Handbook of Comparative and Historical Indo-European Linguistics HSK 41.3 Handbücher zur Sprach- und Kommunikationswissenschaft Handbooks of Linguistics and Communication Science Manuels de linguistique et des sciences de communication Mitbegründet von Gerold Ungeheuer Mitherausgegeben (1985−2001) von Hugo Steger Herausgegeben von / Edited by / Edités par Herbert Ernst Wiegand Band 41.3 De Gruyter Mouton Handbook of Comparative and Historical Indo-European Linguistics Edited by Jared Klein Brian Joseph Matthias Fritz In cooperation with Mark Wenthe De Gruyter Mouton ISBN 978-3-11-054036-9 e-ISBN (PDF) 978-3-11-054243-1 e-ISBN (EPUB) 978-3-11-054052-9 ISSN 1861-5090 Library of Congress Cataloging-in-Publication Data Names: Klein, Jared S., editor. | Joseph, Brian D., editor. | Fritz, Matthias, editor. Title: Handbook of comparative and historical Indo-European linguistics : an international handbook / edited by Jared Klein, Brian Joseph, Matthias Fritz ; in cooperation with Mark Wenthe. Description: Berlin ; Boston : De Gruyter Mouton, 2017- | Series: Handbücher zur Sprach- und Kommunikationswissenschaft = Handbooks of linguistics and communication science, ISSN 1861-5090 ; Band 41.1- | Includes bibliographical references and index. Identifiers: LCCN 2017042351| ISBN 9783110186147 (volume 1 : hardcover) | ISBN 9783110261288 (volume 1 : pdf) | ISBN 9783110393248 (volume 1 : epub) | ISBN 9783110521610 (volume 2 : hardcover) | ISBN 9783110523874 (volume 2 : pdf) | ISBN 9783110521757 (volume 2 : epub) | ISBN 9783110540369 (volume 3 : hardcover) | ISBN 9783110542431 (volume 3 : pdf) | ISBN 9783110540529 (volume 3 : epub) Subjects: LCSH: Indo-European languages--Grammar, Comparative. | Indo-European languages-Grammar, Historical. | BISAC: LANGUAGE ARTS & DISCIPLINES / Linguistics / General. Classification: LCC P575 .H36 2017 | DDC 410--dc23 LC record available at https://lccn.loc.gov/2017042351 Bibliographic information published by the Deutsche Nationalbibliothek The Deutsche Nationalbibliothek lists this publication in the Deutsche Nationalbibliografie; detailed bibliographic data are available on the Internet at http://dnb.dnb.de. © 2018 Walter de Gruyter GmbH, Berlin/Boston Typesetting: Meta Systems Publishing & Printservices GmbH, Wustermark Printing and binding: CPI books GmbH, Leck Cover design: Martin Zech, Bremen www.degruyter.com Contents Volume 3 XIII. Slavic 80. 81. 82. 83. 84. 85. 86. The The The The The The The documentation of Slavic phonology of Slavic . . morphology of Slavic . syntax of Slavic . . . . . lexicon of Slavic . . . . dialectology of Slavic . evolution of Slavic . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1397 1414 1538 1557 1571 1585 1600 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1622 1640 1651 1668 1681 1698 1712 documentation of Albanian phonology of Albanian . . . morphology of Albanian . . syntax of Albanian . . . . . lexicon of Albanian . . . . dialectology of Albanian . . evolution of Albanian . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1716 1732 1749 1771 1788 1800 1812 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1816 1832 1839 1850 1854 1857 XIV. Baltic 87. 88. 89. 90. 91. 92. 93. The The The The The The The documentation of Baltic phonology of Baltic . . morphology of Baltic . . syntax of Baltic . . . . . lexicon of Baltic . . . . dialectology of Baltic . evolution of Baltic . . . XV. Albanian 94. 95. 96. 97. 98. 99. 100. The The The The The The The XVI. Languages of fragmentary attestation 101. 102. 103. 104. 105. 106. Phrygian . Venetic . Messapic Thracian . Siculian . Lusitanian . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . vi Contents 107. Macedonian . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 108. Illyrian . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 109. Pelasgian . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1862 1867 1873 XVII. Indo-Iranian 110. 111. 112. 113. The The The The phonology of Proto-Indo-Iranian morphology of Indo-Iranian . . . syntax of Indo-Iranian . . . . . . lexicon of Indo-Iranian . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1875 1888 1924 1942 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1960 1974 1985 2000 2012 119. The shared features of Italic and Celtic . . . . . . . . . . . . . . . . . . . . . . 120. Graeco-Anatolian contacts in the Mycenaean period . . . . . . . . . . . . . . 2030 2037 XVIII. Balto-Slavic 114. 115. 116. 117. 118. Balto-Slavic . . . . . . . . . . . The phonology of Balto-Slavic Balto-Slavic morphology . . . The syntax of Balto-Slavic . . The lexicon of Balto-Slavic . . . . . . . . . . . . . . . . . . . . . . XIX. Wider configurations and contacts XX. Proto-Indo-European 121. 122. 123. 124. The The The The phonology of Proto-Indo-European . morphology of Proto-Indo-European syntax of Proto-Indo-European . . . lexicon of Proto-Indo-European . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2056 2079 2195 2229 125. More remote relationships of Proto-Indo-European . . . . . . . . . . . . . . . 2280 General index . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Languages and dialect index . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2293 2387 XXI. Beyond Proto-Indo-European Contents vii Volume 1 Preface . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . I. 1. 2. 3. 4. 5. 6. 7. 8. v General and methodological issues Comparison and relationship of languages . . . . . . . Language contact and Indo-European linguistics . . . . Methods in reconstruction . . . . . . . . . . . . . . . . . The sources for Indo-European reconstruction . . . . . The writing systems of Indo-European . . . . . . . . . Indo-European dialectology . . . . . . . . . . . . . . . . The culture of the speakers of Proto-Indo-European . The homeland of the speakers of Proto-Indo-European . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 7 15 20 26 62 75 85 . . . . . . 93 98 105 114 121 129 . . . . 138 . . . . . . . . 171 210 . . . . 220 . . . . . . 239 249 256 274 291 298 II. The application of the comparative method in selected language groups other than Indo-European 9. 10. 11. 12. 13. 