P3" - ^v 3'2j5 1 WILLI AMS AJNi^ i>uii aiii.Do smu. 1 C' fCV ^^S^^^'s T^^ Student's Comparative Grammar of the I O y French Language, with an Historical Sketch of the -^ Formation of French. For the use of Public Schools. With Exercises. By G. ETJ6fiNE.-FASNACHT, French Master, Westminster School. 11th Edition, thoroughly revised. Square crown 8vo. cloth - 5s Or, Grammar, 3« ; Exercises, 2s 6d. " The appearance of a Grammar like this is in itself a sign that great advance is being made in the teaching of modem as well as of ancient languages The rules and observations are all scientifically classified and explained." — Educational Times. "In itself this is in many ways the most satisfactory Grammar for beginners that we have as yet seen." — Athenceum. Eugene's French Method. Elementary French Lessons. Easy Rules and Exercises preparatory to the " Student's Comparative French Gtrammar." By the same Author. 7th Edition. Crown 8vo. cloth 1« 6d ' Certainly deserves to rank among the best of our Elementary French Exercise-books."— J?(i!tcattonal Times. " To those who begin to study French, I may recommend, as the best book of the kind with which I am acquainted, ' Engine's Elementary Lessons in French.' "— Dr. Brey mam,, Lecturer of the French Language amd lAteratii/re, Owens College, Manchester. The Student's Graduated French Reader, for the use of Public Schools. I. First Tear. , Anecdotes, Tales, Historical Pieces. Edited with Notes and a Complete Vocabulary by Leon Delbos, m.a., of King's College, London. Crown 8vo. cloth 2s The Student's Second French Reader. By the same. Crown 8vo. cloth 2s Lemaistre (J.) French for Beginners. Lessons Systematic, Practical and Etymological. By J. Leisiaistre. To precede Eugene's Method and the various Elementary- French Books. Crown 8vo. 2» 6d Little Eugene's French Reader. For Beginners. Anec- dotes and Tales. Edited, with Notes and a complete Vocabulary, by Leon Delbos, m.a. Crown 8vo. cloth, Roget (F. F.) Introduction to Old French. History Grammar, Chrestomathy, Glossary. Cloth 6* Tarver. Colloquial French, for School and Private Use. By H. Tarver, B.-es-L., late of Eton College. 328 pp Crown 8vo. cloth 5^ Victor Hugo. Les Mis6rables. Les Principaux Episodes. Edited, with Life and Notes, by J. Boielle, Senior French Master, Dulwich College. 2 vols. Crown 8vo J23 °^°^^ each 3s ei Krueg - 6i iBoi'elh E H' (T- -f^ Belbos E(- Roussy Fr Vinet (. tir 35 Williau Co Le on iFreuch du( Co 1. Ea 2. Co 3. Co 4. Ml 5. Co 6. m4_-. *?. Chateaubriand. Voyage en Amerique 8. De Maistre. Les Prisonniers da Caucase, Lepreux d'Aoste 9. La Fontaine's Select Fables. — 1» oa ls6d and le ls6d U6d Williams and Norgate's School Books and Ma})s. Fleury's Histoire de France, racontee k la Jeunesse, edited for tte use of English Pupils, witfi Grammatical Notes, by Beljame. 3rd Edition. 12mo. cloth boards 3s 6d Mandrou (A.) French Poetry for English Schools. 2nd Edition. 12mo. cloth 2s Weisse's Complete Practical Grammar of the German Language, with Exercises in Conversations, Letters, &c. 4th Edition. Entirely re-written. 12mo. cloth 6s New Conversational Exercises in German Com- position, 2nd Edition. 12mo. cloth CKey, 5s) 3s 6d Schlatter's German Class Book. A Course of Instruction based on Becker's System, and so arranged as to exhibit the Self-development of the Language, and its Affinities with the English. By Fr. Schlutter, Royal Military Academy, "Woolwich. 4th Edition. 12mo. cloth (Key 5s) 5s MoUer (A.) A German Reading Book. A Companion to Sohlutter's German Class Book. With a complete Vocabulary. 150 pp. 12mo. cloth 2s Ravensherg (A. v.) Practical Grammar of the German Language. Conversational Exercises, Dialogues and Idiomatic Expressions. Third Edition. 12mo. cloth (Key, 2s) 5s Rose's English into German. A Selection of Anec- dotes, Stories, Sac., with copious Notes. 2nd Edition. Cloth (Key, 5s) 4s 6d ■ German Reader, Prose and Poetry, with copious Notes for Beginners. 2nd Edition. Crown 8vo. cloth 3s Hein. German Examination Papers. Comprising a Complete set of German Papers set at the Local Examinations in the four Universities of Scotland. By G. Hein, Aberdeen Grammar School. Crown 8vo. cloth 2s 6d Ahn's German Method by Rose. A New Edition of the genuine Book, with a Supplement consisting of Models of Conjugations, a Table of all Regular Dissonant and Irregular Verbs, Rules on the Prepositions, &o. &c. By A. V. Rose. 2 Courses in 1 vol. Cloth 3s 6d German Method by Rose, &c. Eirst Course. Cloth 2s Apel's Short and Practical German Grammar for Beginners, with copious Examples and Exercises. 2nd Edition. 12mo. cloth 2s Gd For Continuation see the end of the Volume. [41 1 Date Due BEP19 19'5t P I - ^^^^ Cornell University Library PJ 3251.D35 1889 Assyrian grarhmar "'* Kniilfliiif iiiKb «l 3 1924 026 783 211 PORTA LINGUARDM ORIENTALIUM INCHOAVIT J. H. PETEEMANN CONTINUAVIT HEEM. L. STRACK. ELBMBNTA LINGUARUM HEBBAIOAE, PHOENICIAE, BIBLICO-ABAMAICAE, TAEGUMICAE, SAMAEITAivTAE, SYEIACAE, ABABIOAE, AETHIOPICAE, ASSTEIACAE, AEGYPTIACAE, COPTICAE, ABMENIACAE, PEESICAE, T0ECICAE, AilARDM STDIIIIS ACADBMIOIS ACCOMMODAVBBDNT J. H. PBTERMANN, H. L. STRACK, E. NESTLE, A. SOCIN, F. PRAETORIUS, AD. MEEX, AUG. MUELLER, PRIEDE. DELITZSCH, C. SALEMANN, W. SHUKOWSKI, TH. NOELDEKE, G. JACOB, alii. PARS X. ASSYRIAN GRAMMAR BY FRIBDRICH DELITZSCH. BERLIN, H. REUTHER'S TERLAGSBUCHHANDLUNG. LONDON, NEW TOKK, PARIS, WILLIAMS & NOEGATE B. WESTEEMANN & Co. MAISONNEtTVE H, HKMEIETTA STREET, 838, BEOABWAY. & OH. LBCLBEC OOVEHT OAEDEH. 25, QUAI VOLTAIEE. 1889. 4SSTEIANGRAMMAR WITH PAEADIQMS EXERCISES GLOSSARY AND BIBLIOGRAPHY BY FRIEDRICH DELITZSCH. TRANSLATED PEOM THB GERMAN ABCHg. E. s. KENNEDY, B. D. PROr. OP ORIENTAL LiNGUAGES, DNIV. OP ABERDEEN. BERLIN, H. BEUTHER'S VERLAG8BUCHHANDLUNG. LONDON, NEW YORK, .PARIS, WILLIAMS & NOBGATE B. WBSTERMANN & Co. MAISONNBaVB 14, HENBIETTA STBBBT, 838, BEOADWAY. & CH. LEOLBRC OOVEMT QAEDEN. 25, QtTAI VOLTAIEE. 1889. — All rights reserved,. — DEDICATED AS A TOKEN OF EESPECT AND INDEBTEDNESS TO MY EEIEND PAUL HAUPT. The original of tliis book is in tine Cornell University Library. There are no known copyright restrictions in the United States on the use of the text. http://www.archive.org/details/cu31924026783211 AUTHOR'S PREFACE. The present work is intended to be of use not only to Assyriologists but to Semitic scholars generally, by presenting them, in as brief compass as possible, with a summary of the latest results of research in the department of Assyrian grammar. In return, it appeals not merely or chiefly to Assyriologists but to every Semitic scholar for their co-operation in the solution of numerous unsolved problems, some of which, at least, are of the utmost importance for com- parative Semitic philology. Its aim is thus the same as that of my "Assyrisches Handworterbuch" (Leipzig, Hinrichs 1889) the compilation of which has become possible by the appearance of this grammar. My larger dictionary, which is also of the nature of a con- cordance, will he continued without interruption and with renewed energy. The brevity of the Chrestomathy may appear strange. But even had I extended it to many times the size, it would still have been practically useless. For acquiring familiarity with Assyrian literature, even as represented by the so-called historical texts, and with the monuments of the Old and New Baby- lonian empires, not to speak of the so-called Sumerian texts, it is indispensable to have a much more com- prehensive table of characters on the one hand, and on the other, a more extended chrestomathy such as my "Assyrische Lesestiicke". By-and-by it will be VIII adthok's preface. possible to recommend simply the first or fifth vol- ume of Rawlinson's "Inscriptions of Western Asia", supposing, that is, that one or other of these volumes will again be obtainable. I have accordingly selected only a couple of historical texts, an easier and a more difficult one, furnishing these, however, with footnotes and a glossary in such a way that they will, I trust, afford the student his first lesson in the reading and interpretation of Assyrian, and introduce him to the use of this grammar. I am well aware of the many defects incidental to this first essay but I intend to continue with un- wearied diligence the investigation of such problems as have not yet been satisfactorily solved — among which I partially include the discussion in §§ 12 — 14. It shall be my earnest endeavour to bring this manual abreast of the newest results of Semitic philology in general and of Assyriology in particular, and to main- tain it thenceforth in that position. Leipzig, Easter 1889. Fricdricli Delilzsch. Note by Translator. It was the intention of the publishers that this English edition should appear almost simultaneously with the original. The delay has been caused by the demands of my own work during the winter. I shall consider myself amply rewarded for my labour of love in translating and editing Professor Delitzsch's book, if in its English dress, it should prove a boon to the aspiring student in this country and America to whom the German original is a sealed volume. Aberdeen, 12tt June, 1889. A. E. S. K. TABLE OF CONTENTS. Grammar. Introduction (§§ 1 — 5). Page § 1. Remarks on Assyrian 1 § 2. Short history of the excavations 1 § 3. Short history of the decipherment 4 § 4. Short history of the grammatical investigations . . 6 § 5. Brief summary of the contents of cuneiform literature 8 The Written Character (§§ 6—25). § 6. The viredge 11 § 7. Origin of the Assyro-Bahylonian wedge- writing . . 12 § 8. Development of syllabic wedge-writing 14 § 9. Table of characters 17 § 10. Long vowels — how written 41 § 11. Doubled consonants 43 §§ 12 — 14. The signs ia, a-a, a-ia, ia-a 44 § 15. The vowel e 47 § 16. The vowel u 49 § 17. The syllable 50 §§ 18 — 22. The consonants — how written 51 § 23. Eeading-signs 56 § 24. Practical hints 59 § 25. Invention of wedge-writing (the 'Sumerian' question) 61 X TABLE OF CONTENTS. Pago Phonology C§§ 26-52). A. Vowels. §§ 26 — 31. Vowels sounds . . 72 § 29. e in Assyrian 74 § 30. Fusion of e and i 77 § 31. Diphthongs 79 §§ 32' — 39. Phonetic changes affecting the vowels ... 81 § 32. Change of a to e 81 § 33. Change of a to e . 84 § 34. Change of a to e . 85 § 35. Transition of unaccented short a to i . . . 88 § 36. Transition of i to e before r and h . . . . 89 § 37. Syncope of short (and long) vowels ... 90 § 38. Contraction of two vowels 92 § 39. Complete loss of vowels 94 B. Consonants. §§ 40 — 46. Consonantal sounds 96 § 41. The semi-vowels M and i 96 § 42. The aspirate or hreath 100 § 43. Explosives (the nS3"tJ2 in Assyrian) .... 101 § 44. The labial nasal m . . 103 § 45. The liquids 105 § 46. The sibilants 106 . §§ 47 — 52. Phonetic changes affecting the consonants . . 110 §47. The breath (BawcWmrf) 110 § 48. b, d and t 112 § 49. Nasals 113 § 50. Liquids 118 § 51. Sibilants 118 § 52. Compensation for doubling 121 § 53. Place of the tone 122 TABLE OF CONTENTS. XI Page Morphology (§§ 54—118). A. Pronoun. § 54, Preliminary remark 129 § 55, Independent personal pronouns 129 § 56, Suffixed personal pronouns 133 § 57. Demonstrative pronouns 138 § 58. Relative pronouns 140 § 59. Interrogative pronouns 142 § 60. Indefinite pronouns 142 § 61. Significant roots 144 B. Noun, § 62, Primitive nouns 147 § 68. Nominal stems from verbs med, geminatae . . . 151 § 64. Nominal stems from verbs mediae 1 and "•.... 153 § 65. Summary of nominal stems in Assyrian . . . , 158 § 66. Inflexion for case 181 § 67. Plural of nouns (except those with feminine ending) 183 § 68. Formation of the feminine 187 § 69. Plural of feminines in atu 189 § 70. Mixed plural formations 190 § 71. Gender 192 § 72. Status constructus 193 § 73. Composition of words 195 § 74. Union of the substantive with the pronominal suffixes. Appendix to the pronoun and noun: Numerals and Particles. 205 §§ 75—77. Numerals 205 § 75." Cardinals 205 § 76. Ordinals • 207 § 77. Other numerals 208 §§ 78—82. Particles 211 §§ 78—80. Adverbs 211 XII TABLE OF CONTENTS. » Page § 81. Prepositions 224 § 82. Conjunctions 229 C. Verb. § 83. Principal verbal stems 231 § 84. Signification of the principal verbal stenis . . . 233 § 85. Shafel and Ishtafal formed from the Piel .... 236 §§ 86, 87. Permansive and present-preterite themes of the Qal 237 § 88. Permansive and pres.-pret. themes of the augmented stems 242 § 89. Signification of the permansive 247 § 90. Conjugation of the pres.-preterite theme .... 252 § 91. Conjugation of the permansive theme 255 § 92. The modus relativus 256 S 93. Precative 258 § 94. Imperative 261 § 95. Participles. Iniinitives 263 §§ 96 — 98. Strong verbs including verbs medise geminatse 264 §§99—101. Verbs primse 3 275 §§ 102—104. Verbs primse gutturalis 281 S§ 105—107. Verbs mediae gutturalis 292 §§ 108—110. Verbs tertise infirmse 297 §§ 111—113. Verbs primse 1 and i .308 §§ 114—116. Verbs medise 1 and i 314 § 117. Quadriliteral verbs 319 § 118. Union of the verb with the pronominal suffixes . 323 Syntax (§§ 119—152). A. The individual Parts of Speech. §§ 119—127. The Substantive 325 §§ 119 — 120. Subst. with pronominal suffixes . . . 325 §§ 121—122. do. with adjectives 326 § 123. do. with another subst. in subordina- tion (constr. state) 328 TABLE OP CONTENTS. XIII §§ 124 — 126. Subst. -with another subst. in apposition . . 330 § 127. do. with another subst. in co-ordination . 333 §§ 128—129. Numerals 334 § 130. The adverb 337 §§ 131—133. Verbal nouns 337 § 131. Participle 337 §§ 132—133. Infinitive 339 §§ 134—139. The finite verb 340 § 134. Meaning and employment of the tenses and moods 340 §§ 135 — 136. The pronoun governed by the verb . . 342 §§ 137—139. The substantive governed by the verb . 345 B. The Sentence. §§ 140- — 149. The simple sentence 348 §§ 140 — 142. Declaratory sentences 348 § 143. Negative declaratory sentences 351 § 144. Prohibitive sentences 352 § 145. Optative and cohortative sentences .... 353 § 146. Interrogative sentences 355 § 147. Attributive relative clauses 355 § 148. Relative clauses introduced by a conjunction 357 § 149. Conditional clauses 360 §§150 — 152. Union of several Sentences .... 362 §§ 150 — 151. Sentencesjoined together by a conjunction 362 § 152. Circumstantial clauses 364 Paradigmata 1* A. Pronomen 3* B. Verbum trihtterum 8* C. Verbum cum pronominibus suffixis 32* Chrestomathia 33* XIV TABLE OF CONTENTS. Glossarium 33* Litteratura 41* A. De inventione atque effossione monumentorum cuneatorum 53* B. De iuitiis ao progressibus explicationis .... 60* C. Editiones textuum 64* D. Libri grammatici et commentationes grammatioae 67* B. Translationes et interpretationes textuum . . . 70* F. LexicograpMa 73* G. Scriptiones periodicae et collectanea .... 74* Appendix 76* Corrigenda 78* Addenda 79* ABBREVIATIONS. ABK vid. Litteratura 134. — ALs v. Litt. 127. — ASKT v. Litt. 110. — Assurb. Sm. v. Litt. 169. — Assurb. S. A. Sm. 11 V. Litt. 123. — Assurn.: Large Alabaster Liscription of Assur- nasirpal I B 17 — 26. — Assxirn. Balaw. : do. Slab Inscription from Balawat V K 69. 70. — Assurn. Moil.: do. Monolith In- scription III B 6. — Assurn. Stand.: do. so-called Standard In- scription, Layard 1 (with variants 2 — 11). — Beb., NR and the other Achaemenian inscriptions, D, K, etc. are cited in the usual way: for Beh. v. Ill B, 39. 40, for the rest Bezold's "Achaeme- nideninschriften" (■;;. Litt. 113). — Ca, C": Assyrian Eponym Canon, published in the second edition of my "Lesestiicke" (AL^). — Desct.: Legends of Ishtar's Descent to Hades IV B 31. — Diet. V. Litt. 211. — E. M.U V. Litt. 84. — Esarh. : Six-sided Prism of Esarhaddon I B 45—47. — Hanim. Louvre : Inscription of Ham- murabi, V. MSnaut's "Manuel etc." (Litt. 143), pp. 306 — 312. — K.: Tablets of the Kuyunjik collection in the British Museum ; for the books etc. in which they have so far been published, as also for the tablets indicated by S. (Sm.) orM, see Bezold, Kurzgefasster tjberblick fiber die babylonisch-assyrische Literatur nebst einem chronologischen Excurs, zwei Begistern und einem Index zu 1700 Thontafeln des British-Museums. Leipzig 1886. (XV, 395 pp. 8). From the third edition of my "Lesestiicke" (AL^) are quoted: K. 3437 (p. 97fif.). K. 4378 (p. 86 ff.); also Fragm. 18 (p. 95 f.) and Sm. 954 (p. 134 fif.); from ASKT: K. 56 (= II B 14. 15, p. 71 ff.). K. lot (p. 115 f.). K. 133 (p. 79 ff.). K. 246 (= II B 17 f., (p. 82 ff.). K. 3927 (p. 75). K. 4350 (= HE ll, p. 45 ff.); — from Pinches' Texfe; K. 196. K. 823. K. 831; — from Assurb. S. A. Sm. 11: K. 95. K. 359. K. 509. K. 538. K. 562. K. 2867. Note also: K. 64 = II B 62 No. 3. K. 245 = n E 8. 9. K. 4341 = II B 36 No. 3. K. 4386 = U B 48. — Khors. v. Litt. 106. — Lay, V. Litt. 104. — 1 Mich.: Caillou de Miohaux I E 70. — Nabon. : Cylinder inscription of Nabonidus I B 69. — Neb.: Slab XVI ABBREVIATIONS. Inscription of Nebuchadnezzar I E 53—58 (50—64). Neb. Bab. or Bors. or Senk. : do. Cylinder Inscriptions from Babylon (I B. 52 No. 3), Borsippa (51 No. 1), and Senkereh (51 No. 2). — Neb. Grot. : do. Cylinder Inscription, first published by Grotefend, I K 65 — 66. — Nerigl. : Cylinder Inscription of Neriglissar I K 67. — Ninir. Ep.: v. Litt. 116 (Nimr. Bp. XL XII. denotes the eleventh and twellth tablets of this epos as copied by me; for the former see AL3 p. 99 ff. Tablet XII is now edited by P. Haupt in Delitzsch-Haupt's "Beitrage zur Assyriologie r.nd vergleichenden semitischen Sprachwissenschaft", I, 1889, 48—79). — NR v. Beh. Pinches, Texts v. Litt. 112. — Proll. v. Litt. 210. — IR, IIR, etc. V. Litt. 105; the numerals after R denote the page and the line, the letters the columns. — S. or Sm. (tablets of the Smith collection of the British Museum) v. K. — S", S^, S""-, syllabaries published in AL' pp. 41—79. — Sarg. Cyl. or Bull Inscr., cited as in Lyon's "Sargontexte", «. Litt. 115. Sarg. Cyp.: Inscription of Sargon on the monolith found in Cyprus IIIE,ll,cf. Schrader's new edition of the text mentioned in Litt. 111. — Senhb. : Six-sided prism of Sennacherib IE, 37 — 42. — Scnhb. Bav.: do. Eook in- scription of Bavian III E 14. — Senhb. Bell. : Lay. 63—^64 (my numbering of the lines omits the heading and consequently differs by one line from Layard's). ■ — Senhb. Const.: do. Slab inscription, now in Constantinople I E 43. 