion in Garr Dialect Geography, 84. He suggests that in all first millennium NWS dialects the plural demonstrative seems to be derived from *ˀi/ull. The final nun is an innovation in the Sefire treaty either on analogy to the masculine singular זנה or the m.pl nominal ending ן-.
12. Kaufman, “Aramaic” in The Semitic Languages , 123.
13. See Kaufman, “Aramaic” in ABD, 177.
14. Ibid., 177.
15. For some discussion see Stephen A. Kaufman, “Paragogic nun in Biblical Hebrew: Hypercorrection as a clue to a Lost Scribal Practice,” in Solving Riddles and Untying Knots: Biblical, Epigraphic, and Semitic Studies in Honor of Jonas C. Greenfield. (ed Ziony Zevit, Seymour Gitin, and Michael Sokoloff, Winona Lake: Eisenbrauns, 1995), 95-99.
16. Kaufman, “Aramaic” in ABD, 177.
17. See Mark Smith, The Origins and Development of the Waw-Consecutive: Northwest Semitic Evidence From Ugarit to Qumran. Harvard Semitic studies 39. Atlanta: Scholars Press, 1991.
18. Garr, Dialect Geography, 119.
19. Kaufman, “Aramaic” in The Semitic Languages, 127.
20. Ibid., 129.