14. The The The The The The comparative comparative comparative comparative comparative comparative method method method method method method in in in in in in Semitic linguistics . . . Uralic linguistics . . . . Caucasian linguistics . . African linguistics . . . Austronesian linguistics Australian linguistics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . III. Historical perspectives on Indo-European linguistics 15. Intuition, exploration, and assertion of the Indo-European language relationship . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16. Indo-European linguistics in the 19 th and 20 th centuries: beginnings, establishment, remodeling, refinement, and extension(s) . . . . . . . . 17. Encyclopedic works on Indo-European linguistics . . . . . . . . . . . . 18. The impact of Hittite and Tocharian: Rethinking Indo-European in the 20 th century and beyond . . . . . . . . . . . . . . . . . . . . . . . . . . IV. Anatolian 19. 20. 21. 22. 23. 24. The The The The The The documentation of Anatolian . . . . . . . . phonology of Anatolian . . . . . . . . . . morphology of Anatolian . . . . . . . . . syntax of Anatolian: The simple sentence lexicon of Anatolian . . . . . . . . . . . . dialectology of Anatolian . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . viii Contents V. Indic 25. 26. 27. 28. 29. 30. 31. The The The The The The The documentation of Indic . . . . . . . . phonology of Indic . . . . . . . . . . . morphology of Indic (old Indo-Aryan) syntax of Indic . . . . . . . . . . . . . lexicon of Indic . . . . . . . . . . . . . dialectology of Indic . . . . . . . . . . evolution of Indic . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 309 325 344 377 409 417 447 VI. Iranian 32. 33. 34. 35. 36. 37. 38. The The The The The The The documentation of Iranian phonology of Iranian . . morphology of Iranian . syntax of Iranian . . . . lexicon of Iranian . . . . dialectology of Iranian . evolution of Iranian . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 471 481 503 549 566 599 608 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 625 638 654 682 695 710 717 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 733 743 751 804 828 835 858 VII. Greek 39. 40. 41. 42. 43. 44. 45. The The The The The The The documentation of Greek phonology of Greek . . morphology of Greek . syntax of Greek . . . . . lexicon of Greek . . . . dialectology of Greek . evolution of Greek . . . Volume 2 VIII. Italic 46. 47. 48. 49. 50. 51. 52. The The The The The The The documentation of Italic phonology of Italic . . morphology of Italic . syntax of Italic . . . . lexicon of Italic . . . . dialectology of Italic . evolution of Italic . . . . . . . . . Contents ix IX. Germanic 53. 54. 55. 56. 57. 58. 59. The The The The The The The documentation of Germanic phonology of Germanic . . morphology of Germanic . syntax of Germanic . . . . . lexicon of Germanic . . . . dialectology of Germanic . evolution of Germanic . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 875 888 913 954 974 986 1002 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1028 1037 1080 1097 1115 1132 1146 X. Armenian 60. 61. 62. 63. 64. 65. 66. The The The The The The The documentation of Armenian . . . phonology of Classical Armenian morphology of Armenian . . . . syntax of Classical Armenian . . lexicon of Armenian . . . . . . . dialectology of Armenian . . . . evolution of Armenian . . . . . . XI. Celtic 67. 68. 69. 70. 71. 72. 73. The The The The The The The documentation of Celtic phonology of Celtic . . morphology of Celtic . . syntax of Celtic . . . . . lexicon of Celtic . . . . dialectology of Celtic . evolution of Celtic . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1168 1188 1203 1218 1250 1264 1274 documentation of Tocharian phonology of Tocharian . . morphology of Tocharian . syntax of Tocharian . . . . . lexicon of Tocharian . . . . dialectology of Tocharian . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1298 1304 1335 1352 1365 1389 XII. Tocharian 74. 75. 76. 77. 78. 79. The The The The The The XVI. Languages of fragmentary attestation 101. Phrygian 1. 2. 3. 4. Introduction Phonemic inventory Morphonology Historical development 5. Morphology 6. Syntax 7. References 1. Introduction Phrygian is an extinct Indo-European language of West and Central Anatolia, the written sources of which span the period between the 8 th century BCE and 3 rd century CE. 1.1. Greek sources refer to Phrygians either as Βρίγες (Herodian, Strabo, Stephanus Byzantinus), Βρύγες (Strabo), Βρῦγοι (Strabo), Βρίγαντες (Herodian) or as Φρύγες (Homer). According to Herodotus (VII 73), the Phrygians originally were neighbors of the Macedonians and were called Βρίγες as long as they dwelt in Europe. When they changed their home to Asia, they also changed their name. A similar account is also given by Strabo (VII 3, 2). 1.2. The time of the Phrygian migration to Anatolia is heavily debated, as is also the question of whether we can identify the Muški of Assyrian sources with the Phrygians. Homer has the young king Priam aiding the Phrygians against the Amazons (Il. III 189); in return, Phrygians come to Trojan aid (II 862 ff.). If true, these two facts would place the Phrygian migration before the collapse of the Bronze Age, i.e. the 12 th c. BCE; but the Homeric account can easily be anachronistic. At any rate, in the 8 th c. BCE, Phrygians established a powerful kingdom with the capital Gordion (Gk. Γόρδιον, now Yassıhüyük) at the river Sangarios (now Sakarya), where Alexander the Great famously severed the knot on his way to Egypt. Other ancient sites include the so-called Midas city (near Yazılıkaya in Eskişehir province), Daskyleion (near Bandırma), and Dorylaion (now Eskişehir). Thriving under the legendary king Midas, the Kingdom of Phrygia was sacked by the Cimmerians around 695 BCE and then frequently changed hands: it was first a part of Lydia (7 th−6 th c. BCE), then of the Persian Empire (6 th−4 th c. BCE) and of the Empire of Alexander (4 th c. BCE). Later, Phrygia was ruled by the Kingdom of Pergamum (2 nd c. BCE), until it was added to the Roman province of Asia during the late Republic. During the Roman period, Phrygia, lying to the east of Troas, bordered on its northern side with Galatia, on the south with Lycaonia, Pisidia, and Mygdonia and on the east, it touched upon Cappadocia. 