44. — Senhb. Kuy : do.: Inscription on the Kuyunjik bulls III E 12—43. — Senhb. Rass. : Sennacherib Cylinder in Eassam's collection. — Scnhb. Sm. : v. Litt. 175. — Shalm. Balaw. : Inscription of Shalmaneser II on the bronze gates of Balawat, v. Litt. 109. — Shalm. Co.: do. Two inscrip- tions on colossal bulls, published Lay. 12 — 16, 46 — 47, cited according to my own arrangement. — Shalm. Mo.: do. Monolith inscription III E 7—8. — Shalm. Obv. : do. Obelisk inscription Lay. 87—98. — Shalm. Throne Inscr.: v. Litt. 121, I9i. — Shams.: Obelisk inscription of Shamshi-Eamman 1 E 29 — 31 (32—34). — Strassm. v. Litt. 208. — Str. L II. v. Litt. 118. 125. — T'S"' Eight-sided prism of the elder Tiglathpileser I E 9—16. — Tig. jun.: Tablet inscription of the younger Tiglath- pileser II E 67. — Zurich. Voc. v. AL3 p. 84 f. Familiar abbrevations such as cf. = compare (confer), V. or vid. = see, etc. call for no further notice. INTKODUCTION. Assyro- Babylonian or simply Assyrian is the § name given to the language of the Semitic literary monuments in Babylonian or Assyrian wedge-writing. Though known to us for only a few decades, in a ' literature of surpassing richness, Assyro-Babylonian was the tongue of Semitic empires on the Euphrates and the Tigris, which reach back into the fourth millenium B. C. — the Old Babylonian, the Assyrian, and the Neo- Babylonian. After the destruction of Nineveh (c. 608) and the fall of Babylon (c. 538), it continued to be spoken in Babylonia during the period of the Achsemenian kings (cf. the Persian cunei- form inscriptions of the so-called third species), till at last, in the second century B. C, it had gradually to give way before the Aramaic dialect, leaving behind it its last important monuments from the times of the SeleucidEe. It is almost exclusively to excavations that we § 5 owe our possession of an Assyro-Babylonian litera- ture. The following are the principal dates: Assyria. Final identification of Nineveh with the two mounds Kuyunjik and Nebi Yunus by Rich Delitzsch, Assyrian Grammar. 1 2 IHTKODDCTIOH: § 2. EXCAVATION. in 1820. Excavation of Lur-Sarruken^ the city of Sargon, in the mound of Khorsabad by the French- men Emil Botta (1842—1845) and Victor Place (1852). Excavations in Nimrud (Kelach) and Nineveh by the Englishmen Austen Henry Layard (1845 — 1847; 1849—1851), Hormuzd Eassam (1852 — 1854), George Smith (1873; 1874; 1876, f 19. Aug. 1876), Hormuzd Eassam and those under his direction (Nov. 1877 — July 1882) : discovery of the palaces of Assur- nazirpal, Shalmaneser, Esarhaddon and others in Nimrud, the South-West Palace of Sennacherib and the North Palace of Assurbanipal (Sardanapalus) in Kuyunjik; discovery, among the ruins of the latter, of the library of clay tablets belonging to Assurbanipal by Eassam 1854 (several thousand literary works, partly of Assyrian origin, partly copies of Babylonian originals ; hitherto only a part recovered — c. 30000 (?) fragments, hidden away in the British Museum). Be- ginning of the excavations in Kileh Shergat (Assur) by Layard and Eassam 1853. Eassam's discovery of "The Bronze Gates of Shalmaneser 11" in the mound of Balawat 1878. Babylonia. Exploration of the extensive ruins, Babil (Babylon), Birs Nimrud (Borsippa), Niffer (Nippur), Warka (Erech), Senkereh (Larsam), Ur (Mugheir, al-Mukayyar), Abu Sharein (Eridu), by the Englishmen Loftus and Taylor under Sir Henry Eawlinson's superintendence (1849 — 1855) and by the French expedition (sent out under the leadership of Fulgence Fresnel and Jules Oppert 1851 — intkoductjon: § 2. excavation. 3 1854 ; loss of the collection in the Tigris May 23, 1855). Discovery by Arabs in the mound of Jumjuma (Baby- lon) 1874 of more than 3000 tablets (1"— 1' square) containing records of private business transactions ; the collection was secured by George Smith in 1876 for the British Museum, and is being added to from year to year. Kassam's Babylonian expeditions (1879 — July 1882), which resulted in the discovery of Sepharvaim in the extensive ruins of AbuHabba 1881; the excavation of the temple of the Sun, and the find- ing of the temple archives, consisting of clay cylinders and (ace. to Kassam's calculation) c. 50,000 unfortu- nately very badly burnt, clay tablets; exploration of the two largest mounds of Babylon, Babil and Kasr, and of Tel Ibrahim, the site of Kutha ; discovery of the palace of Nabonidus in Borsippa. E. de Sarzec's excavations of the ruins of Tello or Tel Loh (1875 to 1880; 1882 purchase of the collection for the Louvre). North American (Wolfe's) expedition 1884 — 1885. Systematic excavations are at present being continued only in Tello ; in stead of which there has been for years an active trade in Babylonian antiquities, especially in clay tablets and cylinders, often of the highest scientific value, which have been dug up by the Arabs in situ, and purchased in the East or in Europe, for the museums in London, Paris, Berlin and elsewhere. Of rock inscriptions — apart from those of Darius on the rook- hewn graves of Naksh-i-Bustam (near Persepohs) and on the face of the rook at Behistun (Media) — the following are the most notable: Inscriptions and sculptures of Tiglath-pileser I and of 4 INTRODUCTIOS ; § 3. DECIPHERMENT. three of his successors at the entrance to the fountain grotto c the Sebeneh-Su, the left parent stream of the Tigris; Senbacherib Bavian inscription (Assyria) of 60 lines; two inscriptions of Nebi chadnezzar in the Wady Brissa (Lebanon), together comprisin 19 columns. — Details as to excavation and editions of the tea V. Litteratura A, b. and C. — Museums -with Assyro-Babyl. ant: quities: British Museum, London; Louvre, Musee de Ckrqc an Bibliothfeque nationals, Pai-is; Museums in BerKM, ConstantinopL New York, Liverpool, the Hague, St. Petersburg, Zurich (Vatica in Eome, Leyden, Brussels, Graz and others). 3. The decipherment of the Assyro-Bahylonian wedge writing, i. e. of the wedge-writing in the third columi of the Achaemenian inscriptions, is based on th deciphering of the Old Persian wedge-writing in th first column, a work of genius on the part of G e o r j Friedrich Grotefend (who made out the name Darius, Xerxes, Hystaspes; 14. Sept. 1802, translatio: of the first two Achsemenian inscriptions), Eugen Burnouf, Christian Lassen (both 1836; firs employment of the Darius inscription J with its lis of satraps) and Henry Rawlinson, who copiei (1835—1837) and (in 1846) explained the Behistun in scription. After the Old Persian alphabet of 40 sign had been fully deciphered, Hincks and Jules Op pert, Benfey and Spiegel carried on and com pleted the thorough investigation of the Old Persia: language. The observation made by Botta and others, tha the characters on certain monuments excavated i] Assyria and deposited in the Louvre, notwith standing manifest differences, were identical wit] IBITR0DUCTI0K4 § 3.. DECIPKEKMENT. 5 those of the third column of the trilingual inscrip- tions, was the starting-point for the decipherment of this third species. The Old Persian monuments with their phonetically written proper names, especially since the number of those clearly made out had been raised from ten to ninety by Sir Henry Raw- linson's publication of the Babylonian part of the Behistun inscription, were, in this connexion, what the Eosetta stone with its Greek text was to the de- cipherers of the hieroglyphics. Moreover it was soon remarked that, in the Baby- lonian translations of the Old Persian texts, all the names of countries, cities, deities and persons were indicated by a particular sign (the so-called deter- minative) placed immediately before them, a discovery which resulted in fixing, without much difficulty, a considerable number of signs with their approximate values. While, however, the proper names above referred to would have been tolerably sufficient for the decipherment of inscriptions in alphabetic characters, it soon became apparent that the Baby- lonian wedge-writing was not alphabetic. One dif- ficulty surmounted, the decipherer found himself face to face with another. Sir Henry Rawlinson, who arranged the signs occurring in the Babylonian Text of the Behistun inscription, to. the number of 246, was the first to detect the polyphony of the Babylonian signs (Sept. 1851), while Hincks, with equal acuteness, first detected the ,syllabic character of the Babylonian writing (1849 — 1852), and thereby put an end to the 6 inteodootion: § 4. gbammatioal investigation. grand mistake of supposing that tke Babylonia writing was alphabetic, with several signs for eac individual letter (de Saulcy, and, for a considerab] period, Kawlinson). The work of decipherment w£ brought to a close by Jules Oppert (1859) with ti aid of Assyrian sign-lists or "syllabaries", which ha meanwhile been discovered. These lists, inter alu explained the signs representing so-called compoun syllables, that is, syllables containing two consonant with a vowel between, by the signs for the corresponc ing simple syllables, e. g. No. 162 of the table of sign by da-an, ka-al, ri-ib, No. 206 by ha-ab, ki-ir, ri-in, and alongside the signs representing whole words, th so-called ideograms, they gave a phonetic reprodm tion of their meanings, e. g. No. 165: = a-hu and nc sa-ru. All Assyriologists of note (Menant, Norrii Talbot, George Smith, Sayce, Schrader and others have been, and still are, active gleaners in this ric and fruitful field, removing old and false, and discover ing new syllabic and ideographic values. A new tas' — perhaps the last — was undertaken in the increas ingly successful deciphering of the archaic signs oi the oldest bricks, clay cones, and cylindrical seals and especially on the monuments discovered b; de Sarzec at Tello, a task which, thanks to the labour of Amiaud and others, is, like its predecessors, grad ually approaching completion. For details see Litteratura B, a and b. The beginnings oi grammaticalinvestigation coincidi with the early attempts at decipherment, notabb IHTRODUOTION: § 4. GEAMMATICJAL INVESIiaATION. 7 those of de Saulcy, who was the first to read the personal and possessive pronouns, and to recognise the relative and a few verbal forms. The same scholar made a number of observations regarding gender and number which still hold good, and led the way (1849) in the analysis and explanation, as Semitic texts, of such Achsemenian inscriptions as were then accessible. Sir Henry Eawlinson, thereafter, in his translation of the Babylonian Behistun text, succeeded in referring words and phrases to Semitic forms. The first to elaborate on a firm basis a system of the pronoun and verb, and consequently of the elements pf the grammar was Hincks (1854 — 1856); he even thus early com- pared the Assyrian verbal inflexions with those of Hebrew, Syriac, Arabic and Ethiopic, and sought in this way to determine more precisely the position of Assyrian in the group of the Semitic tongues. It is, however, to Jules Oppert (1860), that we owe the first connected presentation of Assyrian grammar; he it was who, aided by the results obtained by other Assyrian scholars, noted the various parts of speech, and discovered the mimation, the pronominal inflexion, the formation of the derived verbal stems, the femi- nine forms in the inflexion of the verb, and the general features of the syntax and composition of words. Oppert and Hincks continued their grammatical in- vestigations, while Schrader tested the reliability of the decipherment, and the results it had yielded for the Assyrian grammar and lexicon. Among younger A«syriolpgists Pognon substituted the only admis- 8 INTKODUCTION : § 5. LIIEEATUEE. sible reading ma for the false reading and suppose( copula ya. It is, however, Paul Haupt in particula who, in a series of papers distinguished by observation of great acuteness, has not only freed the gramma of Assyrian from numerous errors, more especially ii the department of phonology and morphology, bu has roused grammatical research to new life by hi breadth of view, and the originality of his suggestions DetaOs in Litteratura D. Por the foundation of Assyria Lexicography and the carrying on of the work v. ibid. E and I § 5. The literature preserved in the Assyrian tongu occupies on the score of age the first place amon, the literatures of the Semitic peoples. The oldest o the longer, phonetically written, Semitic texts know: up to the present time are, in Babylonia, those o Hammurabi (c. 2200), in Assyria the large stone in scription of Rammannirari I (c. 1350) containing eighty lines. These are followed by the inscription on the octagonal clay prisms of Tiglathpileser (1110 B.C.), each containing eight hundred lines, am by the bulk of the Assyrian and Neo-Babylonia; literature, from Assurnazirpal to Assurbanipal, fror Nebuchadnezzar to Nabonidus, and later from Cyru to Artaxerxes and even to Antiochus I. Soter— comprising, therefore, a period of more than tw^ thousand years. Engraved on bricks, prisms and cylin ders of clay (the latter usually built into the corner of palaces and temples), on slabs of marble am alabaster, on statues, obelisks and colossal bulls, and bstkoduction: § 5. literature. 9 above all, on clay tablets of every shape, a literature has been recovered -which already far exceeds in compass the whole of the 0. T. scriptures, and which, as now from year to year, will doubtless continue to increase for an indefinite period, till at length the wealth of material shall become inexhaustible. Texts of many hundred lines record the wars of Assyrian and Babylonian kings, their buildings, their deeds at the chase and elsewhere, and unfold a lifelike picture of the politics, culture and geography not only of Babylonia and Assyria, but also of all the neighbouring peoples of Western Asia. In addition to these, chronological lists and records of every sort (eponym-lists, chronicles, synchronous histories, tables of kings) render it possible to fix the dates of individual dynasties and monarchs, and afford unimpeachable chronological data reaching back into the fourth millenium B. C. Prayers and psalms, legends of the gods, stories of the creation, lists of the gods and exorcisms of varied content, a great epic in twelve books, along with a great number of astrological tablets, curious lists of secret remedies, oracular deliverances and calendars afford a profound insight into the religion, mythology and superstition of those nations. As companions to the tablets of purely scientific, astronomical, and mathematical contents, we may place long lists of words from the same or a similar stem or having the same or a similar ideogram, lists of synonyms, of the names of occupa- tions, persons, stars, animals, plants, clothes, wooden 10 INTBODUOTION : § 5. LITERATURE. instruments and utensils, paradigms, collections of signs, all well adapted to lead us, as they led the pupils in the priestly schools of Babylonia and Assyria, to a more and more thorough understanding of the Assyrian language as it was written and spoken. Numberless letters and contract tablets, finally, reports of generals and astronomers, proclamations and petitions, deeds of purchase and sale of every description, marriage certificates, bequests, wills, house inventories, receipts and so on reveal the social life of the Assyrians and Babylonians even to its most secret recesses. THE WRITTEN CHARACTER. The characters in which the literary monuments § of Assyria and Babylonia are written have the form of wedges — the writing running from left to right.. Under the name of wedge-writing are comprised all the species of writing in which the fundamental ele- ment is a so-called wedge (Latin cuneus, whence the now usual name cuneiform). This wedge may be described as a stroke proceeding in a straight line from a hollow three-sided pyramid, or, as it appears when graphically reproduced on paper, from a triangular head, the stroke being attached to the apex of the triangle, opposite the initial base. In addition to the horizontal (^> — ), perpendicular (Y) and sloping ( A^ , J^, y ) wedges, there is the double wedge (< m^adminumyfia^m axj noma sufymmmfi 4/imia-. Mdadd dmrv. 111- n" J uJrjfL.CUC 'C ILU OMil /dMotlL ten ion. (t4).jmwm. mstiAltufi. Tvzacfc. •U^ 2^ t 19 4i du. u^l-tii \1L1 fzssmrc juji- ^id/^Mn).-ajkm ao, fe£r ^cz fc ^Msu.tt> .- „.. ixeciaize,. Mni./aU).-£nu/au^^t.Mtiu ami- te/iwict . MOM ieiact. 'numaAont. t ■ rii '" "Li Jkmiset. £t JLL coin. ^\^^)iMfMi^maaOui(i ^ ^ ^ 4ajnmiamvmu. ^esam.. ^ i^.Jk^i^CLuuu^iMA. 27 'V" srL y,. «< ^^ sj, ^ J^i as / K V ILLS tCL tuL Cc and JduaiafiAic Utiuet—ViauJir o^ xanf. Jg^ ^^uidub to Juwjeit. tkectane^). /fcat^mit.-jKaiu. JLad.aisoMt.- M uAcbrui (jioait o^tL -pi^ez, ^fimc.^ Aic&7mn,mec. ain&cLSwtt. ZUM-d^il—iruL 131. Cmtiaduiii,^ Msuz JtamtL'.(^>-(^^./hi ruums o^jfus^cmsahiJat tie. doiJtsux.T'wiajm, aax.and.afiIu,Sffti. >f^>2.._«< ^-fJumssLL dmsum.. tiii—liUi^^ TtuJkJmu. t^ rmidsL 60. j[£tu.