Akkadian

Overview

Akkadian is the oldest Semitic language attested in writing and the principal language of ancient Mesopotamia for nearly two millennia. A member of the East Semitic branch of the Afro-Asiatic family — within which it forms, together with Eblaite, the East Semitic subgroup — it is distinguished from the West Semitic languages (Hebrew, Arabic, Aramaic) notably by its SOV (Subject-Object-Verb) word order. This order, atypical for a Semitic language (generally VSO), is usually attributed to prolonged contact with Sumerian (itself SOV), a hypothesis that is dominant though not verifiable in detail. It takes its name from the city of Akkad (Agade), capital of the empire founded by Sargon (Šarru-kīn) around 2334 BC (Middle Chronology). From the end of the 3rd millennium onward, Akkadian split into two major regional dialects: Babylonian in the south and Assyrian in the north, each passing through several phases (Old, Middle, Neo-, Late). Written by means of the cuneiform script borrowed from Sumerian, it left behind an immense corpus: letters, contracts, law codes (Code of Hammurabi), religious texts, epics (Gilgameš, Enūma eliš, Atra-ḫasīs), scientific texts (astronomy, medicine, mathematics), and diplomatic texts (the El-Amarna letters). It is a fully deciphered language, equipped with reference grammars, exhaustive dictionaries (CAD, AHw), and freely accessible digital corpora.

History of the people

The speakers of Akkadian were the Semitic populations of Mesopotamia (present-day Iraq), in millennia-long interaction with the non-Semitic Sumerians. AKKAD: around 2334 BC (Middle Chronology), Sargon (Šarru-kīn) unified the cities and founded the empire of Akkad, often presented as the first great territorial empire in history, reaching its apogee under his grandson Naram-Sîn; the empire collapsed around 2154, notably in the context of the Gutian incursions. There followed the Sumerian renaissance of the Third Dynasty of Ur (Ur III). At the beginning of the 2nd millennium, Amorite kingdoms emerged, among them BABYLON, raised to hegemony by Hammurabi (ca. 1792–1750 BC, Middle Chronology, author of the famous "code"). The Old Babylonian kingdom fell in 1595 (Hittite raid), after which the Kassite dynasty governed Babylon for several centuries. In the north ASSYRIA developed, first as a merchant city-state (Aššur, with the Anatolian trading colonies of Kaneš in the 19th century), then as a great imperial power: the Middle Assyrian empire, and above all the Neo-Assyrian empire (ca. 911–609 BC), the largest of its time under Tiglath-Pileser III, Sargon II, Sennacherib, Esarhaddon, and Ashurbanipal (whose library at Nineveh preserved thousands of tablets). Nineveh fell in 612 BC before the Medes and Babylonians. The Neo-Babylonian empire (Nebuchadnezzar II) then held sway until the Persian conquest by Cyrus in 539 BC. Under Persian, then Macedonian, then Parthian rule, cuneiform culture declined but persisted in the temples until the beginning of our era.

History of the language

Akkadian appears in onomastics and glosses from around 2600–2500 BC and becomes a full written language under the empire of Akkad (ca. 2334–2154 BC, Middle Chronology; the period known as "Old Akkadian"). After a phase of Sumerian renaissance (Third Dynasty of Ur), Akkadian split into two major dialectal branches: BABYLONIAN (south) and ASSYRIAN (north), each subdivided into periods. Babylonian: Old Babylonian (ca. 2000–1595, the "classical age" of the language, that of the Code of Hammurabi), Middle Babylonian (ca. 1595–1000), Neo-Babylonian (ca. 1000–600), and Late Babylonian (down to the last texts). Assyrian: Old Assyrian (ca. 2000–1750, the commercial archives of Kaneš/Kültepe in Anatolia), Middle Assyrian (ca. 1500–1000), Neo-Assyrian (ca. 1000–600, the language of the empire down to Ashurbanipal). A prestigious, archaizing literary register, STANDARD BABYLONIAN, served as the common literary language from the 2nd to the 1st millennium for epic and scholarship. Babylonian was also the international diplomatic language of the Near East in the Late Bronze Age (the El-Amarna correspondence, 14th century). EXTINCTION: from the 8th–7th century BC onward, Aramaic (a West Semitic language with an alphabetic script) progressively supplanted Akkadian as a spoken and administrative language. After the fall of Babylon (539 BC) and then under the Achaemenids and Seleucids, Akkadian survived mainly as a learned, liturgical, and scientific (astronomical) language, written by an elite of scribes. The latest dated cuneiform texts are astronomical texts; the last datable cuneiform tablet identified is an astronomical almanac from Uruk (W22340a) dated to the Seleucid year 385 / Arsacid year 326, that is, around AD 79/80. Cuneiform production then died out completely, and the language fell into oblivion until its decipherment in the 19th century.