1.3. Phrygian is most closely related to Greek. The two languages share a few unique innovations, such as the vocalization of the laryngeals (4.3), the pronoun auto- (5.2) and https://doi.org/10.1515/9783110542431-022 101. Phrygian 1817 the 3sg. imperative middle ending (5.3). It is therefore very likely that both languages emerged from a single language, which was spoken in the Balkans at the end of the third millennium BCE. 1.4. Written in two distinct scripts − one native and the other Greek − Phrygian inscriptions can on the whole be divided into two corpora: the Old Phrygian (OPhr.) corpus written in the native script, and the New Phrygian (NPhr.) corpus written in the Greek script. Old Phrygian, as opposed to New Phrygian, is customarily romanized with the exception of the disputed signs ↑, Φ and Ψ. 1.5. The native script is an alphabet consisting of 21 characters: A B G D E V I J K L M N O P R S T U C F X a b g d e v i y k l m n o p r s t u ↑ Φ Ψ Similar to the archaic Greek alphabets, the native script is essentially distinguished by the arrow and the yod. The last two letters of the table above, which look like Greek phi and psi, are very rare. Φ occurs only once as a variant of the arrow, while Ψ (ten occurrences) most probably stands for /ks/. The yod does not appear in the oldest OPhr. inscriptions and was introduced somewhere during the 6 th c. BCE (Lejeune 1969), first in prevocalic and word-final positions (e.g., areyastin, kuryaneyon, yosesait; tedatoy, aey, materey, avtay, etc.), later also as a second element of i-diphthongs (ayni, ktevoys, etc.; Lubotsky 1993). Most inscriptions from the North-West of Phrygia (Vezirhan, Daskyleion, etc.) show some deviations from the usual OPhr. alphabet. The yod has a different shape, and there are two types of s, usually transcribed as s and ś (for an overview and discussion of these peculiarities, see Brixhe 2004: 26−32). Since these inscriptions normally lack the arrow sign, it seems reasonable to assume that ś and the arrow indicated the same sound. Words are often separated by a colon consisting of 2, 3, or more vertical dots and occasionally by spaces. About two thirds of the OPhr. inscriptions run from left to right (dextroverse) and one third from right to left (sinistroverse); a few are written boustrophedon. In NorthWest Phrygia, however, the proportion is exactly the opposite, two thirds of the inscriptions being sinistroverse. The OPhr. corpus currently comprises more than 400, unfortunately mostly very short and fragmentary, inscriptions and dates from the 8 th to the 4 th c. BCE; ca. one fifth of the inscriptions are on stone and the rest on pottery or other small objects. The inscriptions are found across a huge area, far outside Phrygia proper: as far east as Boğazköy and Tyana (Hittite Tuwanuwa), as far south as Bayındır (near Antalya) and as far west as Daskyleion. The largest number of inscriptions comes from Gordion (ca. 80 %). The standard edition of the OPhr. corpus is Brixhe and Lejeune (1984). The inscriptions are cited by the region where they are found and by a number. Each inscription is hence assigned a siglum: B − Bithynia; G − Gordion; P − Pteria; M − Midas City; T − Tyana; W − West Phrygia; HP (i.e. hors de Phrygie) − from outside of Phrygia; NW − North West Phrygia (Dorylaion); Dd (i.e. documents divers) − of unknown origin. The corpus continues to be updated by means of supplements (Brixhe 2002, 2004). 1818 XVI. Languages of fragmentary attestation 1.6. NPhr. inscriptions are written in the Greek alphabet, of which only 21 characters are used: <α, β, γ, δ, ε, ζ, η, ι, κ, λ, μ, ν, ξ, ο, π, ρ, σ, τ, υ, ψ, ω>. Greek aspiratae are notably absent, except for Greek names (e.g., Αδιθρερακ, dat.sg. Κλευμαχοι) and loanwords (e.g., dat.sg. θαλαμειδη ‘sepulchral chamber’). The letters ξ and ψ are very rare (found only in the name Ξευν- and υψοδαν ‘above’, respectively), while eta and omega are practically confined to final syllables. New Phrygian by default does not practice word separation. Dating from the 2 nd−3 rd c. CE, the NPhr. corpus currently comprises 113 inscriptions, all of them found in the highlands roughly between Eskişehir and Konya. They are numbered from 2 to 129: occasionally, a number is skipped since certain inscriptions are in the meantime considered Greek. A new edition of NPhr. inscriptions is a desideratum. The largest collection (up to No. 110) is presented in Haas (1966: 114−129); editions of Nos. 111−129 are scattered across various publications (Nos. 111−114 = Brixhe 1978a: 3−7; No. 115 = Brixhe and Waelkens 1981; No. 116 = Brixhe and Neumann 1985; No. 117 = Laminger-Pascher 1984: 35; No. 118 = Mitchell 1993: 186, fig. 33; Nos. 119− 125 = Brixhe and Drew-Bear 1997; Nos. 126–128 = Drew-Bear, Lubotsky, and Üyümez 2008; No. 129 = Brixhe and Drew-Bear 2010; cf. also an overview in Brixhe 1999). Typically opening with ιος νι σεμουν κνουμανει κακουν αδδακετ ‘whoever inflicts harm upon this grave’, NPhr. inscriptions usually consist of a curse following a Greek epitaph, but there are a few Phrygian epitaphs, too. 1.7. As expounded in the preceding sections, the chronological difference between the OPhr. and the NPhr. corpora is normally matched by the use of different alphabets: the native alphabet in the case of OPhr. inscriptions and the Greek alphabet in the case of NPhr. inscriptions. There is, however, one exception: the Dokimeion inscription from the 4 th century BCE, which most probably represents an epigram, is written in the Greek alphabet. This must no doubt be due to the increased influence of Greek during Alexandrian times. 