^u^oWt;cf(,idso:iMLmA-^aui in, in dt camfutftf 0^. <^n rmittai ^TTT. piAu. a^'oA-^m fftMu (kid. midtu, dive. 28 ^.^^ofdojuutoisB. miA tfi£. aiccefitOm. ocm/iit given. tavUii/i . 101. M*9ffT ffijMiJric and. MtMUpAic Tallies.—. ^zauM of !^ns: lujd./Ln.zuj^i-. ^ Ml). f^.Mi& 0^ ^,iuiaetijuB. 29 i/Uti. 11k: ►?- 11».*^1r- za jfucdbct.^ tk, tcitid'(a^an. oMiim), /uc. AiLim.- iiUatii taxi. r if 111. flj. ^ ^ tL^jiiikatu.jJ«htvicaoti^zaIeic,. imi.zlJuL.-ZuJkl^aJlia)f[aoit,rwik.^^ ■ ZiJu&niaanktef.-i^>-^aSfBiMcAc 30 ft*. iW ni. ^ •00. • or. 1:31^ moM. dufi^ibi^iuL toMetJajjioM, taMm.fM£ out. ij.3al&. e^ Chmadoii'B ■ 31 %» iia. 1*1. I'm-. m.^ ffl.*?r ^yix,.^ £cc. Mz . _ sit&ui wzax.. Ms. zoj&I^ AmdruiMait, am/uU^n.Jimtm. tcLvi zhuL taJmt. tuzl^^ai^.-(dm.£t!mt.dstem.. Se^.tuvnad (ilnMe. 7Mdk({).^ ^ mnuJaac ytC/^^.di. -ifi^UL imK£,JarwL cwit.knuttL 32 i^. Jaite, a^ Cha/LactaieB. •%» 4Si. ^ «♦•■ ^ ftj. ^m i6k\ a&MiUc(,mmus. mat. ffisCfi^. rwL. SjjfditiiCHMoxteM'B. 33 ^- J uc. .-.^t wamatj. 34 «j.^ «* its. ^1^ namui, rmduL Jid^. cafi. (jim).^'ma,'am mum. laliz.muji.Mih. 7ruik._£k afurri^avm,. Mtll.xm,. Ami. %^- Mtti.^/fdmiJm., as. if. JilMt M CkuuiitJUuB . 35 <5I ■^ dun.J/id. djtll, tuijv . -.am to. ^&un. Je^.Tnascfuiofm 'nmes. M.^(ily^£Mah. wsi^.MmSJiLtbMcnesir. 4iynlitii tarn,. mjj.- ^n o(dt phid. 4ljial{uucl,._3idem.Iel.fni.fAjfu. rwmJx^ jemlf^^mmmjiL cau^mt^ mCmma. ang- djtOffli.- aMc[tu,,Mjd3i. mmm,;) Jkm.Aimi.^ ^ms:fm)- 36 §^Xlit ^tt^Afitau C. 5f^.. f^HJic W OdeofufiCi^ 'KLu— ^iff^ ^• i^Jfc tLUl/^.-.MJc.,yko(.. ^».w^ ff- . . «^-A>^ M,iUi,M.(jim\^. ytfL. ■as. 221.' 32). U0. CJd&omms ToiiLAi ji3xejCtjJm,(i(tLattuvenmuljoiAiad.D. ■wc [M.^o).^»^,%*^'^ cibj.JmiMik,. toMjaAiitk.Mo ^. j^juhuLtdtjMjr. ^A, '(^:.Juitii hem,. ij. JaMt 1^ CHa/Uj^iiU C. 37 m..\ 243.' W.^ J)dtu.iM/£ute,.tidveit. yfamdJa -^S. Jcnm^ Jim. Jcmvt. aiso^^^ Jami xbif. dmimiM. /tm,.fed^,JSodu-.Mm..Je^..fia€icf;lh ^Carij mm. zmku. (iss^akot$^^)yaiso a. ijieascxe(nrin'^.. ?r ^^ ^nu Jii-ass. t^lS:^^ vi^jfmmtuL 4tcuutnu ■mini,. 38 if.JaHl oi CHauctuis C 2^1 4m Vi^ti Jt ncdc, Tii/ilu. Tdk^JfiMi. wrvrm irustkc. 4CUa/iu. y^&m..(f.aj!£o Msf ahjviMde&im..^ ^u/umJeam-. alfuiax^JUl. iwui Zdmn .(fyMu. value, at). 4/j^ JiaxM. q.aisoX-. So. am£w,Mman.Idn^^rnari. Qetimi.(dso^, Kami nwUty mcoantei, .^^. Mhykut,mdit.J^^ ^(edso <^)at^ / rum. djiinK-o^n^.o^nf. jamdiai momm,. jKaJku. Mdv^j Amcuzed. Cf-oho Jo.SJh mamL MiJcd, diMaJt. nujisa iMnmr. i^. JMt ol OMUxtvis C. 39 %»• ford vo^. ^ <*^:Smamaihi£.. daymti mudca£,^vauxjd&,. durmu. amct. CafuA u(ci)imcL.(^^Mi: vaht m). mj/ULiit ziJdC.(^MisoJ(o.6o. ^ o^zefidlUen,,mt Jo. i£i . jpiTT t^. i)jkdaic aid Md.3!>duimmi Jnm ofcodmi ^ians:ii>.wd.. THE WKITTEN OHAEAOTER: § 10. VOWELS-HOW WRITTEN. 41 How the Vowels are written. — In Assyrian the § 10. vowel of a syllabic sign may, per se, be regarded as long or as short. Even the breaking up of shut syllables like kar, kir, kur into ka-ar, ki-ir, ku-ur does not necessarily imply that the vowel is long. When it is intended to indicate expressly that a vowel is long, this is done, in the case of open syllables in the middle and end of words {Wortin- und -auslaut), by adding the sign for the vowel in question (a, i, e or m): thus we find either li-Sa-nu or H-Sa-a-nu 'tongue', ni-ru or ni-i-ru 'yoke', be-lu or be-e-lu 'lord', nu-nu or nu-u-nu 'fish'; so la and la-a 'not', ma-ha-za 'cities' and Se-la-Sa-a 'thirty', ki-i 'like, as', mal-ke 'princes' and mu-u'-di-e 'multitudes'. The verbal endings i (2 f. sg.), M, d, of the present, preterite, permansive and imperative, when standing at the end of a word, and therefore to all appearance unaccented, are never written plene: we may write ik-Su-du-u-ni but never anything but ik-su-du, ik-ka-lu, Hi-ku-nu. On the other hand the long vowels discussed in § 38 a, which are the result of contraction, are but rarely found written defectively at the end of a word, as in kus-si for ku-us-si-c 'of the throne'; ka-bi for ka-U-e 'speak' (Nimr. Ep, 48, 178). In shut syllables in the middle and end of words, % i, u are practically never ex- pressly indicated {U-im-iu 'fate', i-Um 'he appointed' ; 42 THE WRITTEN OHAKACTER : § 10. VOWELS-HOW WRITTEN. be-el-tu 'lady, mistress', i-be-el 'he ruled'; pu-ur-iu 'wild cow', i-du-uk 'he killed'); with a alone do we find side by side e. g. iam-tu and ta-a-am-tu 'sea', da- an and da-a-an perm, 'he is judge', etc. In both open and shut syllables in the beginning of a word {JVort- anlaut), the length of the vowel may be indicated in the same way if the breath (Hauchlaut) is expressed: of. 'a-a-ru 'go out' = 'arM, 'u-u-ru 'send' = 'MrM (^ urru) — cf. also tu-'a-a-mu=tu-d-mu 'twin'—. This method was, however, seldom employed. In general the in- dication of the breath was dispensed with at the beginning of a word {v. § 20), consequently all indi- cation of the length of the vowel had also to be dis- pensed with: thus a-B-pu 'conjurer', i-nu, e-nu 'eye', ii-ru 'nakedness'. It is worth noting, in particular, that the prefixing of the simple vowel sign, which takes place not unfrequently with shut syllables in the beginning of a word, does not necessarily imply that the vowel of the initial syllable is lorig. It is their etymology, not their orthography, that tells us that the vowel of the construct forms a-ar (from inf. dru st. lis), a-al 'city', i-in 'eye' is long, = «r, dl, in\ for in spite of such orthographical forms as i-iS-ta-lal 'he plundered' (VK 55, 43), a-a i-in-nen-na-a 'let not be oppressed', e-en-iu 'lady', u-us-ziz 'I set up', u-ul 'the highest', also 'not', and in spite of the fact that THE WBITTEN CHARACTEK: § 11. DOUBLED CONSONANTS. 43 Nebuchadnezzar and his successors were extremely fond of writing e-es-si-is 'anew' (adv.), e-ek-du 'stropg', e-ep-ti-ik 'I built', the initial vowel of all these words is short. Also in ki-a-am 'thus', ti-a-am-tu 'sea', the vowel sign a seems to be inserted far more for the purpose of emphasizing the breath or the hiatus, , than of indicating the length of the vowel. Strangely enough, d is the only long vowel for whose expression there is special provision in the Assyrian system of writing; for details see §§ 13 and 14. Very frequently a vovyel may be recognised as § 11. long from the fact of the following consonant being doubled. In spoken language the length of a vowel is often compensated for by such a sharpening of the immediately following consonant (cf. Hebr. Article •n = n; •» = sd\ D'^SS plur. of ps) ;• syllabic writing, moreover, was specially adapted to the reproduction of words according to their actual pronunciation. Hence ru-uk-ku 'far' = ruku^ ur-ru 'light' = wrw, Si- du-un-nu = ifiT^t, Lu-ud-du = "Vlb, kurbannu pi.)?, ba- ba-at-te 'the gates' = bdbdti, pa-nu-us-su 'his face', ta-ba-ah-hu Inf. 'sacrifice' (IV R 68, 33 a), i-na-ar-ru 'they subjugate'= marw, mu-ni-ih-ha Shams. Ill 29 and mu-ni-ha ibid. IV 23. Instances like issanundu = issanudu (§ 52) show that we have not to do here with a mere orthographical freak. 44 THE WRITTEN CHARACTER: § 12. THE SIGN lA. Considerable caution is necessary in dealing with the doubling of a consonant, in so far as such doubling is not demanded by the grammatical form, inasmuch as it may have its raison d'etre not only in the length of the preceding vowel, but also in the accen- tuation of the latter (§ 53); or it may be due simply to inaccuracy of the orthography, or to a desire to reproduce more exactly the ordinary pronunciation (§ 22). § 12. When the two vowels i and a come together, which happens most frequently when a genitive sing, takes the pron. suff. of 1. pers. sing, (i-a), they are generally expressed by a single sign ia (v. § 9 No. 38): cf.: ahi ta-lim-ia 'of my full brother' (VR 62 No. 1, 22. 26). When ia is joined to an ideogram, the i may at the same time do duty as a phonetic complement (§ 23): zer-ia 'of my family' (Beh. 3) = zeri-a. If, however, the i vowel is already expressed in writing, as, for example, in M-ti-ia 'of my house', then the i of ia is purely pleonastic; in other words, the sign ia repre- sents the simple a vowel. Perhaps this method of writing the pronominal suffix of the 1. pers. sing, is the cause of what,"^ at first sight, appears sufficiently striking, namely that the sign ia is in many other cases employed to express the vowel a, nothwith- standing the fact that — at least in the first of the forms given in § 9, No. 38 — it may easily be recog- nised as made up of i + a. Thus always after plural forms in e: ume-ia = ume'a 'my days'; cf. also ir- ba-'a-ia, var. ir-ba--a, certainly irba'd (erbad) 'forty'; THE WRITTEN OHARACTEE: § 13. LONG A. 45 re-ia 'shepherd' = re' -a (Tig. I 34) ; ka-ia-an (IV R 45, 42), perm, from 'jia, certainly = W«, like da-a-ri perm, from 111; ia-u and 2«-«m 'where?' the latter = dnu, Hebr. ]«, ia-um-ma 'any one' = a'umma. See also § 14 and cf. § 41. An exceptional position in the representation of § 13. the long vowels is that of d (v. end of § 10), inasmuch as for this vowel a special sign, a double «(?;.§ 9 No. 6), is usually employed, without, however, super- seding the methods discussed in § 10. Examples of initial d: fa-a-u=d-u name of the sign «; fa-a-Si=dsi 'for my part', pron. ; a-a-nu = dnu 'where?' (v. end of § 12). — Medial and final: ^ta-a-a-ra (V R 33, 11) 'compassion' (pronounce tdra) alongside of ta-a-ru (V R 21, 54a); ^ta-a-a-ar-tu 'return' (pron. tdrtu) alongside of ta-a-ar-tu, ta-ia-ar-tu, all = tdrtu st. cstr. ta-rat ; ^na-a-a-lu and na-a-lu i. e. ndlu name of an animal ; ^ka-a-a-nu 'firm', '^ka-a-a-ma-nu 'everlasting', ^da-a-a-nu 'judge' alongside of ka-ia-nu, ka-ia-ma-nu, da-ia-nu (pron. kdnu, kdmdnu, ddnu); fsa-a-a-i-du 'hunting', also sa-i-du, both = sd'idu, showing that da-a- a-i-ku 'killing', fern, da-a-a-ik-tu and similar forms are to be read daiku, daiktu\ '^ba-a-a-ar-tum part. Qal (same form as rdmtu, IV R 57, 46 a); fka-a-a-an perm. of 113, also ka-ia-an (§ 12); ^u-ka-a-a-an alongside of u-ka-a-an, u-ka-an all = ukdn 'he established'; fBa- 46 THE ■WRITTEN CHARACTER: § 13. LONS A. za-^ki-a-a-u, 'Hezekiah' (Lay. 61, 11); country of tiV«- ba-a-a-ti=Naba'dU (Wa?) and then Nabdti; the tribe- name Ha-a-a-ap-pa-a also Ha-ia-pa-a = Ha'dpd (Hebr. nS'iy), Hdpd; ^u-kd-a-a = ukd 'he waits' [u-ka-a-a-ki = ukdki 'he waits for thee', u-ka-a-a-u = ukd'u 'they wait') ; camels Sa^Su-na-a-a, i. e. ^und (Sunnd) se-re-Si- na 'with double back' (Lay. 98, L III) ; the river U-la- a-a and U-la-a = Uld (''b1^5); goddess Na-na-a-a and Na-na-a (even Na-nd) = Nana (Navaia) ; hence it would seem that Ma-da-a-a and Man-na-a-a^ which are the names of two countries and their inhabitants, should — nothwithstanding the Hebr. ii^, isti — be read Madd (Mddd) and Mannd (= original Mdddi, Manndil). Further, having regard to such orthographical forms as Za-za-a — cf.' also $al-sa-a-{a) Assurb. Sm. 130, 1 — nomina relationis ending originally in d'l, like ^i-du- un-na-a-a, Za-za-a-a, should be pronounced simply $idund, Zdzd. In the same way, perhaps, the plural forms in d with pron. sufiix of the 1. pers. siqg. e. g. : se-pa-a-a 'my feet' simply sepd\ at all events sSpai is a monstrosity, and uncontracted forms like sepd'a are at variance with the elsewhere usual fusion of two contiguous vowels (cf. §§ 38 and 47). Proper names like fApla-a-a 'my son', fSu-ma-a-a 'my name' were certainly pronounced Apia, Sumd, as is proved by the fact that they are often written Jp-la-a, Su-ma-a. THE WRITTEN CHARACTER: § 14. A-IA AND lA-A. § 15. E. 47 In all the words and forms denoted, by f, it seems to me im- possible to regard a-a as equivalent to ai, as is done by so many Assyriologists ; it appears to me, further, that even elsewhere there is not a single instance where it is necessary to read a-a as ai {v. especially § 31). Some of the words written with a-a far rather suggest a'a (a'a, a'a), though even in these cases contrac- tion to a must very soon have taken place. Since, according to § 12, the sign ia has very § 14. frequently the syllabic value «, we find not' only a-a but also a-ia and ia-a (even a-ia-a) written for a, apparently at hap-hazard. Examples of the anlaut: ia-a-bu (e. g. Assurn. I 28), even a-ia-a-lu (I]R 27 No. 2, 68), alongside of a-a-hu st. cstr. a-a-al) (pronounce dl Esarh. II 43), all = dbu 'enemy'; a-ia^ru and a-ru, both = dru 'child, offspring'; a-ia-si (Assurn. II 26) and ia-a-si alongside of a-a-Si, all = dsi 'as for me'; a-ia-um-ma (Shalm. Bal. V 3), a-a-iim-ma and ia-um-ma all = aumma 'any one'; ia-a-nu 'it is, or was, not' — dnu perm. ; ia-a-ri, also a-ar = dru, dr(i) 'forest' (n:^i); — inlaut: ta-ia-a-ru 'merciful' (I R 35 No. 2, 7) alongside of ta-a-a-ru (IV R 66, 42 a), = idru:, hence za-ia-a-ru must = z«rM (iT); — auslaut: Ar-ma-a-ia (Tig. V 47), also Ar-ma-a-a, both doubtless to be read simply Armd (v. end of § 13). The fact that the vowel is included with the con- § 15. sonant in the Assyrian characters is to the advantage of the otherwise complicated Assyro- Babylonian wedge-writing. This advantage is not impaired by 48 THE WRITTEN CHARACTEE: § 15. THE VOWEL E. the fact that about twelve signs admit two different vowels {a and i Nos. 26 Us, 108. 141. 159. 162. 182; a and u Nos. 102. 143; u and i Nos. 10. 101. 174. 193; cf. 199), and that four signs may be pronounced with as many as three different vowels («, i and u Nos. 7. 36. 161. 188), inasmuch as the grammatical form and variants scarcely ever leave us in doubt as to the proper selection. Still the provision made for distin- guishing graphically the e and i vowels is, we must admit, very unsatisfactory. There are, it is true, twelve signs (viz. those for e, le, te, me, ne, se, Se, ie; el, en, er, es), which were specially intended to repre- sent the e vowel, but in practice the same sign was used indiscriminately for syllables containing e and i. We advise the beginner to remember that all the syllabic signs given in the table with i, such as ki, pU, may also be pronounced with e (jke, pes), and that some of them, in fact, must be so pronounced (e. g. res No. 131). This lack of precision is less felt with long e, since the sign specially coined for e is pretty frequently added (v. § 10): the manifold variations M-lu, U-e-lu, ie-e-lu 'lord', or orthographical forms like ri-e-Su 'head', ri-e-mu 'grace', si-e-nu flock (sheep and goats), point assuredly to lelu, risu, remu, senu. It is quite otherwise, we regret to say, with short e, in- asmuch as the forms with this vowel, which almost THE WRITTEN CHAEAOTER: § 16 THE VOWEL tj. 49 always arises by modification from an original a, completely coincide in their graphic representation with those forms, of which i is the characteristic vowel. For this reason it is extremely difficult, or even impossible, to pronounce with certainty regarding a number of complex problems presented by Assyrian morphology, especially as regards the formation of the noun. For the transliteration of Assyrian we may lay it down as a rule, that the above twelve e signs should in all cases be transcribed with e, the cor- responding twelve i signs (i, hi, ti and so on), on the other hand, with i; with the other signs, e. g. U, ir, we may make the choice of the vowel (e or i) depend on the form of the word in each individual instance. For the reason of this misplaced parsimony as regards the indication of the vowel e, see the appendix to the orthography (§ 25), which we have devoted to the Sumerian prohlem. Of the two signs for m (§ 9 Nos. 4 and 5) the first § 16. is practically never used for the copula u («2), while the second practically never serves as a syllabic sign in the beginning of a word (an exception is found in Esarh. VI 24). The third sign for u {u No. 267) was originally ideographic. — The accents (mostly acute) over the vowels of certain signs, u, tu etc. are not meant to indicate the position of the tone or the quantity of the respective vowels; they merely serve, DeUtzsch, Aesyrian Grammar. 4 50 THE WRITTEN OHABACTEB : § 17 OF THE STLLABLE. in the transcription of Assyrian words, to distinguish a sign for a simple syllable from another and more fre- quently occurring sign for the same syllable. Besides u, cf. bu, pu Nos. 11. 70; •■ phonetic reproduction of the pronunciation as handed down by oral tradition. In pro- ductions of the higher, poetical style, above all, where great im- portance attaches to the finer shades of meaning, it was simply indispensable that the pronunciation of the original should be added in some such clearly intelligible fashion. It is becoming increasingly evident that the so-called bilingual texts are simply Semitic texts with a twofold orthography: the one form being the ancient and sacred ideographic writing of the priests, which, however skilfully conceived and cleverly elaborated, remained enigmatical to the last; the other, the ordinary syllabic writing. We are led to the same conclusion, with almost greater certainty by a consideration of the groups of ideograms. It is manifest § 25. THE INVENTION OP WEDSE-WEITINO. 67 that combinations of signs like SIG. DUB. SIG. DUB. BA. i. e. olothes-rending-olothes-rending' or LU. SAG. BI. DUL. LA. i. e. 