The script

Akkadian is written in cuneiform (from Latin cuneus, "wedge"), a system invented by the Sumerians (the Uruk region) around 3300 BC and adapted to Akkadian from the middle of the 3rd millennium onward. The signs are formed from wedge impressions made by a reed stylus with a triangular cross-section pressed into fresh clay and, for monumental inscriptions, carved on stone. The script is logo-syllabic and operates according to four combined principles: (1) SYLLABIC signs noting open syllables (V: a, i, u, e; CV: ba, ki, tu; VC: ab, il, ut) or closed ones (CVC: kan, lum, bir); (2) LOGOGRAMS (or Sumerograms), signs borrowed from Sumerian that note an entire Akkadian word — conventionally transcribed in CAPITALS (e.g. LUGAL = šarru "king," É = bītu "house," DINGIR = ilu "god"); (3) DETERMINATIVES, unpronounced signs classifying the following or preceding word by semantic category (e.g. DINGIR/ᵈ before a divine name, m or disz before a man's name, KI/ki after a place name, GIŠ before wooden objects); (4) PHONETIC COMPLEMENTS specifying the reading or inflection of a logogram. The direction of reading is left to right, in horizontal lines (the archaic signs, at first arranged in vertical columns, rotated 90° over the course of the 2nd millennium). Two major features complicate reading: POLYPHONY (a single sign has several values — e.g. the sign UD/UTU has among others the syllabic values ud, tu, tam, pir, laḫ, ḫiš… and the logograms UD "day," U₄, BABBAR "white," as well as UTU/Šamaš "sun/sun-god") and HOMOPHONY (several distinct signs note the same syllable, distinguished in transliteration by indices: u, ú, ù, u₄…; šu, šú, šù…). Akkadian does not systematically note long vowels or the gemination of consonants, which are often left to interpretation or marked by "full" spelling (doubling of the sign). The number of signs actually in use at a given period is roughly 200 to 400 (an order of magnitude varying with period and text genre).

Sign system

The Akkadian cuneiform repertoire is catalogued in standardized sign lists. The current standard reference is the "MesZL" (R. Borger, Mesopotamisches Zeichenlexikon, AOAT 305, 1st ed. 2003/2004, 2nd ed. 2010), which assigns a number to each sign; older lists such as the "ŠL" (A. Deimel, Šumerisches Lexikon) are also used and, for bibliography, the HKL (Borger, Handbuch der Keilschriftliteratur). Each sign possesses: (a) one or more syllabic values (phonetic readings), (b) one or more logographic values (a Sumerian/Akkadian word), (c) sometimes a determinative function. Assyriological transcription convention: Akkadian syllabic values are in lowercase italics (šar-ru-um); logograms/Sumerograms in upright CAPITALS (LUGAL); determinatives in superscript (ᵈEnlil, Bābiliᵏⁱ, ᵐSîn-aḫḫē-erība). The numeric indices (accents and figures: u, ú, ù, u₄, u₅…) distinguish homophones and note NO difference in pronunciation — they refer to graphically distinct signs, ordered by conventional usage (largely based on frequency). Examples of frequent and versatile signs: the sign AN = an, ilu ("god," logogram DINGIR), divine determinative ᵈ; the sign A = a, mû ("water," logogram), and serving to note aplu/māru ("son," via DUMU or A); the sign É = bītu ("house"); the sign KUR = mātu ("land"), šadû ("mountain"), with the syllabic values kur, mat, šad, lat, nat… Deciphering a sign therefore requires choosing, among its values, the one that yields a coherent Akkadian word in the grammatical context.