1.8. Apart from the inscriptions, Phrygian words are known from Greek sources as well. Plato (Kratylos 410a) quotes πῦρ ‘fire’, ὕδωρ ‘water’ and κύνες ‘dogs’ as shared lexical items. The dictionary of Hesychius quotes some forty words and names with a remark like Φρύγες οr παρὰ Φρυξί, e.g. γλούρεα· χρύσεα. Φρύγες; ζέμελεν· βάρβαρον ἀνδράποδον. Φρύγες; Μαζεύς· ὁ Ζεὺς παρὰ Φρυξί, etc. These glosses are of questionable value, however. The remark “Φρύγες” does not guarantee Phrygian provenance of the gloss, because it could also refer to Anatolians, or even to foreigners in general. Of these glosses, βέκος ‘bread’ (also mentioned by Herodotus II 2) is arguably the most famous one. 2. Phonemic inventory 2.1. Vowels − OPhr. /a/, /e/, /o/, /i/, /u/ − NPhr. /a/, /e/, /o/, /i/, /u/ /ā/, /ō/, (/ī/, /ū/ ?); (/ē/, /ō/ ?) 101. Phrygian 1819 Although vowel length is not expressed in writing, Old Phrygian must have had long vowels, at least /ō/, as follows from the fact that OPhr. <o> has two different reflexes in NPhr.: ο or ου, depending on its origin. OPhr. o that goes back to Proto-Indo-European (PIE) *o corresponds to NPhr. ο: OPhr. yos, ios, relative pronoun < PIE *ios ~ NPhr. ιος; OPhr. 3sg. med. ending -toi, -toy < PIE *-toi ~ NPhr. -τοι. On the other hand, OPhr. o that goes back to PIE *ō corresponds to NPhr. ου: OPhr. 3sg. med. imperative ending -do < *-sdhō (cf. Gk. -σθω) ~ NPhr. -δου; OPhr. dat.sg. ending -oi, -oy < PIE *-ōi (cf. Gk. -ῳ) ~ NPhr. -ου. Presumably, there was no OPhr. phoneme /ē/, since PIE *ē and *eh1 merged with *eh2 into OPhr. a, cf. OPhr. matar nom.sg. ‘mother’ < PIE *meh2tēr, NPhr. αναρ ‘man’ < *-ēr, OPhr. daΨet ‘to place, make’ < *deh1-k-, etc. There are no certain examples of OPhr. /ī/, /ū/, so that their existence remains hypothetical. During the NPhr. period, short and long vowels of OPhr. apparently merged and gave rise to a vowel system without a length opposition: /a/, /e/, /o/, /i/, /u/, at least in initial syllables (thus already Brixhe 1990: 98). The absence of long vowels is further indicated by the use of the hexameter in New Phrygian, where we only find dactylic rhythm and where the function of long vowels was taken by vowels which are long per positionem and by diphthongs (Lubotsky 1998). The status of NPhr. η and ω, which mostly occur in final syllables, is unclear, but they might have represented close long [ẹ̄] and [ọ̄]. NPhr. η most often appears in the final syllable: in the dat.sg. ending of the consonant stems, where it varies with -ε/-ι and -ει, cf. κνουμανη dat.sg. ‘grave’ ~ κνουμανε(ι), κνουμανι, Τιη dat.sg. ‘Zeus’ ~ Τι(ε), Ξευνη PN ~ Ξευνε; in the ending -αης, cf. δεκμουταης ~ δεκμουταις; in the ending -ης, cf. πατερης (No. 98) nom.pl. ‘parents’ (< *-eies), for NPhr. μανκης (No. 86) see 5.1; the function and meaning of δ[α]κερης (No. 116) and παρτης has not yet been clarified. A few times NPhr. η is found in prevocalic position: μαιμαρηαν, τιηιον, εκατηας. For a discussion of this grapheme, see Lubotsky (1998). In contemporary Greek, η had already merged with ι. NPhr. ω is confined, with very few exceptions, to the dat.pl. ending -ως, which goes back to PIE *-ōis (this ending is spelled with <ο> only three times). It typically occurs in the formula με ζεμελως κε δεως κε ‘among men and gods’. The Phr. short diphthongs are: /ey/ = <ey, ei, ει >, /ew/ = <ev, ευ>, /oy/ = <oy, oi, οι >, /ay/ = <ay, ai, αι >, and /aw/ = <av, αυ>. The existence of the diphthong /ow/ is uncertain. In NPhr., it would at any rate be indistinguishable from ου = /u/. In OPhr., we find it once in the nom.sg. Vasous PN (P-03), next to Vasus (P-05) < *u̯asōus (?), and once in final position in otekonov (B-01). This enigmatic ending -ov is reminiscent of forms like tubetiv and derạliv (B-05) or apelev (B-07) and is likely to be due to a dialectal North-Western development. Besides short diphthongs, there must have been at least two long diphthongs in OPhr., that is, /ōy/, cf. OPhr. o-stem dat. sg. -oi, NPhr. -ου < PIE *-ōi (see above), and /āi/, cf. OPhr. ā-stem dat.sg. -ai, NPhr. -α (see further 4.1). 2.2. Consonants − /p/ = <p, π>; /t/ = <t, τ>; /k/ = <k, κ> − /b/ = <b, β>; /d/ = <d, δ>; /g/ = <g, γ> 1820 − − − − − XVI. Languages of fragmentary attestation /z/ or /dz/ = <↑, ζ> /s/ = <s, σ> /m/ = <m, μ>; /n/ = <n, ν> /l/ = <l, λ>; /r/ = <r, ρ> /w/= <v, ο(υ)>; /y/ = <y, i, ι > The phonological interpretation of the OPhr. arrow <↑>, which is probably identical with <ś> (see 1.5), and of NPhr. <ζ> remains controversial. Since OPhr. <↑> only occurs before front vowels (i, e), it is likely that this letter represents a sound which arises through palatalization. In NPhr. inscriptions, <ζ> appears in two words, ζεμελως dat.pl. ‘men’ (< PIE *dhǵ(h)emelo-, cf. Gk. χθαμαλός, Lat. humilis ‘low, humble’) and ζειρα(ι) of unknown meaning and function. If OPhr. śirạy = ↑irạy of the Vezirhan inscription is the same word as NPhr. ζειραι, we have to assume that NPhr. <ζ> and the OPhr. arrow indicated the same sound, presumably a voiced affricate or /z/. For a possible voiceless geminate nn, see 4.2. In the Greek alphabet, /w/ is written either <ου> or simply <ο> in prevocalic position, cf. ουεναουιας, ουανακταν (No. 88), οαν οε αυται (No. 116), κοροαν (W-11). 3. Morphonology Morphemes, both suffixes and roots, show ablaut. In nominal inflection of the consonant stems, the suffix changes its vocalism, e.g. nom.sg. -tar (<*-tēr) vs. obl. -ter- in matar, materey, materan ‘mother’; nom.sg. -an (< *-ēn) vs. obl. -en- in ορουαν, ορουενος, ορουεναν ‘warden’; iman, i(n)meney, imenan ‘monument’; nom.sg. -ōu- vs. obl. -u- in Vas(o)us, Vasos (< *u̯asu̯os). Examples of ablaut in the root are: da- < *dheh1- ‘place, do’ (t-e-da-toy, e-daes, αδδακετ) vs. de- <*dhh̥1- (δετο(υ)ν ‘monument’, a to-participle); teik- < *deik̑- ‘show’ (ιστεικετ, thus to be read in No. 88, cf. Brixhe 1999: 304, fn. 46 ~ Gk. ἐκ-δείκνυμι) vs. tik- < *dik̑- (τιτ-τετικμενος ‘condemned’ ~ Gk. δια-δικάζω ‘I judge’). In Phrygian, word final *-on is raised to -un, for instance, in the acc.sg. ending of ostems, cf. OPhr. acc.sg. akaragayun (M-02) ‘part of the monument’, avtun ‘himself’ (W01b), NPhr. κακουν ‘damage, wrong’. The latter word often appears as κακον in NPhr. inscriptions and sometimes as κακιν, κακων. A parallel raising of e to i before nasals is possibly attested in OPhr. iman, imen- ‘monument’, if we assume with Vine (2010) that it goes back to *en-mēn, en-men-os (~ Gk. ἐμμενές ‘continuously’), and