'man-head-be-cover' cannot possibly be 'Sumerian words' for 'bitter mourning' {uhlu malU), 'mourning of a man' (amelu adir), but are purely ideographic and ingenious representations of the idea of mourning. Were these and the hundreds of ideogram- groups contained in the so-called vocabularies and bilingual texts real compound words , Sumerian must have been a tongue which was practically unable to express ideas and objects by a single word. Supposing, however, that we were to have recourse to the daring assertion that these groups of ideograms are the equi- valents of single Sumerian words with which we are still unac- quainted simply because the necessary glossaries are wanting, we should, when we came to draw the consequences, land ourselves in' a veritable quagmire of impossibilities. These groups of ideo- grams can be nothing more than groups of ideograms, more pre- cisely ideographic equivalents of Semitic words, conceived by Semites and the offspring of the Semitic mind. The symbolical reproduction of mourning by 'covering of the head', and of bitter mourning by 'complete rending of one's raiment' bears on the face of it its Semitic origin, and so it is with the groups of ideograms, one and all, — they are graphic equivalents of Semitic words, sometimes ingenious, sometimes clumsy, not unfrequently punning, but also, at times, with no meaning at all. The vocabularies — the subscriptions to which, by the way, do not breathe a whisper of any other tongue alongside of the Assyro-Babylonian — ^Hke the so-called bilingual texts serve the purposes not of comparative philology but of comparative orthography and comparative editing (nicht vergleichend-linguistische, sondern vergleichend- graphische, vergleichend-redactionelle Zwecke). The glosses, which in the vocabularies here and there accompany simple ideograms and groups of ideograms, and lists of which are found in syllabaries of the species S'' , are still in need of more thorough investigation as regards the manifold ends they serve; we know, however, that they contain, for the moat 6* 68 § 25. THE nrVENTION OP WEDOE-WKITING. part, the conventional readings of the ahove-mentioned signs and sign-groups, readings which are either identical with the Assyrian equivalent of the signs referred to, as given in the column to the right, or borrowed from one of its synonyms. Certain of these glosses are still obscure; others, a,s pisan 'reservoir', in particular, 'water-reservoir', are being proved, with ever increasing certainty, to be genuinely Semitic words. Moreover when the ideogram for sumu 'thirst', made up of 'mouth' and 'day, sun etc.', is accom- panied in VR 31 by the gloss im-ma, we may have been justified formerly in regarding it as a 'Sumerian' word, but now, when we read in Babylonian Semitic texts the words emmu 'hot', immii 'heat', and at once recognize them as derivatives of the Semitic stem d^n, it is clear that the gloss owes its existence to a synonym oi sumu, the genuine Semitic immu. In the same way, as om- acquaintance with the lexicon of Assyrian increases in extent and depth , the result wiU be the clearing up of all the other glosses that as yet remain obscure. Thus even the characteristic 'Sumer- ian' word dingir 'god' has been shown to be good Assyro-Semitio by the equation di-gi-ru-u = hi-li-bu-u = ilu, quite recently found in a vocabulary by Bezold! That the so-called Sumerian connected texts of volumes II, IV and T, the exorcisms, hymns etc. have, one and all, passed through Semitic hands, and from beginning to end present traces of Semitic influence, revision, transposition or whatever one may choose to call it, is an admission which has likewise been gaining ground for a considerable period and may at all events be put down as a support of the auti-Sumerian view. In fact , to admit the existence of a 'monkish' or 'dog Sumerian' swarming with compounds, phrases, arrangements of words etc., even transi- tions of meaning*) either common to all the Semitic dialects or *) The cases in which whole Semitic words with their ter- minations have passed over into 'Sumerian', as, e. g. za-ha-lam-a- ni 'their offering', and the still more repulsive oases in which the Semite, when writing 'Sumerian', mixed up two Semitic words with totally different meanings, airis 'humble' (io^) and airis 'in its place', are intentionally disregarded above. For here there § 25. THE INVENTION OF ■VfEDGE-WKmNO. 69 peculiar to tlie Semitic of Babylonia, be these the medium of the speech, poetry and writing of Semites or of Sumerians or of both, to admit this is, at bottom, the beginning of Anti-Sumerianism. For, not to speak of other impossibilities, this admission, on the usual hypothesis of the contemporaneous existence of the two peoples and tongues, leads to consequences that are absolutely alarming, for they mean that the Sumerian 'language' must simply cease and determine. Moreover it will never be possible for any one to draw a hard and fast line between 'dog Sumerian' and pure classical Sumerian; for the presumably purest Sumerian, that of the unilingual texts of the old kings of TJr , Larsam and Tello (LagaS) is 'dog Sumerian'. Apart altogether from such plays upon words as, for example, da-er 'lasting, eternal', which clearly betray their Semitic origin (st. ddru, 'last', part, dd'ir, dd'er), we meet at every turn, even in these texts, those Semitic forms of thought and speech, with which we are aU familiar from the Assyrian monuments and other sources: cf. in the royal titles such expressions as 'the called one of the true heart, object of the lifting-up of the eye, etc' of such and such a deity. How is it, finally, with the grammatical forms of Sumer- can be no question of any linguistic principle being at work: they are simply examples of most regrettable negligence and disregard of the old rigid ideographic principles, allied, in part, with thoughtlessness and ignorance. A similar declension from the old methods of writing appears also in the so-called 'dialectic' texts, in wljich the old ideograms are confused (e. g. tug 'be' for dug 'speak') and Semitic words and forms admitted with increasing recklessness (e. g. se-ib 'enclosure' from Assyr. stbu, syn. of Upittu,' Su-li-U = mlula IV E 20 No. 1, 15/16). As for the 'dia- lectic' phonetic changes between 'Aocadian' and 'Sumerian', the change of g and d, of w and s, of dug and zib, I regard them as simply impossible from the point of view of the physiology of sound; in all- probability, we must look for the explanation rather in Babylono-Semitic synonyms. The change of m and g, however, — which, by the way, must have been present in Sumerian as early as the time when the Semites 'borrowed' the cuneiform writing, cf. the syllabic values mi, mir, mal—seems to have its analogy in the Semitic tongue of Babylonia (v. phono- logy § 49, a, note). 70 §" 25. THE IKVENTION OF WEDGE-WKIirNG. ian, which still continue to be put forward, along with the . phonetic syllabic values, as the leading argument for the existence of a Sumerian language? These, too, are open to suspicion in many ways. At the very outset we are struck by the singular fact that in the midst of the most genuine Semitic Assyrian texts, where there can be no question of 'Sumerian' originals of any sort, we find such 'Sumerian word-forms' as dam-na 'his wife', al-tur 'he will be diminished', ni-ffal 'it will he' ba-bad 'he will die', na- an-bal-e 'let no one transgress'. Did the Semitic scribes of Baby- lonia and Assyria really go so far as to use full-fledged Sumerian words, with their formative elements, as ideograms to represent their own proper Semitic word-forms? Did they — to use a simple illustration — write: 'the master mourra\ wherein mowrj-a simply represented an ideogram for 'he will die'? Or is it not in- finitely simpler to regard orthographical forms like these as wholly and solely attempts at writing the ideographic equivalents of Semitic words? More important, however, than the above is the circumstance that Sumerian grammar reminds us so very often of Babylonian Semitic. 'Sumerian' employs the characteris- tically Semitic mechanism of the construct state, distinguishes precisely the same tenses as Assyrian, and has, in the Verb, a ^m- stem and a ta-an-siera. Its adverbial ending in es e. g. ul-le-es= elsis ifilses), zi-de-es=kenes corresponds exactly to the Assyrian, e.g. musis 'during the night', iamames 'heavenwards', dabu'es 'like a bear'. The correspondence is rendered all the more com- plete by the express statement in VB, 37, 57 — 59 that es or, as it is customary to say, the 'Sumerian' es is equivalent to i-na as well as to a-na and ki-nia. It is suspicious that the 'Sumerian' he, like the Assyrian lii (from TflAb 'wish, decide'), is not only the preca- tive particle, but like the Assyr. lH — lu, also signifies 'whether — or' (he-a — he-a), not to mention the use of he for the emphatic lu with preterite forms (VB, 62 No. 2). We have lists (cf. that published by Bertin in the Journal of the Eoy. Asiat. Socy. XVII, part ] .) in which the elements of so-called Sumerian inflexion are subjected to a most thorough analysis and set down as preforma- tives , informatives or afformatives e. g. ne and bi-i = ana Su'ati, § 25. THE INVENTION OF WEDGE-WEITINO. 71 bi-ne and ne-e=aUa Su'ati, hi-in and in=H Su'ati; i-ni-ni and mi-ni-ni — i-ni-e and mi-ni-e — i-ni-in and mi-ni-in = anahu atta — Su Su'ati Su'ati; in-na-ni-ni=anahu Su'ati Su'ati u anaku Su'aSum; mu=id'um SapliS etc. How very strange that the Baby- lonians should be thus familiar with the structure of the Sume- rian tongue in its minutest details! Were the Sumerians them- selves such thorough masters of their language that they could thus instruct the Semites in its deepest mysteries? Or did the Semites themselves discover all those significations hy comparative study of the Sumerian texts? It is infinitely more credible that lists like these were drawn up solely for the convenience of the scribes; they were meant to show what meaning was attached to the multifarious syllables and groups of syllables which were employed as the ideographic equivalents of the Semitic forms. There is at present no reason to doubt that what appear above to be elements in the composition of words will turn out to be ideograms artificially designed by the Semitic inventors of the wedge-writing. Here, as elsewhere, we may apply the- saying: dies diem docet. Bertin's list already proves this much, that in 'Samerian' words such as innanlal, haninlal 'he weighed it' (iSkulSu), nan, nin do not, as is universally assumed, correspond to the pronominal suff. Su, thus giving us, in 'Sumerian', an in- corporated pronoun, but rather that an-lal, in-lal are equivalent to iS-kul, and that inna and bani symbolize the object, which in Assyrian, as is well known , precedes the verb (= Assyr. Su'ati Su iSkul 'it, itself, he weighed'). In this wise another support of the 'Sumerian' theory falls away. I do not deny that, especially as regards these supposed Sumerian forms, there are enigmas still awaiting solution, but there is not one among them that can seriously affect the line of argument which we have hitherto pursued. The Semitic Babylonians will be found entirely justified when tiiey ascribed the invention of the art of writing to their god Nebo, and that besides the Cossseans, they never anywhere make the slightest mention of a third, a Sumero-Acoadian , people will in the long run be explained by the fact that such a people was never in existence. Phonology. A Vowels. I. Vowel Sounds. § 26. Assyrian has the following vowel sounds: a, i, u, e; a, i, u, e. Of diphthongs it has perhaps ai. § 27. Examples of short and long a, i, u (to which the beginner may mentally add the corresponding Hebrew words and forms): a: amtu 'maid', sarru 'king', kalldtu 'bride', nahlu and nahallu 'valley, wady', malkatu 'princess', $amsu 'sun', daltu 'folding-door', narkdbtu 'waggon, chariot', altur 'I wrote'; isbatu 'they siezed'; ahu 'brother', kanu 'reed'; taSrup 'thou didst burn'. i: ilu 'god', Mntu 'daughter', sillu 'shadow', par- zillu 'iron'; Hpru 'mission'; timdli 'yesterday', lihittu 'brick', imeru 'ass'. (For i^=ia, e. g. Urup 'he burned', i;. § 41 ; for i from an older a, e. g. Selibu 'fox' V. § 35). phonology: § 26—28. 73 u: mutu 'husband', ^umu 'name', ummu 'mother'; uznu 'ear', Urup 'he burned'; Urupie 'they burned', Purdt 'Euphrates', Ululu 'the month Elul'. d: sdsu 'moth', attd 'thou'; Id 'not', atdnu 'she-ass', aldku 'go', pdMdu 'superintending', bdmdti 'high places'. (For «=«', e. g. rddu 'storm'=ra'\, li?]* (z^); siibu 'finger' yasfii:, gXi.t, ii^l (Si), mpru 'claw' Xp^, JJ, i^ {h), ersitu 'earth' i^ix, jp'X \^if, (h). Of the two sibilants s and S, the former corresponds to the Hebrew o; the latter, sh, is also etymologically of three kinds: sa'dlu 'ask' bsiB, jUi, M-* (^i)) ^^^^ 'o^' 'Tl'li') jp^ V^ol (^2), karsu 'belly' WIS, J'}^: lai.= (^3). In Babylo- nian both 5 and ^ never, we may say, ceased to preserve their original pronunciation. This is best shown by the names of the months, which the Hebrews borrowed from the Babylonians during the exile: Writu I'llBn, Arafisdmna 'jliBti':!^ , Sabdtu t:3lD on the one hand, Ni-sa-an-nu pij, Si-ma-nu llip, Kis{i)- limu "ibps on the other. Cf. further Bel-Sar-usur n^Siaba (also Uten ''ftl!??) for the one sibilant, Sippar D''5':Bp, Sinuhallit t2^55p for the other, as also the Babylonian names of the winds which have found their way into the Aramaic of the Babylonian phonology: g 46 the sibilants. 107 Gemara: Mtu 'South' srjsiia and Sadu 'East' si'i©. (Also lysUJ, i^^=Sumer; here belong further the Aramaic ari©, i^Siffl, and perhaps the Hebrew- Aramaic qm). In accordance with the above, the Babylonians render the sh of a foreign tongue, as we should expect, by S: Kiiiu 'Ethiopia' like ©13, Ddri- dvus (ttJI^'i'l), KuraS (©"lis) = Pers. Ddrayavaush, K^ur'ush etc.; a foreign s, on the other hand, is rendered by s, cf. Babyl. AspaSina and Pers. Aspacand, Babyl. UStaspa and Pers. VHshtdspa etc. l^iJiB'pba is no exception to the rule, for this name in Babylonian may have been equally well BaldtaSu-usur as Baldtsu- usur; it seems, in any case, to be strongly influenced by the similarly pronounced ISi^lBbSl. A real exception, however, is a word which is repeatedly found in NebuchadnezsJar written hursanis 'mountain - like', while 'mountain, mountain range' is, in its original form, f}urSu. The exception has perhaps been caused by the coming together of two sibilants and the effort to obtain greater case of pronunciation by means of dissimilation. Compare, however, usannu (III R 43 col. Ill 21) instead and alongside of usannu (1 Mich. II 14), S?5F)ps<: 'North'=Babyl. iStdnu, and a few others. The two pairs Su'dlu ViXlB and ISfdr Tn'r\VS have heen in- tentionally disregarded, since their explanation as loan-words from Babylonian is hoth uncertain and improhahle. — The name of the Old Babylonian king Samsu-i-lu-na, the son of Hammurabi, 108 phonolost: § 46 the sieilaiits. and that of the Old Assyrian king Samsi-Bammdn (1 B 6 No. 1) render it probable that even in the earliest period the word for 'sun' fluctuated between SamSu, Samsu and samsu. In Assyrian, on the other hand, s has given up, more and more, its pronunciation as sh and has gradually become identical with s. For Assyrian words and forms, it is true, the historical orthography was faithfully adhered to (although after s and s had become identical in pronunciation, they could not fail to get mixed up in writing, cf. iShup 'he cast down' Tig. 11 39, ispunu Shalm. Ob. 21, nashuru 'a turning to' I R.35 No. 2, 7 for ishup, ispunu, nashuru; also askup and Ukupu Tig. VII 24. 22, and again isruka 'he gave' Assurn. II 26 for iSruk, and many others) but the employment of the signs containing s was confined to genuine Assyrian words, while the S of foreign words was rendered simply by s, it being, of course, pronounced like that letter. An Assyrian s, on the other hand, naturally appears as a simple s, when reproduced by foreigners, since these heard only an s. As illu- strations of the latter statement cf. Tukulti-pal-eSara ICSb&'rtin, Sarrukin liSIp, ASur-ah-iddina p'^niDX, saknu, 'viceroy' Dipac pi. ; of the former, DbTB1T| Ursa- limmu, fiyat Sa-me-ri-na, ni'nm AsdMu, yiSin A-u-si-'a, 'Ethiopia' iljiis Kusu, SaSank I^auufxi? Susinku and many others. In Hebr. njPlB-i'i (=Assyr. ral) sake 'chief phonology: § 46 the sibilants. 109 O'fficer') the V may be due to a mistaken popular etymology. In the same way the rendering of ASSur by 111SS is only an apparent exception, since the Hebrews' acquaintance with the name of the country must date from a period prior to the time when S was universally pronounced as s. Such a period may perhaps be found in the reigns of Tiglathpileser II and Sargon; the in the sadly disfigured name, of kinglpSiab© (Ass jr. Sulmdn-aSared) is to be explained partly by dissimilation, partly by the influence exer- cised by the name "ipsbS-nbaPl. In Assurbanipal's time the rule given above, that s, notwithstanding its being pronounced as s, must not be employed to render a foreign s, began to be set aside, and conse- quently we find in a few proper names like Pu-Si-ru 'Busiris', HininU (Ditl), Si-ia-a-u-tu, Pi-sa-an-fiu-ru, ffa-rsi-ia-e-su, in Assurbanipal's prism inscription an Egyptian s rendered by an Assyrian s, pronounced as s. Still the only correct rendering of a foreign s like that just given is found, for example, in the name of Sais, Egyptian Sau, Sai (with d), Assyr. Sa-a-a (with D). Phenomena such as these would be inconceivable, had the Assyrians, as some maintain, pronounced not only S as s, but also s as L That the name , of the moon-god in Assyro-Babylonian was Sin (with D), not Sin, is a fact which nothing can alter; the rendering 110 phonology: § 47 the breath. of the name Sin-ahe-erba by S^lrtb proves, accordingly, that the Assyr. s, like the Babyl. s, was never pronounced and never heard otherwise than as s. So long as no Assyrian word can be pointed out, the s (d) of which is reproduced in a foreign tongue by s (©), we must maintain that in the pronunciation of Assyrian s and s, we have to do, not with an inter- change of sounds (Lautverschiebutiff), but simply with a 'onesided softening of the broad sibilant sh to s' — for which analogies are not wanting in the other languages of the Semitic group. For the rendering of b in Assyrian of. on the one hand li2i!:=)Saw?rM (III B, 5 No. 6, 45), on the other piUB'n Di-ma-ai-M (I R 35 No. 1, 15. 