Phonology

The reconstructed Akkadian phonological system comprises: CONSONANTS — voiceless stops p, t, k; voiced b, d, g; "emphatics" ṣ, ṭ, q (probably glottalized/ejective at an early date, the precise realization being debated); fricatives/sibilants s, z, š, and ḫ (a voiceless velar or uvular fricative); sonorants m, n, l, r; semivowels w, y; and a consonant ʾ (glottal stop / vowel bearer). A major East Semitic feature: the Proto-Semitic laryngeals and pharyngeals (*ʾ, *h, *ḥ, *ʿ, *ġ) have largely DISAPPEARED in Akkadian, often with compensatory lengthening or coloring of the adjacent vowel (in particular *ʿ and *ḥ color a > e). VOWELS — four timbres a, i, u, e (the secondary e derives mainly from coloring by former laryngeals), each short or long. A distinction is made between "naturally" long vowels (arising from contraction, often noted with a circumflex â, ê, î, û) and "ordinary" long vowels (macron ā, ē, ī, ū) — a distinction that is largely conventional in modern editions. Stress (not noted in the writing) is the subject of reconstructed rules, falling in principle on the last heavy syllable. Notable phenomena: vowel harmony in Babylonian (harmony of the preformative vowel a); frequent consonantal assimilations (nC > CC: *inaddin > iddin); syncope of a short vowel in an unstressed open syllable (an important rule for morphology: *parisum > parsum). The phonetic restitution remains partly conventional: the syllabic writing gives a good picture of the consonants and vowel timbres but notes vowel quantity and gemination poorly.

Grammar — essentials

Semitic morphology based on consonantal roots (most often triconsonantal: PRS, ŠKN, DIN), whose vocalic patterns and affixes carry grammatical meaning. NOUN: two genders (masculine; feminine marked in -at/-t), three numbers (singular, residual dual, plural), and a three-CASE declension in the singular: nominative -u(m), genitive -i(m), accusative -a(m) — the final -m (mimation) generally drops after the Old Babylonian period. Masculine plural: nom.-acc. -ū, gen. -ī; feminine plural -ātu(m)/-āti(m). "Construct" state (status constructus) for the genitival relation (bīt šarrim "the house of the king"), and absolute state. VERB: the core of the system, organized into STEMS (Stämme) — G (Grundstamm, base), D (doubling of the 2nd radical, factitive/intensive), Š (prefix š-, causative), N (prefix n-, passive/medio-passive), plus the forms with infixed -t- (Gt, Dt, Št) and -tan- (iterative/habitual Gtn…). Each stem conjugates in several "tenses"/aspects: PRETERITE (iprus, punctual past action/perfective), PRESENT-DURATIVE (iparras, with doubling/lengthening, imperfective/present-future), PERFECT (iptaras, with infixed -ta-, marking a completed action), STATIVE or permansive (paris, predicate of state/result, conjugated with suffixes: parsāku "I am in a state of having cut"), plus imperative (purus), infinitive (parāsum), participle (pārisum "the one who cuts"), verbal adjective (parsum). Persons marked by prefixes/suffixes: i-/ta-/a-/ni- (3rd / 2nd and 1st sg. / 1st pl.). SYNTAX: basic order SOV (verb at the end of the clause), generally attributed to Sumerian influence; adjective after the noun; subordination by ša (relative/genitival) and by conjunctions (inūma "when," kīma "as/that," šumma "if," aššum "because"). Possessive pronominal suffixes (bīt-ī "my house") and accusative/dative suffixes on the verb.