in NPhr. πινκε (No. 116), if it means ‘five’, PIE *penk we. Further, o was raised to u in the position before ri̯ , li̯ , cf. OPhr. kuryaneyon (W-01c), which was borrowed from Gk. κοιρανέων ‘giving orders; ruling’ < *kori̯ - in Mycenaean times (Lubotsky 1988: 23). Another development in word-final position is *-ans > -ais, which follows from the inflection of titles or patronymics in -evais (arkiaevais, memevais, kanutievais): nom.sg. -evais, gen.sg. -evanos < *-evans, -evanos, most probably going back to *-eu̯ants, *-eu̯antos < *-eu̯n̥ts, *-eu̯n̥t-os (for the development of *-nt- see 4.2). For a parallel, cf. Greek Lesbian ταις < *τανς. It is further attractive to assume that the ending OPhr. -ais, NPhr. -αης, -αις is acc.pl. in some contexts and reflects PIE *-ns (Brixhe 2004: 41−42); similarly, OPhr. -ois can go back to *-ons. 101. Phrygian 1821 Final clusters are reduced, cf. OPhr. °vanak nom.sg. ‘king’ < *-kts, cf. OPhr. dat.sg. vanaktei, NPhr. acc.sg. ουανακταν, Gk. ἄναξ, -κτος ‘lord, master’; Βας nom.sg. (name of a deity) < *-ts (acc.sg. Βαταν), ας prep. ‘to, towards’ < *-ts (= αδ + s, cf. Gk. εἰς beside ἐν and ἐξ beside ἐκ); NPhr. δακαρεν 3pl. ‘they made erect’ (No. 98) < *-nt; 3sg. aor. ending OPhr. -es, NPhr. -ες < *-est. The vowels /e/ and /i/ show some vacillation, presumably in pretonic position, both in Old and New Phrygian, cf. kubeleya (B-01) and kubileya (W-04) ‘Cybele’ (epithet of the Mother Goddess), δεως (passim) and διως (Nos. 4, 5, 39), δυως (No. 113) dat.pl. ‘gods’; αββερετορ (Nos. 73, 75) and αββιρετο (No. 25). In clusters consisting of a dental and a stop, the dental becomes completely assimilated to the stop. The resulting geminate is often simplified in NPhr., cf. α(β)βερετ (αδ°) ‘bring’, τιγ-γεγαριτμενος (τιτ°) ‘devoted’. Also other geminates are regularly simplified, cf. α(δ)δακετ (αδ°) ‘inflict’, τι(τ)-τετικμενος (τιτ°) ‘condemned’. In external sandhi, in prepositional phrases, we encounter the same results, cf. NPhr. α(κ) κε οι and α(τ) τιε (for αδ). This loss of contrastive gemination has led to hypercorrect spellings like κνουμμανει, κνουνμανει for κνουμανει or αινι μμυρα for αινι μυρα. More controversial is the assimilation of word-final -s to a following velar, but there are a few good examples in NPhr., cf. -s k- > -k k-: αδιθρερακ ξευνεοι (No. 31), ικ κναικαν (No. 116); -s g- > -k g-: ποκ γονιον (No. 116), presumably via -h k-, -h g-. 4. Historical development 4.1. Vowels The Indo-European vowels seem to be well preserved, except for the changes already mentioned in the preceding sections. Here are a few more examples of vocalic phonemes. − *i : OPhr. kin, NPhr. κιν ‘which’ < PIE *k wim; NPhr. γεγαριτμενος ‘devoted’ < PIE *g̑hrHit- (Gk. ἐν-κεχαρισμένος); − *e : OPhr. ke, NPhr. κε ‘and’ < PIE *k we; NPhr. αββερετ, μεβερετ < PIE *bher-; − *o : OPhr. -os, NPhr. -ος, nom.sg. m. of the o-stems < PIE *-os; − *u : NPhr. (ο)υψοδαν adv. ‘above; on the top’ < PIE *(H)upsodhn̥ (cf. Gk. ὑψόθε[ν] ‘[from] above’); NPhr. κνουμαν- n. ‘grave’ < PIE *knu- (cf. Gk. κνύω ‘I scratch’); − *a (*h2e): NPhr. αδ preverb ‘to, at, by’ < PIE *h2ed (cf. Lat. ad ‘id.’); − *ē : NPhr. ορουαν nom.sg. ‘father, guardian’ (gen.sg. ορουενος; acc.sg. ορουεναν) < PIE *soru̯ēn (cf. Gk. οὖρος ‘watcher, guardian’); − *eh1 : NPhr. (αδ)δακετ 3sg. ‘inflicts’ < PIE *dheh1-k- (cf. Gk. aor. ἔθηκα); − *eh2 : NPhr. βρατερε dat.sg. ‘brother’ < PIE *bhreh2-ter- (cf. Skt. bhrā́tar-, Lat. frāter); − *eh3 : NPhr. acc.sg. μουρου[ν] (No. 100), acc.pl. n. μμυρα (No. 25) ‘stupidity’, cf. Gk. μῶρος, μωρός ‘stupid’. As far as we can see, the diphthongs remain unchanged in Old Phrygian, but in New Phrygian the long diphthongs /āi/ and /ōi/ often lose their second element in final position, while word-final /ei/ gradually becomes monophthongized and is then written as <-ε, -ι, -η>. PIE *-ōis shows a special development to NPhr. -ως: 1822 XVI. Languages of fragmentary attestation − *h2ei/*eh2 i : OPhr. ai, NPhr. αι ‘if’< PIE *h2ei (cf. Gk. Aeol., Dor. αἰ ‘if’); OPhr. ayni, NPhr. αινι ‘and/or’ < PIE *h2ei-ni; NPhr. κναικαν acc.sg. f. ‘wife’ < PIE *g wneh2 ikm̥ (cf. Gk. γυναῖκα); − *h2eu : OPhr. avtoi dat.sg.m., NPhr. αυτος ‘self’ < PIE *h2euto- (cf. Gk. αὐτός ‘self’); − *ei : NPhr. γεγρειμεναν acc.sg. f. ‘written’ < PIE *ghreiH- (cf. Gk. χρίω ‘I touch’); OPhr. dat.sg. ending of the consonant stems, e.g., materey ‘mother’, Tiei ‘Zeus’ (NW101), NPhr. Τιε, Τι, Τιη dat.sg. ‘Zeus’, κνουμανει, -ε, -ι, -η dat.sg. ‘grave’ < PIE *-ei (cf. Lat. -ei, -ī); − *eu : OPhr. bevdos acc.sg. n. ‘statue, image’ (B-01) < PIE *bheudhos; − *oi : OPhr. 3sg. med. ending -toi, -toy, NPhr. -τοι < PIE *-toi; NPhr. τετικμενοι nom.pl. m. ‘condemned’ < PIE *-oi; − “*āi” : OPhr. ā-stem dat.sg. -ai (Midai, Atai), dat.sg. f. pron. °esai-t (W-01b), NPhr. dat.sg. f. dem. pron. σα(ι), pron. αυται, dat.sg. f. μανκα(ι) ‘stele’ < PIE *-eh2ei, cf. Gk. -ᾱι, -ηι, Lat. -ae; − *ōi : OPhr. o-stem dat.sg. ending -oi, -oy, NPhr. -ου < PIE *-ōi (cf. Gk. -ῳ); NPhr. ostem dat.pl. ending -ως < PIE *-ōis. 4.2. Resonants Consonantal resonants have undergone few changes. Word-final /m/ and /n/ have merged into /-n/ in Phrygian, just as in Greek, cf. OPhr. o-stem acc. sg. ending -un, NPhr. -ουν, -ον < PIE *-om. Possibly,*u̯ was lost before a following*o in Phrygian, cf. OPhr. nom.sg. vas(o)us PN (P-03, P-05), gen.sg. vasos (P-02) < *u̯asu̯os (Brixhe 1990: 65). The apparent counterexamples, OPhr. tovo and devos, go back to *toho < *toso and *dehos < *dhh̥1sos, respectively, where -v- is a Hiatustilger. The development of the cluster *nt in Phrygian is unclear. First of all, it is remarkable that this cluster is very rare in Phrygian texts: among well-attested words we find only the possible borrowings OPhr. panta (B-05.4), παντης (W-11), NPhr. παντα (No. 35) ~ Gk. πᾶς, παντ- ‘all, every’ and NPhr. Πουντας (No. 48) ~ Gk. Πόντος ‘Pontic region’ (Lubotsky 1997: 123 with refs.). On the other hand, the ending of the 3pl. imperative, which presumably goes back to *-ntō (parallel to 3sg. impv. ειτου < *-tō), is spelled in NPhr. as -ττνου (αδειττνου No. 12) and -ννου (ιννου Nos. 35, 71). These spellings may point to a voiceless geminate nn, IPA [n̥n̥]. Also the OPhr. spellings tn, ntn, found in apaktneni (B-01.8), ẹventnoktoy (B-06), seem to point in this direction (cf. Lubotsky 1997: 121−122). However, Annelies Hämmig points out to us (p.c.) that αδειττνου in No. 12 must rather be read αδειννου, which would mean that *-nt- > -nn- in Phrygian. See further 3 on OPhr. -evanos < *-eu̯antos < *-eu̯n̥tos. The vocalic nasals have become aN, cf. OPhr. onoman acc.sg. n., NPhr. ονομαν‘name’ < PIE *h̥3nh̥3mn̥ (cf. Gk. ὄνομα ‘id.’); NPhr. κναικαν acc.sg. f. ‘wife’ < PIE *g wneh2 ikm̥. The reflexes of vocalic *r̥ and *l̥ are less certain. OPhr. por, NPhr. πουρ prep. ‘for’ < PIE *pr̥ (cf. Gk. πάρ, Goth. faur ‘id.’) seems to indicate that *r̥ has developed into *or, but this is the only example. For NPhr. γεγαριτμενος ‘devoted, at the mercy of ’ < PIE *g̑hr̥Hit- see the next section. 