21). Yice versa cf. lia(X)>l?i and especially Qi'nbS, the inhabitants of the land of Kaidu. II. Phonetic Changes affecting the Consonants. § 47. The Breath. When ' closes a syllable, either (1) it quiesces in the vowel which precedes it, lengthening this vowel when it is short, e. g. zi-i-bu i. e. zibu ^woW^zfbu, muru 'young animal, esp. io3X\^mu'ru ^^, nddu 'exalted'=na'(?M, na'idu, nikul 'we ate', ^uhuzii, 'cause to take', ndmuru (Inf. Nif.) 'be seen'= na'muru — for other examples of a=d (and then=e), and for ' at the end of a word {Wortauslaut) v. § 32 phonolost: § 47 the breath. Ill P and y — , or (2) it is assimilated to the consonant which, follows it: allik 'I went'=ff'//A-; orthographical varieties like a-lik are to be considered in the light of § 22. Nevertheless, instances are by no means rare in which the breath has been retained: cf. mu'du 'multitude', MSu and M^u 'wicked', bu'Mnu and huMnu 'bad smell', na'butu 'flee', ibd' 'he comes', etc. When ' follows a syllable ending with a consonant, it is mostly assimilated to the consonant which pre- cedes it, and should this doubling of the consonant be dispensed with, the preceding vowel is lengthened: labbu ^li6n'=lab'u, hittu 'sin', nibu ^num.hev'=mbbu= niVu, ziru 'seed'=^«rM, zarru, zaru {v. § 33); inna- mir 'he was seen', innabit 'he Q.ed^=inamir, in'abit (pret. Nif.). Nevertheless instances are found, especi- ally in the conjugation of verbs primae Xi ji where the has been preserved: ifal, tr'ub (cf. §20), im'id 'he, it increased' alongside of i-mi-du, Mam'ida 'may he increase'. Between two «-vowels, ' either maintains itself or is dropped, the result, in the latter case, being the contraction of the two vowels: md'adu 'much', la'abu 'flame', sclciIu 'ask' and mddu 'much', ma-du i. e. mddu 'be much' cf. also ramu 'love'. ' maintains itself, as a matter of course, in cases like ri'd^u. 'vermin', mu'dru, ba'uliu 'subjects'; but it does the same also in na'id 112 phonology: § 48 b, d, and t. 'he is exalted' and rd'imu 'loving', for example, so long as the i is not syncopated. Loss of ' and subsequent contraction seem to occur in ruku 'distant' =rffl'MAM, rauku. We should naturally expect the virtually doubled or sharpened ' to be particularly persistent, and the Piel forms of verbs mediae Xi.2) such as umdUr^ muhiru, mula'it, confirm our expectations. It is hard, notwithstanding, to decide whether buuru 'catch, hunt', even in cases where it is written not bu-'-u-ru or bu-'u-ru, but bu-u-ru, is to be read as bu'uru or, giving up the ', as burn, the reason being that along- side of uma'irdni 'he sent me' we also meet with forms like u-ma-ra-an-ni (V R 34 col. Ill 1). For the con- traction of fdsas and similar forms to iHaS v. § 38, b. For the loss of ' in the aniaut, e. g. in timdii 'yester- day' V. § 39, and for the same in the auslaut in conse- quence of the shortening of the vowel, in which ' quiesces, e. g. nasi, peti (form J^Ij from SillJS, VijrSi) V. ibid. § 48. &, d and t. The labial b readily assimilates itself to the OT of a following ma; the assimilation is esp. frequent in erumma 'I entered and' instead and along- side of irub-ma. Cf. also u-Si-im-ma 'he dwelt and' alongside of u-sib-ma (Senhb. V 4) and pres. uS-Sab-ma (K. 4350 col. I 6. 9). No argument, however, may be drawn from these in favour of reading b as v, and m phonoloqy: § 49 nasals. 113 as V (as e. g. Srumma), in view of other cases of a consonant being assimilated to the m of the copula ma, as e. g. M/cumma=liSkunma (v. § 49, b). Among the dentals, t of the Ifteal and Iftaal is assimilated to a foregoing z or s: iz-zak-kar 'he speaks', assabat 'I, he siezed'; on orthographical varieties like a-sa-bat, a-sab-ta cf. § 22. For the assimilation of the same letter to a foregoing s, see § 51, 2. — Vowelless d, further, is assimilated to a following t, e. g. ma-at-tu fern, of ma'adu 'much'; also to a following S, when the latter is the third radical of a stem containing three consonants: eUu 'new' (=edsu, ediSu), seHu 'sixth' (^sedsu, §dd(i)Su). — After k the t of the reflexive stems becomes t, e. g.akterib 'I approached', after g it becomes d, e. g. agdamar 'I complete'. Also after m and n it is frequently softened to d, e. g. am- dahis 'I fought', umdaSir 'he quitted', amdahar 'I received', but compare with these amtahar (for attahar V. § 49, a) imtalik etc. The same change is presented by the t of the feminine after m and n: tdmtu and (always, it would seem, in ordinary pronunciation) idmdu 'sea', sinundu 'swallow' etc. Nasals, a) m. Of the nasals, the labial nasal m, § 49. before an immediately following dental, passes into the dental n always, we are sure, in pronunciation, and in most cases in writing as well: cf. mundahse DelltzBch, Aasyrian Grammar. 8 114 phohologt: §.49 nasals. [-wariiors' =mumdah(i)se, sindu 'yoke (of oxen), team' (Khors. 124, 153S), sandu (V R 35, lG)=samdu, nakamtu and nakantu 'treasure', hantu 'swift, active' for hamtu etc. Frequently also before a following s or s: unsu 'want' alongside of umsu, hansd 'fifty', i-ri-en-iu 'he presented to him' (III R 43 col. I 13, DXg"!); occasionally the n, which arises in this way from m, is afterwards assimilated to s: su-un-su 'his name', and then into Jms^m, hdnSu and IkdHu 'fifth'. Cf. also na{m)zidti (Assurn. II 67). This transition of m to « must be assumed as an intermediate stage in at{t)ahar 'I received' (Assurn. II 102. Shalm. Ob. 120). m also passes into an n before k: hence dumku and dunku 'favour'; emku and enku 'wise'; cf. also ikkut ffi, when pronounced as v^ seem§ to have been completely dropped in the younger Babylonian texts whenever it occurs in the middle of a word (Maui) between two vowels: thus we find the form usalrhd, usalvd 'I, he caused to hem in' written u-la-al-va-am and u-la-al-am (V R 34 col. I 34. 26); cf. also u-lat- ■vi-ih and u-sai-ih 'he caused to sieze' (V R 65, 5b, st. iamdJiu), iur-i-ni 'cypress' (2. 4b) in place of the usual surmeni, surndni (intermediate form ^urvini, surfini), na-i-ri 'panther' (V R 46, 43b) for namiri^ and a few other forms. Observe also Du'uzu, Duzu (=Duvuzu,. phonolooy: § 49 nasals. 115 Davvuzu'i) in its relation to TIBn (and zu-u'-ri-Su 'his body' III R 43 col. IV 16 for zu-um-ri-su 1 Mich. IV 6). When, on the other hand, a t; is found where the etymology of the word does not lead us to expect it, as e. g. in u-ka-va-an-ni 'he waited for me' (V R 65, 27 a) alongside of u-ga-a-an-ni (V R 63, 28 a) i. e. uka'anni, and especially in fydmiru, Tidviru 'wooer, bridegroom, husband' (cf. e. g. ha-me-ir IV R 27, 2a, ha-mir Descent rev. 47) in place and alongside of ^d'iru (st. Tin 'see, choose', as Haupt rightly assumes, t;. V R 50, 60 a), in this case Haupt regards the v as a secondary development from the intervening spiritus lenis. Or was it the case that the signs ma, mi, mir (va, vi, vir) etc. were simply employed for 'a, 'i, 'ir, just as the sign mur (vur, § 9 No. 188), for example, was on occasion used for ur? This would, to a certain extent, be a parallel to the use of i-a for a. In the Babylonian writing an interchange has heen observed between m and g, on which scant light has yet been thrown. Thus liursam 'mountain range', for example, was ideographically rendered by hur-iag, and, vice versa, halaku (halagu) 'perish' by ha-lam-ma (Haupt, ASKT 181, XII), sahluktu 'destruction' by Sa- ha-lam-ma (see for the latter ideogram III B 60, 71, 65, 4. 22b). The name 125^13, also, whose identity with Sumer still remains probable, at least, seems to point to the conclusion that the Semitic Babylonians, in certain cases, pronounced m as ng or- — without the nasal tone — as g: in writing they kept to the historical or- thography Sumer, but the Hebrews heard Sunger. 8* 116 PHONOLOaT: § 49 NASALS. b) n. The dental nasal n, when it is vowelless, readily assimilates itself to the immediately following consonant. Such is always the case with the n of the Nifal and Ittafal, e. g. iUakin and ittaSkan 'it was made' ; and almost always with the n of verbs primse :, hence iisuk 'he bit', iUu 'they took', attaM 'I named', madattu, mandattu^=mandantu 'tribute' (in the Shafel we also find usansir 'I set a watch', usanhit 'I made to shine'; but cf. im-ii 'he spread abroad' on the one hand, usaUi 'he caused to carry' on the other). As third radical, n is assimilated in mandattu, libittu 'brick', sukuttu 'rubbish, stufif'. From other cases of the assimilation of n we would single out the follow- ing: ni-hi-im-ma 'may he cast down' (viz. his counte- nance) =ft7&m-»«« (V R 56, 55), lis-kum-ma 'may she do &nA'=liskunma (III E 43 col. IV 17. 1 Mich. IV 7), but al-bi-in-ma (V R 66 col. I 11), az-nun-ma (V R 62 No. 1, 13). The converse, viz. progressive assimi- lation, is illustrated by the name of the moon-god Nannaru=Nanmaru (Haupt). Ar re'i=an re'i 'to the shepherd' is read in Pinches' Texts p. 15 No. 4, 9. Assimilation of n after a long vowel is found in um- mdtu'=ummdntu, fem. of ummdnu 'army', Utdtu^istdntu, fern, of istdn, Uten 'one' and a few others. Before b, n passes into m in all cases in pronun- ciation and in many cases in writing as well: cf. phonology: § 49 nasals. 117 however inbu 'fruit', but imbubu 'flute' (st. aaj). The same change takes place before k: Sumkuru 'estrange' and 'make keen' (the glance, v. E. M. II, 339, 1. 6), uSamkir (st. 1S5), and even before dentals and nasals (v. § 52). It is to be carefully noted, however, that the Assyrian script has not in all cases developed two signs for compound syllables ending in m and n (e. g. dam and dan) but has in very many cases been content with a single sign {v. § 9 Nos. 148. 206 and of. 138; No. 182, also, has the two values rim and rin, No. 196 Ian and bam; special signs for han (in addition to that for ^am), Ian, nan, ran, San, tun (v. p. 138), mam, mim etc., not met with as yet, in all probability never existed) — For this reason, we need have no scruples in reading sum, Sam etc. as Sun, san. In the imperative Qal of verbs primse j, n is merged in the spiritus lenis, hence usur 'protect', isi 'lift up', idin 'give thou'; also in the infins. of the Ifteal itpusu (=nitpusu), itanbutu, itanpuJiu (=nitdbutu, nitdpuhu) and Iftaal (?): itappusu; the same applies to the characteristic (Nifal) n in the infinitives of the Ittafal (Intafal): itaplusu 'see' (=mtaplusu), iiaktumu 'faint (=nitaktumu), etc. For the employment of m and n as substitutes for a doubled consonant, or one sharpened by the tone, see § 52. 118 phonology: § 50 liquids. § 51 sibilants. §50. Liquids. For the interchange of r and /, when both go back to an original sibilant, ?;. § 51. Assimi- lation of r to the following consonant is nowhere found: for this, if for no other, reason hattu 'staff', annabu 'hare' cannot be derived from older forms such as hartu, arnabu. That words like kakkaru 'the ground' are no proof of the assimilation of r is shown in § 61, 1 (p. 144). § 51. Sibilants. 1) When immediately preceded by a vowelless dental or sibilant, the s of the pronominal suffixes is always changed to s, hence mdt-su 'his country' (contrasted with ma^a-sw), as-ba(-su; thereon the dental is frequently, the sibilant always, assimi- lated to the s, after which both dental and sibilant, it would seem, are entirely dropped in writing (for the accent v. § 53, a) : hence §al-la-su-nu 'their spoil' (Khors. 47) from and alongside of Sal-lat-su-nu (Khors. 48), kak-ka-su 'his head' (Esarh. I 18), ka-a-su 'his hand' (=kdssu, kdtsu, kdt-su), karassu 'his body' (from karSu), murussu 'his sickness' (mursu), izussu 'he parted it' (=izuz-su), u-sak-m(-is)-su-nu-ti 'I sub- dued them' (tB33), u-lab-hi-su-nu-ti 'I clothed them, Id uS-har-ma-si 'he shall not destroy it, (the palace)' (I R 27 No. 2, 39 u^oin). Exceptions like ap-pa-lis-Su (Assurb. Sm. 290, 55), ar-ku-us-su (V R 8, 12) or M- rit-su-nu (II R 65 No. 1 obv. 3 a) are very rare, and phonology: § 51 sibilants. 119 in Assyr. texts may be explained by the pronunciation which we find in later times of s as s, so that they are quite as reprehensible as the rendering in Assur- banipal of a foreign s by S, (v. § 46 p. 109). Compare, however, in Babylonian ussabbit-sunutu (Beh.87), Msdt- Sunu 'their presents' (V K 33 col. V 46). 2) The radical letter S, which precedes the t of the reflexive forms Ifteal and Iftaal, and also the S of the causative form which precedes the t of the Ishtafal are able to maintain themselves in very many cases (apart from the change to I): astakan {altakan), ultebila etc. There is, however, a marked tendency on the part of this st, especially in the language of everyday life, to pass into ss or s: hence in both Babylonian and Assyrian letters the frequent forms assapar, asapra 'I sent', isaparuni 'they sent', ussibila 'I sent for'; of. usamris (III R 4 No. 4, 41). Among the longer historical texts, the only one that shows a special fondness for these forms is the great inscrip- tion of Assurnazirpal, which is otherwise remarkable for its peculiarities (because reflecting the language of the people?): asakan 'I made' (Assurn. Ill 2 and often), asarap 'I burned' (II 21) etc., etc. 3) Before an immediately following dental, Assy- rian sibilants very frequently become / (cf. vilta in certain Italian dialects, e. g. that of Pisa, alongside 120 phonology: § 51 bieilahts. of vista), hence selalti 'three', hamilti 'five', rapaltu=ra- pastu, fern, of rapsu, maltitu 'drink' from and alongside of mastitu, altur 'I wrote' (Assurn. I 69) from and alongside of astur (Esarh. Ill 48), altanan 'I fought' (Tig. I 55. Ill 77, pffi), manzalti 'stand, position' (V R 2, 43), eldu and es{a)du 'harvest'; tultesera=tusteSera 'thou governest' (IV E 67, 12b). The same change occurs. when two different sibilants come together: ulziz 'I set up' from and alongside of usziz {=usiziz), alsi 'I spoke, called'=«ss2. From iltdnu 'North' in the Assyr. vocabulary II R 29, 2h, contrasted with the talmudic XDnpiil (Babyl. istdnu), from the Babyl. kuS- tdru 'tent' (V R 35, 29), in Assyr. always kultdru, and especially from the name for Chaldaea, hitherto found only in Assyr. texts, viz. Kaldu contrasted with the Hebr. D^'lteS, which presupposes the Babyl. Kaidu, we might be tempted to conclude that this phonetic interchange of s and I was peculiarly Assyrian; such, however, is not the case, for forms exhibiting this particular change occur in the later Babylonian period at least, e. g. in the texts of Nebuchadnezzar. It is long since it was inferred from the Babylonian VraUu, Assyr. Urartu (li'TlS) that an r formed an intermediate stage between I and /; since then other examples of this sort have been found, especially by Pinches: thus even in the same (Neo-Babylonian) phonology: § 52 compensation for doublikg. 121 text IV R 15 we find iSdudu (1. 5) alongside of irdudu (1. 10), an Assyr. duplicate having in both passages iSdudu. Cf. the name of the plant maS-ta-kal (?), mar-ta-kal and mal-ia-kal. As a substitute for the doubling of a con- § 52. sonant, characteristic of the stem or of the inflexion of a word, as well as for the sharpening of a consonant occasioned by the tone, a nasal sound is often given to the vowel preceding the consonant in question: sumbu 'freight waggon, cart'=SM6&M; numbu 'scream, howr=wM&&M, 'hambakuku {^=habl)akuku), Amkarruna 'Ekron' (lilfe?) ; inamdin, inambi, ittanamzaz, ittanamdi (all written with nam^ for which comp. p. 117) from and alongside of inddin, indbi, ittandzaz, itianddi; ittanbit and ittandnbit (I 3) 'he shone' (=ittdbit, ittand- bit), inf. itanbuiu (^nitdbutu), etanamdaru (I 3) 'they were afraid' {=itianddaru or etanddarii); innam- daru, innandaru (IV 1) 'they rage', ittanamdar alongside of ittanddar (IV 3) 'it rages'; issanundu {=issanuddu), astamdih, inf. sitamduhu {^aStddih, Mtd- duhu). For the duplicate forms ndduru, ndhuzu and nanduru, nanfiuzu, for ittananmar 'it is found' (IV 3 =ittandmar, ittand'mar) and other cases consult § 11- There is no instance of the doubling being resolved by means of r. 122 phonology: § 53 place of the tone. Another substitute for the doubling of a consonant is the lengthening of the preceding vowel: note as examples suhu 'cart' (=suibu) and the cases mentioned in §§ 33 and' 41, b, zeru 'seed' (=zdru, zarru), ztmu (=zim.mu, zimiu) etc. {uSdziz, uSeziz § 101 may also come in here as 'bemg=uSazziz, usanziz); for a similar case of compensation for the sharpening of a consonant cf. the forms with enclitic ma mentioned in § 53, d. § 53. By ^^y of appendix we may here add a few remarks as to the place of the t o n e in Assyrian words. a) There can scarcely be any doubt that in words like kdrdu, sdrratu, epiissu ('ei feci'), muMkUd, musdksidu, uttdkkar, uUdklil, tuSahJidssi, and in those like abubu, nakruii, ime're, iksuduni, narkabd'ti, idukit ni, uSamsi- kuni, iksudsunuti the position of the principal tone or stress is really as indicated above. Forms like ulabMssu 'I clothed him' (=ulailis-su), even when written u-lah-hi-su must, according to the above, be accented ulabMsu, ulabbissu. The sharpening of the immediately following consonant, however, which never fails to accompany the accenting of a short vowel, and the fact already repeatedly referred to, that the Assyrian writing adapts itself to the every- day pronunciation are the causes of certain other phenomena. Thus as regards the present of the Qal, the fact that in the great majority of cases the second radical is written twice, as in iSakkal, iballut, inaddin, ilabbin, Uemmu, shows without a doubt that the characteristic «-vowel of these presents was accented. PHONOLOGY-: § 53 PLACE OF THE TONE. 123 The same is proved for the syllable ta of the verbal stem I 2, and for the syllable na of the verbal stem I 3, both in the preterite and in the present, by the extremely frequent forms which are vrritten thus: ikakkan, aUakkan, iltak{k)anu (Assurn. I 30), attak{k)i 'I sacrificed' (Tig. VIII 10), amdah{h)is, mundahhise 'warriors', iktarrdbu 'they blessed', iptallahu 'they were afraid', muitabUl 'leading, ruling', itaUuku 'go to and fro', cf. altamdih, istamdahu 'they drew' (§ 52); — ihta- nabbata 'plundered', iStanappara 'he sent', imta- nallu, cf. ittanamdi (§ 52). The forms akterib, ilteki {§ 34, a) as alternatives of aktarib etc. render it also in a high degree probable that the syllable to in I 2 was likewise accented. That in the present Nifal, further, the tone rested on the second syllable is shown by orthographical forms like innakku 'there are poured out', innemmedu (rel.) 