Reading tips

OPERATIONAL READING METHOD, step by step: (1) First TRANSLITERATE each sign in accordance with the conventions (lowercase italics for syllables, CAPITALS for logograms, superscripts for determinatives, hyphens between the signs of a single word, dots between the components of a compound logogram). Do not yet seek the meaning. (2) Identify the DETERMINATIVES (unread) in order to segment and categorize: ᵈ = deity, ᵐ/disz = man's name, ᶠ/MUNUS = woman, KI = place (after the word), GIŠ = wood, URU = city, KUR = land. (3) Resolve the POLYPHONY: for each sign, test its values until a sequence forming plausible Akkadian words is obtained; rely on the word boundaries suggested by the grammar. (4) NORMALIZE: convert the transliteration into a normalized Akkadian form (šar-ru-um → šarrum), restoring lengths and geminates from the morphology. (5) ANALYZE morphologically: identify the root (often 3 consonants), the verbal stem (G/D/Š/N), the tense/aspect (prefix i-, doubling = present iparras, -ta- = perfect, suffixes = stative), the nominal case (-u/-i/-a + possible mimation). (6) Render the text taking account of the SOV order and the subordinators (ša, šumma, inūma). PITFALLS: never read a logogram phonetically or vice versa without checking; beware of rare values; distinguish dialect/period (Old Babylonian vs Neo-Assyrian change spellings and grammar); long vowels and geminates are often under-noted. HONESTY: for a broken or ambiguous sign or an uncertain value, signal it explicitly (brackets [ ] for restoration, ellipsis for a lacuna, "?" for a doubtful reading) rather than inventing.

Pitfalls

Major pitfalls in reading/translating Akkadian: (1) LOGOGRAM vs SYLLABIC — the same sign may be read as an entire word (LUGAL = šarru) or as a phonetic value; confusing the two falsifies everything. (2) POLYPHONY — a sign has several syllabic values; one must choose the one that yields a valid Akkadian word in context. (3) HOMOPHONY — several signs render the same syllable; the indices (u, ú, ù, u₄) do NOT change the pronunciation, so do not over-interpret them. (4) Unpronounced DETERMINATIVES: do not incorporate them into the word being read. (5) Dialectal and diachronic VARIATION: Babylonian ≠ Assyrian, and Old ≠ Neo- (verbal forms, spellings, mimation present or dropped). (6) UNDER-NOTED long vowels and geminate consonants: restore them from the morphology, not from the spelling alone. (7) SOV order: the verb is at the end of the clause, so do not copy European order. (8) Secondary e: an a may appear as e through ancient coloring. (9) Assimilations (n before a consonant) and vowel syncope obscure the root. HONESTY REQUIRED: signal lacunae [ ], doubtful readings (?), and restorations; do not produce a smooth "translation" where the sign or passage is uncertain.

Decipherment history

Akkadian was deciphered in the 19th century, in the wake of the decipherment of OLD PERSIAN. The Achaemenid royal inscriptions were trilingual (Old Persian, Elamite, Akkadian/Babylonian), in different cuneiform scripts. The German philologist Georg Friedrich GROTEFEND communicated as early as 1802 (to the Royal Society of Sciences of Göttingen) partially correct readings of Persian royal names (Darius, Xerxes, Hystaspes) in the alphabetic Old Persian. Henry Creswick RAWLINSON copied (from 1835 onward) and published the great BEHISTUN INSCRIPTION (Bīsotūn) of Darius I, a trilingual text that provided the decisive key. The deciphered Old Persian gave access, by correspondence, to the third, Akkadian text, which was far more complex (logo-syllabic, polyphonic). The decisive contributions of the 1840s–1850s came from Rawlinson, from Edward HINCKS (who established the syllabic and non-alphabetic nature, polyphony, and the existence of determinatives and multiple values), and from Jules OPPERT (who systematized the grammar and popularized the term "Assyriology"). Public proof of the decipherment was furnished in 1857 by the Royal Asiatic Society: four scholars (Rawlinson, Hincks, Oppert, and W. H. Fox Talbot, at the latter's initiative) independently translated one and the same recently discovered text (an inscription — on a prism — of Tiglath-Pileser I); the concordance of their translations, verified by a jury (report of 29 May 1857), validated the method. The discipline (Assyriology) then took shape with grammars and dictionaries in the 19th–20th centuries.