101. Phrygian 1823 4.3. Laryngeals Vocalization of the Indo-European laryngeals shows the same “triple representation” as in Greek and, being a common innovation of the two branches, it is an important indication of the dialectal position of Phrygian. Initial laryngeals develop a prothetic vowel, i.e. *h1C- > eC-, *h2C- < aC-, *h3C- > oC-: OPhr. eu- ‘well’ (?) < PIE *h̥1su- (cf. Gk. εὐ-, Skt. su- ‘id.’); NPhr. αναρ m. ‘husband’ < PIE *h̥2nēr (cf. Skt. nár-, Gk. ἀνήρ ‘id.’); OPhr. onoman acc.sg. n., NPhr. ονομαν- ‘name’ < PIE *h̥3nh̥3mn̥ (cf. Gk. ὄνομα, Skt. nā́man- ‘id.’). In a similar fashion, interconsonantal laryngeals are vocalized to e, a, o, respectively: NPhr. δεως instr.pl. m. ‘god’ < PIE *dhh̥1so- (cf. Gk. θεός ‘id.’); NPhr. δετουν m./n. ‘monument’ < PIE *dhh̥1to-; OPhr. -meno-, NPhr. -μενο- middle ptc. < PIE *-mh̥1no- (cf. Gk. -μενο-); NPhr. πατερης nom.pl. ‘parents’ < PIE *ph̥2ter- (cf. Gk. πατήρ ‘id.’); NPhr. τιτ-τετικμενα nom.pl. n. ‘condemned’ < PIE *-h̥2 (cf. Gk. -α, Lat. -a, Skt. -i); OPhr. onoman, NPhr. ονομαν- ‘name’ < PIE *h̥3nh̥3mn̥. Also in other positions, the development of the laryngeals in Greek and in Phrygian is identical, cf. NPhr. γεγαριτμενος ‘devoted, at the mercy of’ < PIE *g̑hr̥Hit- (cf. Gk. ἐν-κεχαρισμένος ‘id.’, χάρις, χάριτος ‘love’); NPhr. γλουρεος ‘golden (?)’ (for the meaning, cf. the above-mentioned gloss by Hesychius γλούρεα· χρύσεα, Φρύγες ‘golden items [Phrygian]’) < PIE *g̑hl̥ h3-ro- (cf. Gk. χλωρός ‘green’). 4.4. The single Phr. fricative /s/ is practically restricted to word-final position and to clusters with a stop, cf. OPhr. o-stem nom.sg. -os, NPhr. -ος < PIE *-os, NPhr. 3sg. saor. εσταες ‘established’, OPhr. 3sg. subj. daΨet /dakset/ ‘will do’. In other positions, word-initially and intervocalically, it was lost, cf. NPhr. ορουαν ‘warden’ < PIE *soru̯ēn; OPhr. egeseti, NPhr. εγεσιτ, εγεδου ‘hold, experience’ < PIE *seg̑h-; NPhr. dat.pl. δεως ‘god’ < PIE *dhh̥1so-. PIE *s was further lost in the clusters *su̯- and *-sdh-, cf. OPhr. ven- ‘self’, NPhr. nom.pl. n. ουα ‘own’ < PIE *su̯e/*su̯o- and impv. ending -do, -δου < PIE *-sdhō. The intervocalic /s/ in the s-subjunctives OPhr. egeseti, NPhr. εγεσιτ, mentioned above, has probably been generalized from postconsonantal positions, just like in Greek. 4.5. Stops It is clear that PIE tenues are reflected as Phrygian tenues, and mediae aspiratae as PIE mediae, cf. OPhr. 3sg. primary act. -ti, NPhr. -τι < PIE *-ti; NPhr. πατερης nom.pl. ‘parents’ < PIE *ph̥2ter-; NPhr. dat.pl. δεως ‘god’ < PIE *dhh̥1so-; NPhr. acc. sg. γεγρειμεναν ‘written’ < PIE *ghreiH-; NPhr. βρατερε dat.sg. ‘brother’ < PIE *bhreh2ter-, etc. The fate of PIE mediae is more controversial, but there is a growing body of evidence that they have become Phrygian tenues (cf. Lubotsky 2004 for more examples and a discussion of the counterevidence), cf. NPhr. acc.sg. Τιαν, gen. sg. Τιος, dat.sg. Τι(ε), OPhr. Tiei ‘Zeus’ < PIE *diēm, *diu̯os, *diu̯ei; NPhr. acc.sg. κ̣ναικαν ‘wife’ < PIE *g wneh2 ikm̥ (cf. Gk. γυναῖκα); OPhr. torv- (B-05) ‘wood’ < PIE *doru̯-/*dr̥u̯-; NPhr. (τιτ-)τετικμενος ‘condemned’ < PIE *deik̑-, cf. Gk. δια-δικάζω ‘I judge’, κατα-δικάζω ‘I condemn’. PIE labiovelars have lost their labial feature, cf. OPhr. ke, NPhr. κε ‘and’ < PIE *k we, NPhr. acc. sg. κ̣ναικαν ‘wife’ (116) < * g wneh2 ikm̥. 1824 XVI. Languages of fragmentary attestation In view of the close relationship of Phrygian and Greek, it is likely that Phrygian is a centum language, too, cf. OPhr. egeseti, NPhr. εγεσιτ, εγεδου ‘hold, experience’ < PIE *seg̑h-; NPhr. (τιτ-)τετικμενος ‘condemned’ < PIE *deik̑-; NPhr. γεγαριτμενος ‘devoted, at the mercy of’ < PIE *g̑ hr̥Hit-; NPhr. γλουρεος ‘golden (?)’ < PIE *g̑hl̥ h3-ro-. This implies that ζεμελως dat.pl. ‘men’ (< PIE *dhg̑(h)emelo-) must be due to a special development of the initial cluster and that the Phrygian demonstrative pronoun s- (OPhr. acc.sg.n. si, acc.sg.m. sin; NPhr. gen.sg. f. σας, dat.sg. σα(ι), dat.sg.n. σεμουν, see 5.2) must reflect PIE *k̑i̯ - with palatalization (as indicated above, 4.4, PIE initial *s- shows a zero-reflex in Phrygian). 5. Morphology 5.1. Nouns Phrygian nouns are inflected for case, gender and number. There are at least 4 cases: nominative, accusative, genitive and dative; other cases, possibly unidentified, could have existed as well; cf., for instance, the puzzling NPhr. κναικο ‘wife’ (No. 116) or kạṿarmọyo (B-01) next to acc.sg. kavarmoỵun in the same inscription. There are three genders (masculine, feminine, neuter) and two numbers (singular and plural). Nominal stems can be divided into o-stems, ā-stems and consonant stems (C-stems). Poorly attested stems include i-, u- and “e”-stems (i.e. Anatolian names in -es, like Ates, Bateles, Iktes). We make no distinction between substantives and adjectives, since their inflection is identical. When we cite the actually attested forms and inscriptions, we use brackets as follows: [ ] = reconstructed portion of the text, < > = omitted portion of the text, ( ) = mistake of the engraver. Damaged letters are indicated by a subscript dot. singular o-stems OPhr. Nom. Gen. Dat. Acc. -os ? -ovo -oi, -oy -un ā-stems NPhr. OPhr. -ος -ου -ου -ουν, -ον -a f., -a(s) m. − -ai, -ay -an C-stems NPhr. OPhr. NPhr. -α f. -ας -αι, -α -αν -s, -0̸ -os -ei, -ey -an, -0̸ [n.] -ς, -0̸ -ος -ε(ι), -ι, -η -αν, -0̸ [n.] -ας − − -αις, -αης -a [n.] − − ? -ais -ης − − ? -αης, -αις plural Nom. Gen. Dat. Acc. -oi − ? -oys ? -oys, -a [n.] -α [n.] -ουν -ως α [n.] − − − − Nominative singular: − o-stems: akenanogavos title (M-01a), τιττετικμενος ‘condemned’ (passim) < PIE *-os; 101. Phrygian 1825 − ā-stems: Kubeleya ‘Cybele’ (B-01), OPhr. μανκα ‘stele’ (W-11) < PIE *-eh2; Midas (M-01d), the name of the second king of Phrygia, most probably of Anatolian origin; other Anatolian names appear both with and without -s in the nominative, cf. Baba (M-01b) next to Babas (G-06), Kaliya (B-05), but Kuliyas (G-127), etc.; PN in -es always have a sigmatic nominative: Ates (M-01a), Bateles (W-08), Eies (G-108), Iktes (G-02), etc.; − C-stems: the sigmatic nominative is attested with the i- and u-stems Ṭuvatis PN (G133), Alus PN (W-09), Vasous PN (P-03) next to Vasus (P-05) < *u̯asōus; with stems in stops: Manes PN (B-07), Βας ‘Bat (name of deity)’ (No. 99) < *-ts, Modrovanak ‘king of Modra’ (M-04) < *-kts and with OPhr. patronymics in -evan-: arkiaevais (M01a), kanutievais (P-03) < *-u̯ans < *-u̯n̥ts; the regular r- and n-stems have an asigmatic nominative: matar ‘mother’ (W-04), αναρ ‘man’ (No. 15) < *-ēr; iman ‘monument’, ορουαν ‘father, warden’ (No. 48) < *-ēn; kuryaneyon ‘commander’ (W-01c) is borrowed from Greek. Genitive singular: − o-stems: ?αργου ‘because of’ (No. 30); the ending is pronominal, cf. OPhr. tovo (G02c), NPhr. του (No. 87); Atevo PN (W-10) is probably gen.sg. of Ates, with an ending analogical to o-stems; − ā-stems: Ουεναουιας PN (No. 88) < PIE *-eh2es; the interpretation of μανκης ‘stele’ (No. 86), which is used in the function of a dative, is uncertain: genitive (pro dat.) or, rather, dat.pl.?; − C-stems: Τιος ‘Zeus’ < PIE *diu̯os (with loss of -u̯- before o); Vasos PN (P-02) < *u̯asu̯os (idem); kanutiievanoṣ title/patronymic (P-02), ορουενος ‘father, warden’ (No. 106); Aṛtimitos ‘Artemis’ (B-05), Manitos ‘Manes’ (B-07) < PIE *-os. Dative singular: − o-stems: adoikavoi PN (G-02a); κορου ‘ground for the grave’ (No. 92), a loanword from Gk. χῶρος; σορου ‘sarcophagus’ (Nos. 21, 124), probably borrowed from Gk. σορός < PIE *-ōi; − ā-stems: dumeyay adj.f. ‘of the religious community’ (G-01a); μανκα(ι) ‘stele’ < PIE *-eh2ei; cf. also midai ‘Midas’ (M-01a); − C-stems: Tiei (NW-101), Τι(ε), Τιη dat.sg. ‘Zeus’ < PIE *diu̯ei (with analogical loss of *-u̯- due to leveling with other cases); materey ‘mother’ (W-01b), inmeney (B-05) ‘monument’, βρατερε ‘brother’ (No. 31); μα̣τ̣[ε]ρε (thus to be read in No. 129, instead of μα̣γ̣ρε of the edition); κνουμανε(ι), -η, -ι ‘grave’, δουμ(ε) ‘religious community’ (No. 48); vanaktei ‘king’ (M-01a) < PIE *-ei. Accusative singular: − o-stems: akaragayun ‘part of the monument’ (Μ-02), δετον̣ (No. 116) and δετουν (No. 31) ‘monument’ < PIE *-om; − ā-stems: ạkinanogavaṇ title (M-04), κοροαν ‘girl’ (W-11), μανκαν ‘stele’ (No. 15) < PIE *-eh2m; − C-stems: areyastin epithet of Cybele (W-01a), ευκιν ‘vow’ (No. 30), possibly a loanword from Gk. εὐχή ‘id.’ < PIE *-im; Τιαν ‘Zeus’ < PIE *diēm; materan ‘mother’ (W-01a) < PIE *-er-m̥; imenan (B-05) ‘monument’, ορουεναν ‘warden’ (No. 128) < PIE *-en-m̥; Batan (T-02b), Βαταν (No. 33) ‘Bat’, duman ‘religious community’ (B01), ουανακταν ‘king’ (No. 88), κ̣ναικαν ‘wife’ (No. 116) < PIE *-m̥. The neuters 1826 XVI. Languages of fragmentary attestation keneman ‘niche (?)’ (M-01), κνουμαν ‘grave’ (No. 31); bevdos ‘image, statue’ (B01), βεκος ‘bread’, have a zero ending. Nominative plural: − o-stems: τετικμενοι ‘condemned’ (No. 71) < PIE *-oi of pronominal origin; n.pl. τετικμενα (No. 12) < PIE *-h̥2; ā-stems: ουελας ‘relatives (?)’ (No. 120) < PIE *su̯el-eh2-es; − C-stems: πατερης ‘parents’ (No. 93), most probably reflecting PIE *-eies. Neuter kena ‘generation’ (No. 35), if correctly analyzed, < PIE *g̑enh1es-h̥2. Genitive plural: − o-stems: τετουκμενουν (No. 28) < PIE *-ōm, although the interpretation of the final part of this inscription (ιος νι σεμουν κνουμανε κακουν αδακετ ιος τιτετουκμενουν ειτου) is far from certain. Dative plural: − o-stems: δεως ‘god’ (No. 40) < PIE *-ōis. Accusative plural: − o-stems: kṭevoys ‘property (?)’ (B-01), pạtriyiọis ‘paternal (?)’ (B-04) < PIE *-ons; neuters kạka ‘harm’ (B-05), μμυρα ‘stupidity’ (No. 25) < PIE *-h̥2; − ā-stems: δεκμουταις ‘?’ (No. 9), δεκμουταης (No. 31); − C-stems: ḅṛạterạis ‘brother’ (B-04) < *-ans < PIE *-n̥s. 5.2. Pronouns The proximal demonstrative pronoun (‘this here’) has the stem *se-/si- in masculine and neuter, and *sa- in feminine. Since initial PIE *s- seems to disappear in Phrygian, the stem is likely to go back to PIE *k̑i̯ - (Goth. hi-, Lith. ši-, Gk. σήμερον ‘today’ < *κi̯ άμερον, etc.) + *e-/i- (Lat. is, ea, id). The Phrygian demonstrative pronoun is often followed by an emphatic particle appearing in the inscriptions as OPhr. t, NPhr. του, το, τι, τ. The attested forms are: − acc.sg. m. sin-t (B-05) < PIE *-im, n. si (M-01b, B-01) < PIE *-id; NPhr. σεμουν (No. 31) in the function of acc. must be due to generalization of the oblique stem. − dat.sg. m./n. σεμουν, with the variants σεμον, σεμυν, σεμιν < PIE *-smōi + n (reminiscent of Greek νῦ ἐφελκυστικόν); − gen.sg. f. (pro dat.) σας, dat.sg. f. σαι or σα, acc.sg. (pro dat.) f. σαν (No. 60). In OPhr. inscription W-01b, we encounter dat.sg. f. e-sai-t (materey) ‘to this very (mother)’, with yet another pronominal stem e- added (type French celui-ci). If NPhr. ειαν (No. 31) is to be read ε(σ)αν with Neumann (1986: 81), the same pronoun is also attested in NPhr. In enclitic position, we find NPhr. dat. sg. ιοι/οι and, possibly, OPhr. yọỵ (B-05). The distribution among the two NPhr. forms is determined by the phonological context. In clear cases, οι always appears after a vowel, whereas ιοι is found after consonants. This means that we have to start with *ioi, which presumably is an enclitic dative of the type Skt. me, te <*h1moi, toi (cf. Lubotsky 1997: 126), built on the stem of the*e-/i- pronoun. 