'there is hidden', and especially innamdaru, innamdu 'they are founded' (V R 64, 27 b), cf. § 52. In the continuous transliteration of Assy- rian words, forms with the consonant written twice should be given as in isakkal or isdkal, so that isakal might signify that the second radical was written only once. b) When a consonant is uniformly written once, on the other hand, we may with certainty infer that the preceding short vowel is unaccented. Consequently, 124 phonolog-t: § 53 place op the tone. in the case of verbal and nominal forms like iskulu (rel.), Ukulu, iskuM ; hatanu 'son-in-law', laUru 'old' etc., this much at least is certain that the tone did not rest on the middle syllable. That the tone, moreover, did not rest on the last syllable we know from the fact already mentioned in § 10 regarding the verbal forms, that the length of the verbal endings i, u, a is never expressly denoted in writing, when they are part of the last syllable of the word: even with verbs tertise i we find, though rarely, forms like ib-nu 'they built', an impossible orthography were the accentu- ation ibnu. The same is proved with still greater certainty by the contraction of forms originally ending in e and i, such as iSme, iSmi, ibne, ibni to ismi, ibm etc. We should therefore read: ikiud, tdksud, tdksudi, ik- sudu etc. c) Special attention must in future be paid to those cases in which the verbal forms now mentioned are nevertheless — in contrast to the course pursued in the overwhelming majority of cases — written with the third radical doubled. As against the supposition that these are simply cases of inaccurate and defective orthography (v. § 22) we have, in the first place, the circumstance that forms written in this way are, after all, not so rare as one might suppose, and, in the second place, the fact that, in some cases at least, the phonology: § 53 place op the tone. 125 sentence accent is seen to be, without a doubt, the determining factor. I shall confine myself here to a few examples, in which the verbal forms in dispute are distinctly brought out by means of spaced type. 'An art, which among the kings, my fathers, none ihuzzu had learned' (end of sentence). 'District and border iskunnu did they establish' (end of section II R 65 obv. col. I 23); 'such an one iskunnu they set up to bear rule over them' (end of sentence, ibid, col. II 32, supplement). 'Whirlwind and hurricane isabbannu (end of sent., Nimr. Ep. XI, 122)'; 'what I tell them, ipp uHd they do' (NR 24) ; 'm? i llikku 'they have not come' (end of sent. K. 831 obv. 7); 'in the evening uSaznan(n)u Samutu kibdti' (Nimr. Ep. IX, 83); 'on the street ittanamzazzu lu-nu they tread' (IV R 2, 17b); immalillu, ittanahlallu (end of sent. IV R 15, 38, 40a). Examples are often found in the contract tablets: uSzizzu (Strass. II 13, 6); 'till the creditor A:«s^« iSallimmu is paid in fuir(Str. I. 118, 11), inamdinnu 'they must pay', and many others — all of them pausal forms. Also at the end of relative clauses: 'their landmarks which ibtillu had been removed' (Khors. 136); 'where my father ipkiddusu had appointed him' (Assurb. Sm. 46, 62); 'Auramazda who iddinnu hath created this earth (or these heavens etc.)', in D, 2f., for example; 'what I 126 phonology: § 53 place oi?, the tone. epussu have done here, and what I ^j9 m s sm in another land, all that I epuHu^ (E, 16 — 18); 'what I ipussu and what my father ipussu^ (D, 14. 19. C, a, llf. C, b, 21/23). Cf. also iskunna Assurn. Ill 110. This orthography is not unfrequently found in the second of two verbs joined by ma: 'Assurbanipal, on whom Nebo and Tasmet have bestowed (iSrukus) great intelligence, ihuzzu enu namirtum hath received a clear-seeing eye' (often in the subscriptions to tablets) ; ikbusuma usaknissu sSpuswi 'they trod (them) down and made them submissive' (Esarh. IV 36); 'Sargon, who brought the king to his city of Assur and Muski emiddu apsdniu (Lay. 33, 11). Cf. also I R 49 col. IV 6. Attention is even called to the position of the tone by lengthening the vowel in place of sharpening the immediately following consonant: cf. u-si-i-bu K. 13 (IV R 52 Nr. 2) 1. 6; and ul-te- zi-i-iil (Assurb. Sm. 293, a c), also U-i-li (IV R 5, 39b)? Of permansive forms the following demand a place here: 'Istar isdta Ut-bu-sat melamme na-sa- a-ta (var. nasat) was clad with fire, wore (a mantle of) radiance' (V R 9, 80) where nasdta seems to be simply the equivalent of naMta. From the contract tablets cf. the phrase Uten bu-ud sani nasi, for which we find also ?ia-a-si, na-as-si, fem. na-Sa-a-ta. In the same way is explained the phrase often met phouoloby: § 53 place of. the tone. 127 with in the subscriptions of tablets satirma la-a-ri (IV R 16, 67b). d) The enclitic particle ma^ including both the copula ma and the ma of emphasis, draws the tone to the syllable immediately preceding it ; vowels originally long then re-appear, though often enough it is only to disappear once more in the sharpening of the m of ma: cf. on the one hand ma-ti-e-ma, ak-ri-e-ma st. Slip (Sarg. Bull Inscr. 99), ap-te-e-ma (Senhb. I 27), iS-me-e-ma (often), al-me-e-ma (V R 3, 127), adkema, ahrema, aS-te--e-ma (often), ah-ri-e-ma (Neb. Senk. II 3 and often), u-mas-si-i-ma (Sarg. Cyl. 46), on the other hand Sanumma 'some other one' {=sanu-ma), ilamma 'he came up and' {=ila-ma) ; vowels originally short remain, the tone, of course, causing at the same time the sharpening of the m of ma, cf. ameluiumma (Nimr. Ep. XI, 182) ilUkamma 'he went and', ikkisu- nimma 'he refused and'; in some cases, however, the lengthening of the short vowel is substituted for the sharpening of the m (cf. § 52 note). Thus in mi-tu-ti- i-ma (IV R 67 No. 2, 60b), i-ha-ru-{u-)-ma 'he went out and' (rel., Sarg. Cyl. 21), 'when that house i-lal- M-ru-{u-)ma shall become old and', 'whoever u-ma-a- ru-u-ma (III R 43 col. I 32) will send a friend and', alongside of u-ma-'-a-ru-ma. — In many cases we may be in doubt as to whether the length of the vowel. 128 phonologt: § 53 place of the tone. preceding the enclitic ma, is to be explained by the first or by the second of the methods now exemplified. This applies, for instance, to the verbal suffix of the 3. pers. m. sing., which, united to ma, is often written summa or suma; of. Uskunsumma 'may he appoint unto him' (V R 56, 43); ar-H-su-u-ma (V R 3, 20), tam-nu- su-u-ma (V R 3, 7), liskipu-su-u-ma IV R 6, 68 a. 63, 55 a): does the original long vowel of su make its appearance here? And how is it with sarri eni-ia-a- ma 'of my lord the king' (K. 823 obv. 5 etc.), Sumi-ia- a-ma (also ^umi-a-md) 'of my name'? And how with kaldma 'all together' (declined ka-la-mu, gen. ka-la- a-mi Nimr. Ep. 1, 4)? How far it may be safe to draw conclusions of a more general nature in regard to the position of the tone from orthographical forms like ina bi-ri-in-ni 'between us' (V K 1, 126), is a question which, on the one hand, it is difficult to decide, and which, on the other hand, it is better to reserve for the future (cf. § 74). In general, neither nominal nor verbal suffixes draw the accent of the word to the syllable immediately preceding tliem: kin-na-as- $u gabbi 'his whole family' (IV B 62 No. 2, 8) like ab-bi-e-S'U, 'I addressed him' (V R 64 col. Ill H) is evidently influenced by the sentence accent. It would be of special importance if we could infer from nam-kur-ri-su-nu (e. g. Tig. Ill 3) contrasted with na- am-ku-rum (II R 47, 49 d), that the stress could not, in Assyrian, fall upon the fifth syllable from the end of a word (e. g. ndm- kuriSunu) as is possible in Arabic, but that, in cases like these, the stress was laid upon the syllable nearer the end of the word. MOfePHOLOG^Y. Merely mentioning in passing the only inter- § 54. j actions as yet met with, viz. the exclamations of pain a-a (doubtless=<2) and m'«, we proceed at once to the pronominal stems, and to the pronouns developed therefrom. The former consist either of the vowels d and u simply, or of one of the conso- nants t, n, k, ff, s, I, m, with a short or a long vowel, The latter are to be learned from the paradigms A, 1 — 6; §§ 55 — 60 are meant to be nothing more than ob- servations supplementary to the paradigms. A. The Pronoun. * 1. Independent personal pronouns: a) To §55. express the nominative: sing. 1. c. andku.—2. m. atta; sometimes also used for the fem., e. g. lu asSati atta 'thou art not my wife' (V R 25, 10 b). Thd form dt-tam (IV R 20 No. 3, 18) must be explained as atta with ma {ni) added for emphasis. — 2. f. On the form at-ii-e (IV R 57, 45— 54 b) see top of p. 78. Plur. 1. c. Delitzach, Assyrian Grammar. 9 130 MORPHOLOGY r § 55 INDEPENDENT PERS. PRONOUNS. Notice the personal name Istu-Eammdn-a-ni-nu (Var. ni-ni) C^ 233; ni-i-ni (IV R 53 No. 1, 40). 2. m. at-tu-nu, e. g. lY R 56, 47 a. — For the rare cases where andku and attunu are used in place of the verbal suffix to express the datiye (without special emphasis), see syntax § 135. Tor Su, It, Sunu used as adjectives v. § 57, a. — u (no gender) 'he, it', with emphatic md 'that very man or thing', c. g. ina Satti u-ma 'in that very year' (Senhb. Bav. 34) is very frequently used to indicate the repetition of one or more preceding words (note Neh.III50 where um-ma is written). This meaning — corresponding to our 'ditto' — we also find in su, iuma, especially in the vocahu- laries. Perhaps, too, in the oft recurring phrase ina ume-Su-ma 'on that very day', the su is not the pronominal suff. but the ex- pression should rather be read ina ume suma, analogous to the above-mentioned ina satti lima. For the ideogram of H (H-ma), see the table of characters No. 268 ; for further details, see Dioty. No. 103. 6) To express the genitive and accusa- tive. Sing. 1. c. On the reading oiia-a-H^ a-a-h etc. as d§i, dti (from idsi, idti § 41, b) v. §§ 13 and 14; ia- a-tu written ia-a-pi (v. table of characters No 69) Assurb. Sm, 37, 9. — 2. m. and f. are completely iden- tical: kdti, kdSi; in the 3 m. and f., also, no distinction of gender is implied in the final ^a, si of sdsa and idH, in contrast to Msu — the masculine forms kd^a and dU alone forbidding such a supposition. The fact is, as we learn from the forms of 1. c. plur. occurring in Bertin's list (v. p. 70), viz: ni-ia-ti, ni-ia-sim, a-na ni- mokpbology; § 65 independent pees, peonouns, 131 a-Sim, that all these' pronouns dSi, JcdH q%c. are made up of the nominal suffixes and ati, aSi (or atu^ asu) and ata, dta (dti, dSi etc. or dti, dUl cf. su-a-tu § 57, a). In the case of the 1. c. sing, this is 'another reason for putting aside aisi as an altogether impossihle reading; in the 2 f. we have contraction from ki-asi, and in the 3 m. and f. contraction from Su-aU Qur asu) or sa-asi. The form Su-a-Su is still found e. g. Assurn. Ill 76 {ana §u-a-Su 'to him'). In the 2. and 3. pers. plur. the plural termination is appended to the singular forms. — To say that the pronouns above given serve to express the genitive and accusative is in general correct. Asamatter of fact, in connexion with prepositions governing the genitive only these pronouns are used : cf. ana dU 'upon me' (lift up thine eyes, IV R 68, 29 b), ana kdh 'to thee' (fern., will he draw near), ana sdsu, ana MU 'to him, to her' (he spake), ana kdsunu 'to you' (IV R 56, 46 a), kima ia- ti-ma 'as I' (Tig. VIII 60), kima Msunu 'like them' (Khors. 96) , lanamma eli dSi 'another than I', ela kdti 'besides thee' (0 goddess, there is no deity). In the same way, in such a connexion as: 'him (himself), his wife, etc. he carried off', or: 'her (herself) I took captive alive', we never find, in the accusative other than SdSu, MSa. Still we also find: andku u kdSi 'I and thou' (will do so and so, K. 3437 rev. 3), and 132 morphology: § 55 indbpenbext peks. pkonouns. when one of these pronouns is, for the sake of emphasis, placed before the verbal or nominal suffix, it naturally stands in the nominative, e. g. M$u esir^u 'him I shut in', prop, as for him, (nom. ahsol.) I shut him in (Senhb. Ill 20); SdSu maSakm aktis 'as for him, I flayed him' (Khors. 35), kdtu amdtka 'thy command'. Other examples of this use of the pro- nouns in question will be found in the syntax §§ 119 and 135. For the rare cases — apart from Sultnu dU 'my greeting' — in which these pronouns are used as a circumlocution for the nominal suffixes v. syntax § 119, and for the equally rare cases in which they are used in the same way for the verbal suffixes, where no sort of emphasis can be intended, iMd. § 135. For the rare use Of iaSw as adj. 'selfsame' (usually Su'atU) V. § 57, a. c) In still another way do we find the nominal suffixes transformed into independent pronouns, a) Joined to rdmdnu (rdmenu, rdmnu) i. e. 'fear-com- pelling might' (st. QXil), the nominal suffixes express the idea of 'self: rdmdni 'I myself, rdmdnka 'thou thyself etc. Cf. Khors. 77: ina kdt rdmdniSu napiUa^u ukatti 'with his own hand he took his life (committed suicide)'; inBeh. 17 we read: 'Cambyses mitutu ra-man- ni-Su miti died by suicide'; rdmdnkunu 'you yourselves' (IV R 52, 23 a); — saknu sa rdmeni'a 'my own viceroy' MOKPBOLOGT : § 56 SUPFKED PEHS. PEONOUNS. 133 (Assurn. I 89); — ramnu e. g. Khors. 125. P) Joined to atiu in the forms attua (1. sing.) attuni or attunu (1. plur. ; not to be confused with attunu 'ye'!) and attukunu (2. m. plur.) they serve to emphasize the nominal suffixes, cf. at-tu-ni aSdbani 'our remaining' (V R 1, 122) ; for at-tu-ku-nu v. K. 312 1. 24. In the Achsemenian inscriptions, however, they are used simply as another way of expressing the nominal suffixes, the latter being even expressed, in some cases, over and above, see syntax § 119. In Beh. 18 we find attunu with the signification of a possessive pronoun: 'from the days of our fathers the supreme power has been at-tu-nu u Sa zer-u-ni ours and our family's', y) As a possessive pron. = 'thine' in ex- pressions like 'heaven and earth are thine', e. g. IV R 29, 26 ff., we find ku-um-mu, made up of the no- minal suffix ku (a bye-form of Ara, v. § 56) and the particle ma which appears in the case inflexion (cf. kaldmu gen. kaldmi and mimmu, mimmu § 58 end) ; for mm see § 53, d. 2. Suffixed personal pronouns, a) No- § 56. minal suffixes. For the way in which they are appended to the three cases of the sing, and to the various forms of the plural, and for the choice between the two forms i and a {=ia, § 41b) of the 1. c, sing., see § 74. For orthographical varieties like 134 mokphologt: § 56 suffixed pees, pronouns. mu-te 'my husband' (mu-ti-ma var. mu-te-ma Nimr. Ep. 42, 9), see p. 78. — For ka of the 2 m. we also find ku; for which note especially the text IV R 46: dl-ku 'thy city' (1.30 a), bit-ku 'thy house' (31a), belut-ku 'thy glory' (28 a) and many others. For the change of s of the suffixes of the 3. pers. singular and plural into s, see § 51. — Plur. 1. c. Besides ni we also find nu, as in atiiinu § 55, c, ^ and in proper names like Sadunu (also Saduni), A-hu-nu (also A-Jiu-ni). This suffix seems also to occur in the name of the Old Babylonian king Samsu-i-lu-na. — 3. masc. The m of bu-sd-su-num 'their (ace.) property' (Neb. VII 20) must be explained like that of at-tam (§ 55, a). An alternative form of Sunu is §unuti; cf, libba-Su-nu(^ti) 'their heart' (V E. 1, 120), [eH-Su-]nu-u-te 'over them' (Assurb. Sm. 35, 14), baltusunuti. — 3. fem. Once we find Hnu, viz: V R 66 col. II 19: mandatfi-si-nu 'their tribute' (countries are spoken of). b) Verbal suffixes. For the way in which these are appended to the verbal forms — both to those ending in a consonant and those ending in a vowel — when the third radical of the verb is strong, as well as when it is weak, see § 118. The forms UkulaUu^ ipta^su etc., which are found alongside of UkuUu, iptUu etc., call for a word of comment. Iskulassu, for example, is not to be explained as if the simple pro- morphology: § 56 suffixed PEKB. PRONOnNS. 