Common formulae & expressions

šumma awīlum …

"If a man (does X)…" — juridical protasis typical of the Code of Hammurabi and law texts (šumma = "if").

ana … qibī-ma umma … -ma

"To (addressee) say (speak)! Thus (speaks) (sender):" — the stereotyped opening formula of Akkadian letters.

ilū liballiṭūka

"May the gods keep you alive / make you live!" — a formula of greeting and blessing in letters.

ana dāriš / ana ūmī dārûti

"forever / for eternal days" — votive and commemorative formula of royal inscriptions.

ina qibīt ᵈMarduk / ina awât DN

"at the command of Marduk / by the word of the god DN" — formula legitimizing royal action.

MU (mu) … lugal

year-name dating formula ("the year in which the king…"), common in Old Babylonian for dating documents.

ištu … adi …

"from … up to …" — temporal or spatial delimitation frequent in contracts and annals.

Corpora & reference editions

• CAD — The Assyrian Dictionary of the Oriental Institute of the University of Chicago (Chicago Assyrian Dictionary), 21 volumes, 1956–2010, freely available in PDF on the website of the Institute for the Study of Ancient Cultures (ISAC, formerly Oriental Institute)
• AHw — W. von Soden, Akkadisches Handwörterbuch, 3 vols., Harrassowitz, 1965–1981
• ORACC — Open Richly Annotated Cuneiform Corpus (oracc.museum.upenn.edu), annotated digital corpora (RINAP, SAAo, etc.)
• CDLI — Cuneiform Digital Library Initiative (cdli.mpiwg-berlin.mpg.de; also cdli.earth), catalogue and transliterations of tablets
• RINAP — Royal Inscriptions of the Neo-Assyrian Period (a series of editions, Eisenbrauns / ORACC)
• SAA — State Archives of Assyria (Helsinki University Press), editions of the Neo-Assyrian archives
• eBL — electronic Babylonian Library (LMU Munich) for Babylonian literary texts

Bibliographic references

• W. von Soden (with W. R. Mayer), Grundriss der akkadischen Grammatik (GAG), Analecta Orientalia 33, 3rd ed., Pontificium Institutum Biblicum, 1995 — The scholarly reference grammar of Akkadian (in German), the most complete.
• J. Huehnergard, A Grammar of Akkadian, Harvard Semitic Studies 45, 3rd ed., 2011 — The standard teaching manual in English, pedagogical and rigorous.
• R. Borger, Mesopotamisches Zeichenlexikon (MesZL), AOAT 305, Ugarit-Verlag, 1st ed. 2003/2004 (2nd ed. 2010) — The current reference list of cuneiform signs (MesZL/MZL numbering).
• The Assyrian Dictionary of the Oriental Institute of the University of Chicago (CAD), 21 vols., 1956–2010 — An exhaustive dictionary of Akkadian; freely available.
• W. von Soden, Akkadisches Handwörterbuch (AHw), 3 vols., Harrassowitz, 1965–1981 — The reference dictionary in German, complementary to the CAD.
• J. Black, A. George, N. Postgate (eds.), A Concise Dictionary of Akkadian (CDA), Harrassowitz, 2nd ed. 2000 — A handy single-volume dictionary, very useful for everyday reading.
• C. B. F. Walker, Cuneiform (Reading the Past), British Museum Press, 1987 — An accessible introduction to the cuneiform script and its decipherment.
• A. R. George, The Babylonian Gilgamesh Epic, 2 vols., Oxford University Press, 2003 — The reference critical edition of the Epic of Gilgameš.