135 nominal suffix Su, — and the same applies to Si, ka etc. — were appended to the verbal form iMula, which ends in short a; for the verbal suffix does n6t draw the tone to the last syllable : tu-na--a-Su-uu (V R 45 Col. II 52) might pass for such a form but never i§kuM$u, iskuMsunu. We should rather assume, in the case of verbs tertise infirmse, that forms like iptaSh are to be read iptd-si {=ipti-a + U) according to § 11. The analogy of verbs with a strong con- sonant as third radical, however, renders it prac- tically certain that here also we have to do with the stronger suffixes aUu, aUi, akka, which run parallel with SM, si, ka: al-ka-Su-nu-n-ti 'I removed them' (Tig. I 87) may be formed directly from alkd, but iptaUi, iptaSSunuti certainly stand for ipti-aEi, ipti-a§Sunuti (like naSanni 'he drove me' perm, for naSi-anni Neb. Ill 19): we even find such forms as: us-si-si-as-§u 'I brought it out' (III R 4 No. 2, 7). What we have just said does not exclude the possibility that in .cer- tain cases — as, for example, in verbal forms in a relative clause — the a of assu, aSSindtu, anndU may at the same time represent the final a of the verb. The verbal suffix of the I. pers. plur. occurs only in this stronger form: anndsi; in the same way the suffix of the 1. sing, is, without exception, anni after verbal forms in the singular. Exceptions are found only 136 MORBHOLOGY: § 56 suffixed PERS. PROSOOlifS. with verbal forms in the plur. (in u\ e. g. Tig, YHI 30: SalmiS litiarruni 'may they lead we in safety'; V R 7, 105: 'the rule over whom the gods iddimni had conferred on me'; Assurb. Sm. 11, 12: 'mighty forces usatlimuni have they granted unto me'; Esarh. IV 41 (uMzizuni). In passages like Assurb. §m. 11 (cf also 217, k) to refuse to recognize the suffix of the 1. pers. would result in a construction extremely harsh and forced, while to do so in the other cases is absolutely precluded by the context. U- sal-la-a-ni 'he implored me' (Esarh. Ill 7) stands for usalldnni. The question as to the origin of the stronger suffixes aHu, aSsi, akka (in certain cases ikkd), anni (in certain cases inni), assunu(tu or ti), asHndtu (or ti) and aHiniti, a?indsi is to all appearance intimately connected with that as to the origin of the Hebrew suffixes J^-r, •"t— etc. Examples of the 3. and 2. pers. are: usS- MlasSu 'he made him bring', (V R 7, 44), nmutu al-ku- na-lu (for alkunaUu) 'mercy he showed unto him' (end of sentence, Assurn. Ill 76), la tan&salli 'break it not' (0 Istar! Descent obv. 23), iptaUi 'he opened to her' (ibid. 1. 39), a-da~na{]i)-ka 'I shall give to thee' (end of sentence IV R 68, 21 a. 58 c), si tu-sa-an-nak-ka 'she gives thee to know' (Assurb. Sm. 125, 63); rimittu as- ku-na-(as-)su-nu (end of a section, Assurn. Ill 56), in- da-na-as-Su-nu-tii 'he gave them' (Beh. 96), 'what MQRPgoLOST: § 56 SUFFJXED PERS. PEONOUNS. 137 I a-kab-la-^S-M-M-a-iu command them' (NR 24), id- dan-)}^ra^-sirni-.ti 'he surrendered them, (yiz : the countries) to me' (NR 21). No distinction can be detected, to all appearance, in tbe use of the weaker and stronger forms of the suffixes. Ad:d,enda: S^ng. 1. c. A^ur-Se-zib-a^ni (C 28). i-ki- p^-an-nim 'he gave me up' (Neb. 1 42), cf. ai-tam § 55, a. After the 3. pars. fern. plur. we find -ini^i: i-Se-im-ma- 'rin-ni 'they obey me' (Beh. 7), 'the countries Sa ik-ki- ra--in-ni which rebelled against me' (Beh. 40). — ^2. m. Contracted tQ k : ak-ii-ba-ak 'I have said to thee' (IV R 68, 39 c) ; ku: lik-M-ku 'may he announce to thee' (IV R 66, 7. 8 a). — 2. f. H-hU-lak-ki 'let him bring to thee' (IV R 65, 38 b). — 3. m. On the change of the S of all the verba-l suffixes of the 3. pers. into s, see § 51; for the long u of Su in forms like liskipu-su- u-ma, § 53, d. The follg. are examples of the verbal suffix contracted to ^: u-sak-ni-su-us 'they subdued him', ak-M-iS (Neb. I 54), u-se-ri-ha-as 'he made him enter' (V R 35, 17); us-mal-M=u^malUsi, yiz. the palace (Senhb. Const. 86). Su is found strengthened by m (ma) in IV R 21, 30 b: Uk-ka-W-sum 'let it be told him'. — Plur. I.e. ikarrahanndU 'he blesseth us' (Nimr. Ep. XI, 181), 'who il-li-kan-na-si hath come to us' (Nimr. Ep. 60, 14); is-pur-an-na-a-Su 'he hath sent to us' (K. 647 obv. 7). — 2. m. ak-lak-ku-nu-su 'I spake 138 moepholosy: § 57 demonstrative peonocnS. to you' (IV R 52, 27b). — 3. m. du-u-ku-su-nu-u-tu 'kill them' (Beh. 48). at-ta-nab-bal-Su-nu-si 'I offer to them' (V E 63, 22 a); note also II R 11, 25— 28 b: id- din-Su-nu-sim, i-na-din-sii-nu-Si etc. — 3 f. ulteHh-h-na- n-tii NR 23. iS-te-ni-'-e-si-na-a-tim 'he provided for them' (V R 35, 14). askun-U-na-B-im (Hammur. Louvre II 6). The form -U-na has as yet been found only with the particle ni appended enclitically : 'the coun- tries Sa a-pi-lu-H^na-ni which I had conquered' (I R 27 No. 2, 23. Assurn. Ill 125. 133). 57. Demonstrative Pronouns, a) Su-a-tu {iu'atu, sudtu, M'atul), a contracted form of which is Mtu cf. § 38, a. It occurs only in connexion with a sub- stantive, after which it always stands, never before. A sufficient number of passages might be quoted in support of all the forms given in the paradigms. For the fem. of the sing. cf. Shalm. Obel. 50. Ill R 4 No. 1, 1. 2 and oft.: ina satti-ma si-a-ti 'in that same year'.- — Plur. m. dldni Su-a-tum also Mi-a-tum or, as I would propose to read, Su-a-tun {v. § 49, b, p. 117) V R 56, 9. 11. In the same sense as suatu fem. Wati, plur. su'atunu fem. Mtina we find ^u fem. si, plur. ^unu, and more frequently Sunuti, fem. Sindti: cf. dlu su-u and §u-u 'said city' (Assurn. Ill 133), dlu sii-u (var. dlu alone) 'this city' (V R 69, 21) — this explains the supposed suffix Sic in Sarg. Bull Inscr. 91 — , ekallum mobphologt: § 57 DEMONSTKATIVE peonouns. 139 Si-i 'that palace' (Assurn. II 5); murdni su-nu (Su-nu) 'said young lions' (Lay. 44, 16), sdbe Su-nu-li 'those people' (Shalm. Ob. 154), dldni Su-nu-ti 'those cities' (Assurb. Sm. 82, 7); as regards the fern., note how ekle Sa-ti-na and ekU si-na-a-ti change places in the two parallel passages III R 15 col. Ill 25 and Esarh. II 49. Very rarely, it would appear, was sdsu used for Su'atu (although they are at bottom identical, the one being=sM + asu, the other su + atu, see § 55, b), e. g. V R 64, 11 a: eli dli u Mti sa-a-Sit 'against that city and that house'. &) annu, from an-ni-u, cf. for example an-ni-u a-hi-u 'this other' (III R 54, 43 b), umu an-ni-u (V R 54, 39 a), genly. ina umi an-ni-i 'to-day', cf. jisoal. It is always placed after its substantive, except in an-na-a ka-M-e 'this speech' Nimr. Ep. 48, 178, III an-nu-tu sdhe 'these three people' (V R 54, 51 a). In an-ni-a-am (IV R 66, 30 a) we again meet with ma; so too in M-ma-mi an-nim (annSm) 'of these heavens' (Neb. Bab. II 2). For the fem. sing, note istu uSmdni an-ni- te-ma 'from that camping-ground' (Assurn. II 39 and oft.). — Plur. m. an-nu-te . . . an-nu-ie, also a-nu-ie 'some . . . others (. . . a third party)', v. Assurn. I 117. 90 f. c) M??M, e. g. D, 20: 'what I have done and what my father did, ul-lu-u-um-ma that may Auramazda 140 MOHPffOLOOT: § 58 BELATIVB PHONQTJN. protect'; D, 15: tabbanutu uHutv, 'those buildings, words' (atCc). — Another contrast to annu 'this' is ammu in the phrase ina pa^an (? § 9 No. 261) or simply padan, also padan^^ am-ma-{&-)te 'on the other side' of a river (Tig. II 4. Assurn. Ill 1), opposed to padan an-na-te (var. ti) Assurn. Ill 49 f. {padan am- ma-te, var. ti). d) agd (in Assurbanipal and especially in the Achffimenian texts) is placed both before and after its substantive: e. g. hit a-ga-a 'this house', a-ga-a Sadu 'this mountain', umu a-ga-a 'to-day', iame a-ga-a 'this heaven' (ace), irsitim a-ga-a-ta 'this earth' (the fern, is always placed after). Plural forms (always put after the subst.): salmdnu agannutu 'these portraits' (Beh. 106); mdtdti a-ga-ni-e-tii 'these countries' (Beh. 8. 9). In these forms of the plural agd is clearly strengthened by annu., as in agdsu by su. agdSu is always placed after a subst. or a proper name, e. g. nikrutu a-ga-su-nu 'these rebels' (Beh. 46. 65). § 58. The relative pronoun Sa (originally sa-a, ace. of SM, V. II R 31 No. 2, 14 c. d, and oft., cf. Hebr. -tp, •©, origly. IB) may also be used to express the genitive relation, e. g. ina silli sa Uramazda. The original demonstrative signification still appears in such ex- pressions as sa bit sibitti 'that (man) of the prison, the JtOEPHOLOat: § 58 iStiLATlVE PBONOniT. 141 prisonef' (IV R 58, 32 a, and cf. t R 13, 8— 10b), in which #« is used like the Arabic ,>. The so-called pronomen telativum generale 'whosoever, whatsoeyer, all that, as inuCh as, as many as' is expressed sometimes by the interi'ogative pro- noun with or without sa, sometimes by the substan- tives ma-la, mal (doubtless =»8^?fl!) and ammar (always without Sa, for which v. Syntax § 147), both of which Originally signified 'fulness'. Cf. tnan-nu Sa iidbalu 'whoever shall take away' (v. Diet. p. 214), man-nu atta sarru 'thou, whoever thou art, that shalt be king' (Beh. 105), ma-nu arkA 'whosoever shall be in the future' (I R 36 No. 2, 12); del tni-na-a ha-SUma 'lord of all that exists' (said of Metodach, Neb. I 35);-— ildni ma-la sum nabu 'the gods as many as there be', 'the living beings ma^ta ina mdti haM\ often in the phrase ma-la (mal) baSu 'as many as there are (or were) of them', gab-U ma^la epuHu 'all whatsoever I have done' (E, 9); sdbe am-mar ipparsidu 'the men, as many as there were of them, had fled' (Assurn. 1 66 and oft.). A third expression, by means of the indefinite pro- noun with or without sa, is found only for the neuter: man-ma (doubtless to be read min-ma or mim-ma. V. § 60) Sa etepusa 'all whatsoever I had done' (Shalm. Ob. 72); mi'im-ma or -^-ma — i. e. mim-ma (v. § 9 No. 212) — or mimma (apparently the sign, nin, v. ibid.) 142 morphology: § 59 INTEBBOG. PBONS. § 60 INDEP. PBON. $umsu '■whatsoeTer is named i. e. exists', tnlmma isu 'whatsoever I possessed' (Nimr. Ep. XI, 77 ff.), mimma sa suma naiu 'all creation'. Note also ^-mu-u i. e. mimmu eppusu 'whatsoever I do' (VR 63, 11a, cf. 41b), ■^-mu-su(-nu) 'his (or their) property' (often in the contract-tablets) and cf. man (i. e. mim)-mu-Su 'all his goods' (K. 245 col. II 68). § 59. Interrogative Pronouns. For mannu and minu (e. g. ina eli mi-ni-e 'on whose account?' V R 9, 70) references are unnecessary. With mannu there is used interchangeably the pron. a-a-u i. e. d-u {v. §§ 13 and 31), e. g. a-a-ii ildmad 'who learns?' (IV R 67, 58a) a-a-ii ilu 'which god?' (IV R 9, 52a), which goes back, directly or indirectly, to the interro- gative stem d, which, again, may be regarded either as contracted from ai (cf. Stade, Hebr. Gramm. § 99, 3), or better, as an interrogative existing independently alongside of ai. — The same applies, of course, to Hebr. is alongside of ']';i«; cf. Idtim 'houses' alongside of hait. §60. The indefinite pronoun is formed partly by the reduplication of the interrogative stem man (per- sonal indef. pron.), partly by appending enclitically the generalizing particle ma to the interrogative stems man (for persons) and min (for things). Illustrations are to be found everywhere (cf. ma-ma sa-na-a 'any man' mokphologt: § 60 indefinite pronouns. 143 IV R 45, 25 ; mi-im-ma or mi-ma lim-na 'anything wicked' Tig. VIII 70). Mu-um-ma 'any one whatever' (Shalm. Mon. rev. 71) stand's quite alone. For man- man etc. employed as adjectives cf. ilu ma-nu-man ul . . . 'no god' (IV R 6, 14 c). Manman is usually, as here, followed by the negative. When the latter stands at the head of the clause, la mammana etc. also signifies 'nobody'. — Both manma {mamma) and mvmma are very frequently represented ideographic- ally by -^ with the phonetic complt. m«, and this when closely written looks very like the sign, nin (v. § 58 above). For -^-ma (nin) = mamma see, e. g., V R 6, 66 (mamma ahu 'some stranger') and Diet, p. 293 f., for nin^=mimma t). V R 63, 23 a (a variant of mi-im-ma Neb. II 32. VIII 11) and many other passages (always so in mimma sumsu 'all sorts of). — The neuter indefinite pronoun is also sometimes written wiaw-ma ; see § 58 above, and also cf. man-ma amdt limutti 'any- thing wicked' (I R 27 No. 2, 80, for which in 1. 42: mimma amdt limutte). Since it is highly improbable that manma is also used as a neuter, we ought cer- tainly to read min-ma or {y. § 49, b, p. 117) mim-ma, especially as there is express testimony V R 37, 34 d that the sign man has also the value min (cf. also man- di-e-ma IV R 53 No. 3, 37 corresponding to mi-in-di- e-ma Nimr. Ep. 65, 13). — For d'umma — as regards the 144 morphology: § 61 SiaNIFICA^tl feOOfs. reading and writing of which §§ 12 — 14 are to be coinpared-^see e. g. ShaL BaL V 3: a-(i)a-'um-ma til Stib 'not one did I spare', sarru ia-um-ma 'soiiie king or other' (Tig. I 67 and oft.), la te-zi-ha a-a-am-tna 'leave not one alive' (M 55 col. I 21)= Most of tlie pronominal stems mentioned in §§ 55 — 60 we shall meet again among the particles (iJ, Sw in the adverb umma and the conjunction Summa, ago, in the adverb aganna, and so on); for details see under §§ 78 — 82. Transition to tlie Noun and the Verb. § 61. In Assyrian as in all Semitic languages roots in which an idea or meaning is inherent are of two kinds. Either they originally consist of three or ' more consonants, or originally of two, which have been subsequently increased to three. 1) Roots composed of two Consonants are still found: a) in those nouns that show complete reduplication — no verbs have yet been found. — Such nouns are: lakalaka 'stork' syn. rakrakku, sarsaru 'cricket', barbaru 'jackal', panpanu 'chamber for gods'; dandannu 'all-powerful', kaSkaS{S)u 'very strong'; kalkaltu 'panting, pining', kamkammatu 'ring'; — birbirru 'brightness of the rising stars', zirzirru name of a tiny insect, dikditcffu, name of a very small bird; — zunzunu and dukdukku synonyms of the two last- mentioned words, mulmul{l)u 'javelin, lance', b) in MOBPHOLOSy: § 61 SIGNimCAUT ROOTS. 145 those nouns and verbs that show imperfect reduplica- tion of the root. Verbs are rare: lahdlu 'bring'; kakdru II 1 'root out', tatdpu 'shut in' (part. II 1 : mu-te-tip- tum, to which add ti-tip-pu, a synonym oi daltum 11^23, 2. 3 c). As regards the nouns, the question sometimes suggests itself, whether it is not the case that the second radical of the biconsonantal root has been assimilated to the first radical, when the latter is re- peated: as, for example, in ka{k)ka'bu 'star', kakkadu 'head', cf. 'ip'ijs, kakkaru 'ground', the masc. prop. noun. Hahhuru Hebr. lirTiti, sissinnu 'palm-branch' of. Di3p5D, kukvb&nu 'stomach' (of animals), cf. Arab. _,Sj^, Aram. KSSfs'llp. We must not, however, from these and a few other similar contractions (cf. li-il-li-du 'child' II E 30, 47 c) formulate any laws for the assi- milation of Assyrian consonants of universal applica- tion, or even as applied to the derivatives from triconsonantal roots (cf. above § 50). Note further papahu 'chamber for the gods', dadmu 'dwelling-place', mamlu 'strong', lallaru 'crier', fem. lallartu 'wailing, loud crying' (also name of a bird and an insect), sis{s)iktu 'dress, robe', dudittu (=dudintu) 'ornament for the breast', pitpdnu (?) 'bow', c) Possibly in some of the so-called primitive nouns to be discussed in § 62. — These biconsonantal roots may also be in- ferred from some of the so-called 'weak' stems or Delitzsch, Assyrian Grammar. 10 146 morpholoot: § 61 biqhipicant boots. verbs, notably from tbe verbs terti* i and 1 (v. § 62), verbs mediae geminatse, wbich probably go back to a biconsonantal root with a strongly accented a vowel {v. § 63), and verbs medise 1 and i, which seem to be developed out of a biconsonantal root with a for the medial vowel (v. § 64). 2) As examples of roots composed of three consonants we have first of all the verbs with three strong radicals. Whether the n of verbs primae 3, and the u, i of verbs primse 1, "i be of secondary origin, and if so, in what cases — are questions which will be difficult to decide. The gutturals, however, were without doubt as inseparably part and parcel of the root in verbs mediae and tertise i{, y, n, as in the case of verbs primse s, y, n. 3) Roots composed of four consonants, em- ployed as verbs, are but sparingly represented in Assyrian ; the two principal examples are nsba IV 1 'free one's self, be torn to pieces; cross over', and IIB-IS IV 1 'flee', cf. also umtj, t253in II 1. Ill 1 'destroy', nobS, UTIB. Of the nouns we may mention; akrabu 'scorpion', harbasu 'fright' (?), paltigu 'travel- ling-chair' (II R 23, 6 a), parzillu 'iron', hdb{h)asillatu 'stalk (of grain and of a flower)', parSumu and pur- Sumu 'old, venerable', sursummu, hurhummatu, pur{par)- lu-u-ii 'flea', sumelu 'left' (bXittIB), and many others. MORPHOLoey: § 62 primitive nouns. 147 From these examples we see the extent to which the liquids / and r have contributed to the formation of roots of four consonants. — For roots like nsbs and "iin©, which have four consonants in appearance only, V. § 117, 1 and 2. — I know of no roots composed of more than four consonants. B. The Noun. On the difficult question as to the existence of § 62. so-called primitive nouns, the following remarks may be made from the stand-point of Assyrian. 1) Primitive nouns alongside of roots ter- tiae 1. We have already, in § 39, referred to the extreme shortening which takes place in the forms of the part, of the Qal (and Shafel) in verbs tertise K and 1, in the perm, of the Qal in verbs tertise i, and in the constr. state of the nominal stem J-^^- This shortening, we found, resulted in the complete suppres- sion not only of the final vowel, whether short or long, but also of the last consonant of the root. Among the derivatives of verbs tertise i, nominal stems formed like tertu 'law' or tudtu 'decision' (from mi and nil V. § 65 No. 32, a) deserve, in this respect, to be singled out. The same disappearance of the auslaut may also be observed in a series of nouns which, after what has just been said, must not, from 10* 148 moephology: § 62 pkimitivb noons. the mere fact that they contain but two radicals, be set down as primitive nouns, in the sense that the corresponding verbs tertiffi i never existed or, at least, had never reached the triconsonantal stage. Impossible as it is to derive forms like Surb-at, ter-iu from other than triconsonantal stems, it is equally unnecessary, to say the least, to regard as primitive nouns, Arm 'god of heaven', fern. An-tu [st. cstr. Anai], sat-tu '■jeaTL\=santu [sana], kai-tu 'bow', plur. kasdti, am-tu 'maid', dal-tu 'door', ^ap-tu 'lip', dar-tu 'rising, revolt' ; enu 'lord' fern, entu, enu 'time', fern, en-tu, ettu, ittu; binu 'son' fern. Mn-tu, ilu 'god' fern, il-tu [Hat], isu 'wood', ir-iu 'breast' [irat], it-tu 'side' pi. itdti, sinu (sind) 'two'; suk-tu 'drinking- trough', ul-tu originally 'direction', then prep, 'from', and others, especially as for the most of these nouns there exists a tricon- sonantal stem. In the latter as in the former case, we may have to do with examples of the extreme shortening of verbs tertiae i, — a fact which no one questions in regard to el=eli, eli, elai; mat=matai, le-at fem. st. cstr. of /e'w 'strong', and many others (cf. also Hebr. 15>, 1j5, SI). While, therefore, daltu, Mnu, Un-tu etc. are not, any by means, necessarily primitive nouns, it is still, from another point of view, very remarkable that nominal formations like the above are not found MOEPHOLOQY: § 62 PP.IMITIVE NOtlNS. 149 from stems (verbs) tertise s (nominal stems like mi-lu 'flood' and ze-ru 'seed' are proved by the variants mi-i-lu and ze-e-ru to be of the same form as zimu, bunu, V. § 65 Nos. 1 — 3). It is evident that the third radical of stems tertise i (also 1?) was' much less distinctly felt to be part of the root, and treated as such, than was final i?; and it was for this reason that in § 61, 1 the stems tertise i (and l) appeared to me to warrant, more than did any others, the assumption of biconsonantal roots. The problem becomes more complicated when we take the case of those biconsonantal nouns, to which we cannot assign a definite and known stem, as, for example, ahu 'brother' and 'side', and emu 'father-in- law'. Are these, by reason of their feminines ahdtw 'sister' and 'side', and emetu 'mother-in-law', to be set down as shortened forms from triconsonantal stems tertise i, or must they be recognized as bicon- sonantal primitive nouns, that are on the point of rising beyond the biconsonantal stage and appearing henceforth as triconsonantal (note the instructive atM 'companion') ? For aMtu in the signification of 'sister' as well as for emetu, the latter alternative seems to me to deserve the preference, because only in very rare cases do we find the form Jlii with a concrete personal signification. The a seems 150 moepholooy: § 62 primitive nouns. to owe its existence to an effort at strengthening, or, so to say, expanding the short word of two con- sonants. In this respect it resembles the a in perman- sive forms like danndta etc. — Abu 'father' (m is tho- roughly well attested) cannot, from an Assyrian standpoint, be regarded as other than a derivative from a triconsonantal stem 7\y& (prob. 'decide'). 2) Other primitive nouns. TJmmu 'mother', origin- ally 'womb', we may easily recognize in Assyrian as derived from the stem oax 'be wide, spacious'; it may be allowed to pass for a primitive noun only in so far as all stems mediae geminatae are at bottom of biconsonantal origin. To set down words like sdsu 'moth' and sumu 'garlic' as primitive nouns in the sense in which the term is usually understood (cf. Stade's 'isolated nouns') is very hazardous, to say the least, since it is perhaps a mere chance that the corresponding verbal stems with medial vowel are no longer to be, or have not yet been, met with in our texts. In the case of ddmu 'blood' 'and dmu 'sea', we might, looking at D'n, D'^, ^OT, ^!>; D';, D^, D'^B';;, regard them as primitive nouns for the reason that the various Semitic languages — to a certain extent, each separate language — have adopted various plans in order to give to these words a greater stability; but who will guarantee that the Hebrew and Arabic forms. MOEPHOLOOY: § 63 NOCNS from stems MED. GEMmATAE. 151 let us say, represent only a comparatively late stage of development under the influence of continuous shortening, analogy etc., and that damn and idmu must, nevertheless, be assumed as the ground-forms in primitive Semitic, from some unknown stem which has long ago disappeared? The same applies to Udtu 'fire' and others. We should be most inclined to see examples of primitive nouns in mutu 'husband', idu 'hand, side', and immu, connected with iimu 'day' (of. itJi, D'^tt'i, ace. to Praetorius from an old word iim), whereas, on the other hand, mdtu 'country', Sumu 'name', mu 'water', pu 'mouth' are altogether uncertain. An intermediate position between the so- § 63. called primitive nouns and the forms of nominal stems discussed in § 65 is occupied by the derivatives of verbs mediae geminatse and mediae i and 1, inas- much as these present unmistakeable traces of their descent from biconsonantal roots {v. § 61, 1). For this reason we shall treat both classes apart from the derivatives of the other stems, and shall begin with the formation of nouns from stems mediae geminatae. While the verbal forms from these stems follow in all respects the analogy of the strong stems (the only exception is the permansive of the Qal, V. § 87 and cf. § 89, also § 37, b), this cannot be said, to anything like the same extent, of the nominal 152 mokpholoqt: § 63 notois tkom stems med. geminatab. forms. Nouns like dannu 'powerful', sarru 'king', sallu 'captured' are formed directly from the root, there being no proof — as is the case in the feminine forms of the nominal stems J^ (§ 65 No. 6) and Jj«i (No. 7) — of an intermediate stage with a vowel between the second and third radicals. Between sarru and dannu^ ellu, emmu ('hot') — the three last men- tioned, being adjectives, cannot by any possibility re- present the form jJis — and between Sarratu 'queen' and dannatu 'powerful', dannat 'she was powerful' (the permansive form is Jjsi), Sullatu 'booty' (cf. the Hebr. masc. bbiB) no difference can be detected: in other words, stems mediae geminatse, in place of all the forms of nominal stems enumerated in § 65 Nos. 1 — 10, are content with three: with jJiS, which combines the meanings of substantive and adjective, and with j^ and Jjti, which form only substantives. Of jjis examples have already been given. For j^As cf. sillu 'shadow', sippu 'threshold', liblju 'heart', hissatu 'perception'; illatu 'might', although found written ellafu, must also belong here on account of its femi- nine ending atu {ellatu would give ellitu), while sirritu § 34, 8, on the other hand, must be put alongside of the Hebr. r\yi. For jji cf. guhhu 'cistern', zumlu 'fly', uzzu and uzzatu 'anger', kullatu 'totality, whole'. morphology: § 64 nouns i-kom stems med. '. and \ 153 Corresponding to these forms we have Sarrutu, §aUutu (§ 65 No. 34); fiarrdnu 'street', Rammdnu, zilldnu (No. 35). Only when a long vowel appears between the second and third radicals, or when the doubling of the second or third radical is characteristic of the nominal stem, are the stems med. geminatte compelled to follow the example of the strong stems. Hence Saldlu, nardru (nerdru) 'helper' (No. 11); dumdmu 'wild cat' (13); hasisu 'intelligence' (14); Mlilu 'setting (of gem), garland', zikiku and zakiku 'wind' (15); Saruru 'brilliance', «6m6m 'deluge', aSidtu 'sorrow' (17); sibubu 'brilliance', sinundu 'swallow' (18); sululu 'shadow, shade, covering' (19); Dan-na-{a-)nu prop, n. m., al-lal-lu 'strong' (25, or is it stem No. 23?); namhuhtu (28) name of a bird; imluhu 'flute' (ans, 30, e). For the nominal stem No. 31, a, cf. masallu 'herdsman's tent', namaddu 'measure', on the one hand, and on the other (like strong stems) manzazu 'stand, place' fern, manzaltu. — The stems kunnunu, luklulu (also namurratu) see under § 88. Stems mediae 1 and i agree with stems med. §64. geminatse in not belying their descent from biconson- antal roots. This descent is most distinctly visible in the permansive form of the Qal : ddr, kdn^ tdrat etc. {v. § 87 and cf. § 89). Impossible as it is to fit these forms into the scheme of the ordinary permansive 154 morphology: § 64 nouns from stems MED. 1 AND -■. JjU, it is equally unnecessary to make nominal stems like tabu 'good' contracted from an hypothetical original taialu (stem Jjii) ; they are further examples, rather, of the oldest form of the root, which is still free from internal rowel change. Also for the semi- nominal infinitive of the Qal: tdru (with feminine ending tdrtu), tabu we must he content to give up the theory of intermediate forms, such as taudru, taidbu. When once the characteristic of the infinitive, the vowel d, had appeared before the last radical ( jCai), tdru was the natural root-stem of the infinitive. In connexion with the permansive forms above referred to, which will be thoroughly discussed in § 89, such as da-{a-)ri, ka-ia-an and ka-a-a-an, ta-ab, ta-a-a-rat etc., which must undoubtedly be read ddri, kdn, tdb, tdrat (v. § 13), we would call attention, in the first place , to the most difficult forms of stems med. 1 and i, that is to the nouns written da-ia-nu, da-a-a-nu 'judge', a-a-bu, ia-a-bu, a-ia-a-bu 'enemy', ha-a-a-ru 'consort' and many others. Looking at the Hebr. 'j^'^, we naturally feel inclined to read the Assyrian word for 'judge' also daianu ; but apart from the fact, that such a form is at variance with the treatment of the intervocalic i elsewhere (§ 41, b), in accordance with which da'anu is all that would be phonetically admissible, it is shown to be altogether MOEPHOLOaT: § 64 NOUNS FROM STEMS MED. 1 AND -. 155 out of the question by the orthographical variation da- a-a-nu. For this form, whether read dotanu or da'dnu — dainu is excluded for reasons both graphical (v. § 13) and grammatical — can never be classed under the form JSJ (§ 65 No. 24). The only way out of the difficulty would be to read da'dnu, which we might regard as a form Jlii; for this we might appeal to za-ia-a-re 'the adversary' (Assurn. 18), to a-ia-a-bu and ta-ia-arru (v. § 14). Moreover, the form jlSi (No. 25), which, in any case, is scarcely found in Assyrian for names of professions, such as 'judge', appears quite unsuitable for words like za-ia-a-ru, a-ia-a-bu, while for a word like ha-a-a-ru 'bridegroom, husband', which can hardly be assigned to a different category from a-a-bu and da-a-a-nu, it is utterly im- possible to assume a form with the second radical sharpened. A vastly more suitable explanation, however, appears to be suggested by the word ha-a- a-ru, just mentioned, which in II R 36, 39 — 42 d, appears alongside of the part, ha-i-ru; it is that these supposed nouns are nothing but participles with syncopated i, that Mru accordingly stands in the same relation to hd!iru as do dlbu to dUbu, ramu 'loving' to rd'imu (§ 37, a). If this explanation is correct, it of course disposes finally of the theory that a-a=ai: and further, whoever believes in 156 mokphologt: § 64 nouns from stems MED. 1 AND ^. the possibility of hairu and ailu being contracted from hcfiru and d'iiu, will be obliged henceforth to relinquish this theory in face of the construct a-a-ah (§ 14). There is, moreover, another important con- sideration which tells against this view of afew, ddnu etc., viz. the fact that it is precisely the parti- ciples of verbs med. 1 and i, formed after the mOdel of the strong stems, that, in contrast to the J^li of all other verbs, are wont, for a reason that is easily intelligible, to preserve in its purity the f-vowel before the last radical : cf. out of a great number of such participles only za--i-re, za-i-re, za-e-re, za-e- ru-ut (IV E 44, 25. Tig. VIII 32. 41. Assurn. I 28. Shalm. Ob. 20. Senhb. V 57. Neb. II 25 etc.), da-i- nu-te 'judging' pi. (Sarg. Cyl. 53), sd'idu, daHku fem. daiktu {v. § 13). There is, consequently, no other course open to us but to read all these nominal forms as ddnu, dbu, zdru ("iT), taru^ hdru in connec- tion with the permansive forms mentioned at the commencement of this paragraph, and in agreement with the two-fold employment of the Hebrew Dp as 3. m. perf. and as participle. — That these nouns have frequently ut in the plur. is entirely in harmony with their character as participles. — In the same way ka- a-a-nam-ma adv. 'continually'=M«ame, ka-a-a-ma-nu (st. § 65 No. 35) adj. 'everlasting; Saturn'=A'«»«a«M etc. morpholo&y: § 64 nouns from stems med. i and ■>. 157 It has already been shown in §§ 12 — 14 that there is no impediment from the side of the orthography: the permansive forms and the Tarieties in the way of writing a word like t&rtu (§ 13) corroborate anew the statements there laid down. A-a-lu 'ram', in accordance with the laws of Assyrian writing and phonetics, can hardly be read otherwise than as alu. In the case of words like a-a-lu 'stag', and the name of the mouth A-a-ru, we may on principle assume the form Jjlai as ground-form, but there can be no doubt, all the same, that they were respectively pronounced merely as cilu and Aru. Even if we were to read A'aru or quite falsely Aim, the Hebr. I'JK (same form as 13N) would have to be set down, in any case, as a free Hebrew transformation of the Babylonian name (as in ^TOti'i^). A'alu 'stag' (Hebr. b^N), we may be sure, was forthwith contracted to alu (v. § 47), which affords the only satisfactory explanation of 'ram' and 'stag' being- written identically : viz : a-a-lu. — The "■ in the first syllable of the name for Saturn 'p^D, i^^^ contrasted with the Assyro-Baby- lonian kdmSn, kdvan (cf. p. 104 f.) goes back perhaps to a bye- form Mvdn, with the first a modified to S, which may have been current in every-day speech (of. ^Silfl in its relation to Burner, § 49, a note). As examples of the remaining forms of nominal stems we would mention: miltu 'death', suru 'ox', urru (= uru) 'light', inu 'eye', imtu 'fright' (stem § 65 No. 1) ; niru 'yoke', dinu 'judgment', Bhtu syn. pirhu 'shoot' (No. 2); suku 'street', nunu 'fish', ru^tu 'breath' (3); mitu, mefu 'dead' (abstr. noun metutu), kenu, ktnu fem. kettu, Mttu 'true, just' (7); tdriu 'return' (11); ki-a-hi 'surname' (? ©"ip 12) ; simu '(purchase) price', fem. Hmtu 158 morpholooy: § 65 summary of nominal stems. 'fate' (prop, that which is fixed), diktu 'fallen host', kistu 'present', Mr(a)tu 'bride, wife' (14) ; makanu 'place', mdkdsu 'rack', mahdzu 'town', mdlu (biS) 'front', mandhtu 'resting-place', also 'care for one' (31, a); mutdnu 'plague, pest', si-da-nu 'hunting-net' (35). For the stem kunnu, fern, tubtii^tubbatu (and the form kut- tin-nu derived therefrom) see § 88. There yet remain many difficulties to solve : for example, does puru 'young -wild-ox' stand to piru 'elephant (st. T'S 'be strong, powerful') , and pulu 'dressed stone' to the more common pilu iptlu) with the same meaning, in the same relation as iLs»i to jj-3«i? For the nominal stems tiduku 'killing', titHru 'bridge', tinuru 'stove' cf. § 83 note. § 65. Summary of the Nominal Stems in Assyrian.*) I. Internal Vowel Change only (Nos. 1 — 19). 1. Short vowels only (Nos. 1 — 10). a) A short accented vowel after the first radical and a short, unessential, vowel after the second *) That is of those corresponding to verbal stems composed of three strong consonants. Weak stems will also be included, with the exception of stems med. gemin. and med. \ i. For no- minal stems of four consonants, in so far as they present internal vowel change only, see § 61, 1 a and 3; note also § 65 No. 35 (at end) and especially § 117, 1 and 2. — Arrangement observed in discussing Nos. 1 — 33: the derivatives of the strong trlconsonantal stems, which include those of stems primse 2, are taken first, and are separated by a period and a dash from the derivatives of the mobphology: § 65 nominal stems nos. 1 and 2. 159 (Nos. 1 — 5): confined most probably to substan- tives. The vowel beard after tbe second radical serves merely to prevent tbe stem from terminating in two consonants, and is syncopated, almost witbout exception, wben inflexional endings are appended. Tbere is entire uniformity among tbe stems intbis division, except as regards tbe addition of tbe femi- nine ending atu: before tbe latter, Nos. 1 — 3 synco- pate tbe second vowel, while Nos. 4 — 5 retain it (in tbe abs. state). 1. J^s (^3-^ st. cstr. Jjis) fern. SSsl- kalbu [A-«to6] 'dog' fem. kalhatu, samSu [Samaf\ 'sun', ma^ku [masak] 'skin', saknu \)akan\ 'viceroy'. — ahnu [aban] 'stone', anbatu, but also ersitu; eklu [e-ki-el] 'field'; enzu, erpu fem. erpitu (v. §§ 34, y. 35) ; — re^u 'head' fem. reitu; siru 'back'; remu, Seru; helu fem. 'beltu, but also rddu {v. § 32, y); — mdlu 'abundance', Idbbu 'lion'; zeru, di-mu ^tGa,v'=demu, ddmu (v. §§ 33. 47); — beru 'glance' (IVR45, 43), bSru 'middle' fem. berit; perhaps also menu, minu (=mdnu) 'number' (v. §§ 33. 41); — arhu [araJi] 'month'. 2. JJi (jJi st. cstr. Jjti) fem. S^Jd. zikru weak stems ; the latter, separated from each other by a semicolon and a dash, are taken in the following order: primse, mediae and tertlae K, tertise "■ and \ primse 1 and i. The forms of the con- struct state are always placed in square brackets, as in § 62. 160 morphology: § 65 NONINAL. stems NOS. 3 AND 4. [zikir] 'name', sibtu 'staff', kirbu [Mrib] 'interior', kibratu 'point of the compass, direction, region', zibbatu 'tail'. — igru 'pay'; — rimu 'wild ox', siru 'flesh' (v. § 47); ■ — ■ hittu, hitu 'sin', milu 'flood, overflow' {v. § 47) ; — simmu 'blindness' (