101. Phrygian 1827 The pronoun to-/ta- < PIE *to- seems to have an anaphoric function, which is most clear in relative clauses, where we often find ιος νι …, τος νι … in NPhr. malediction formulae. Other forms are less clear, cf. gen.sg. m. tovo (G-02c), του (No. 87), dat.sg. f. ται (No. 116), acc.sg. f. ταν (No. 15), acc.pl. n. ta (B-01). If the gen.sg. tovo, του is correctly identified, it probably goes back to *toso > *toho > *to-o, with v as a Hiatustilger. The relative pronoun is *io- from PIE *(h1 )io-: nom.sg. m. yos (W-01), ios (P-04a), ιος (passim), acc.sg. f. ιαν (No. 31). It also once occurs reduplicated: yosyos (B-03). The pronoun *auto- ‘self’ (< PIE *h2euto-, cf. Gk. αὐτός) inflects like a thematic adjective: nom.sg. m. αυτος (No. 33), dat.sg. avtoi (T-03); dat.sg. f. avtay (W-01b). It can be reinforced by a reflexive pronoun /we-/ < PIE *su̯e (cf. also Gk. ἑαυτόν): acc.sg. m. ven-avtun (W-01b), dat.sg. f. οε-αυται (No. 116). The same possessive pronoun may be found in OPhr. acc.sg. n. ove-vin (W-01b) < PIE *su̯in, cf. also Phr. kin below; NPhr. ουα ‘his own’ nom.pl. n. < *su̯eh2. Finally, the interrogative pronoun in indefinite function is acc.sg. n. kin (B-01), κιν (No. 100: [αι]νι κακουν κιν ‘or whatever harm’) < PIE *k wim (cf. Skt. kím). 5.3. Verbs Phrygian verbs are marked for tense, voice, and mood. Identified categories include 3 tenses (present, perfect, aorist), 2 voices (active, middle), and 4 moods (indicative, imperative, optative, subjunctive). Since the stem formation and the function of the majority of verbal forms are still unknown, they are grouped below in accordance with their endings. − 3sg. -es: edaes (passim), εδαες (2 × No. 116) ‘put, placed’; eneparkes (G-125, M01d), ενεπαρκες (No. 31) ‘engraved’; εσταες ‘erected’ (No. 31); εκανες (No. 116) ‘dug (?)’; unclear are εγ̣δ̣αες (No. 18) and δδικες (No. 31). These forms are characterized by an augment e-, which immediately precedes the root, and appear in preterital contexts, except for εγ̣δ̣αες (No. 18), but the reading of this inscription is uncertain. The ablaut of the root is ambiguous in edaes and εσταες (full or lengthened grade), but the lengthened grade is probable in eneparkes (<*pērḱ-) and εκανες (< *kēn-). It is obvious that this category goes back to the sigmatic aorist (-es < *-es-t), but details are far from clear (cf. for a discussion Lubotsky 1988: 17−18, Gorbachov 2005). − 3sg. -toi: edatoy ‘put, placed’ (B-05.2), t-edatoy (W-01a), tit-edaṭ[oy] (B-05.1); egertoy ‘?’ (W-01c); ektetoy ‘possessed’ (B-01.3); epaktoy ‘?’ (B-01.9); estatoi ‘erected’ (G-144). The augment and the preterital contexts make it probable that we are dealing with a middle counterpart of the -es-forms. The root usually has full grade, but zerograde in ektetoy (< *h1e-tk̑h1-toi). However, the ending *-toi is primary in Greek dialects and in Indo-Iranian, and its appearance in the aorist is unexpected. See further below on -etor. − 3sg. -et: dạket (B-05.11), (αδ)δακετ ‘do, inflict’, αββερετ ‘bring’ in the protasis of NPhr. maledictions ‘whoever will inflict/bring harm upon this grave’. There are two cases of με-βερετ (Nos. 86, 111), which occur in an apodosis Βας ιοι βεκος μεβερετ ‘Bat will take away his bread’. Maybe, βρειτ ‘break (?)’ in the protasis ιος κε βρειτ περβεδαν (No. 114) belongs to the same category. Because of the contexts, the -et- 1828 − − − − − − − − − XVI. Languages of fragmentary attestation forms are usually considered subjunctives, but it is by no means certain that they are morphological subjunctives. Once, in a NPhr. quasi-bilingue No. 48, αδδακετ seems to be used in parallel to the Greek aorist παρεθέμην. 3sg. -etor: αδδακετορ (Nos. 40, 63, 121) and αββερετορ (Nos. 73, 75) appear in exactly the same contexts as αδδακετ and αββερετ. Moreover, we also find three times αββερετοι (Nos. 91, 113, 129) there. The difference between forms in -et and those in -etor/-etoi is generally interpreted as a difference of voice (active vs. middle), but this leaves unexplained why active and middle forms are used in the same contexts. Probably, we must rather assume that all these forms, i.e. -et, -etor, -etoi, belong to the middle paradigm, which is further confirmed by the forms in -seti/-set. 3sg. -seti/-siti: egeseti ‘will hold, experience’ (P-04a); dedạsitiy ‘will do’ (B-05.9; thus to be read with A. Hämmig, p.c., instead of dedạpitiy of the edition); με-τοτοσσειτι ‘will give away (?)’ (No. 99) are likely to be subjunctives. These forms show that final -i has not disappeared in Phrygian. 3sg. -set/-sit: daΨet /dakset/ ‘will do’ (W-01b), εγεσιτ ‘will hold’ (No. 58) are very similar in form and function to the preceding group, and are likely to be their middle counterpart. 3sg. -oi: kakoioi (G-02c), kakuioi (P-04b) are often considered optatives to a denominative verb ‘to go bad’ (< *oit), but the syntactic analysis of these inscriptions is uncertain. 3sg. impv. act. ειτου ‘let become!’ (passim) < PIE *-tō(t), cf. Gk. -τω, Skt. gachatāt. 3sg. impv. med. lakedo (W-01b, B-03), εγεδου ‘let hold!’ (passim). The ending has a close parallel in Gk. -σθω (cf. Rix 1992: 265) and represents a common innovation of the two languages. 3pl. impv. act. αδ-ειττνου (No. 12), ιννου (Nos. 35, 71) < PIE *-ntō(t), cf. Gk. -ντω (Rhod. γραφόντω, Lac. ἀναθέντω, γραψάντω, cf. Rix 1992: 265). 3pl. ind. perf. act. δακαρεν ‘put, placed’ (No. 98) < PIE *-ēr (cf. Lat. -ēre) + an additional 3pl. ending *-ent. 3sg. -ei: aey ‘be (?)’ (W-01), etitevtevey ‘?’ (B-03) might be perfects (for a discussion see Lubotsky 1988: 17−18). Perfect middle participles are athematic and reduplicated, nom.sg. γεγαριτμενος ‘devoted, condemned’, τιτ-τετικμενος ‘id.’, acc.sg. f. γεγρειμεναν ‘written’, acc.sg. f. οπεσταμε̣να̣ν̣ ‘erected’ (No. 9; cf. also σεσταμεναν in No. 15 with restored reduplication); possibly also αργμενα ‘?’ (No. 116). For this reason, αιδομενου (No. 116) probably belongs to the system of the thematic present. 6. Syntax 6.1. Word order The unmarked word order seems to be SOV, cf. with a direct object: OPhr. baba … sikeneman edaes (M-01b) ‘Baba has established this niche (?)’; with an indirect object: OPhr. ates … midai … edaes (M-01a) ‘Ates has established for Midas’. An indirect object normally precedes a direct object, cf. OPhr. yos-esai-t materey … onoman daΨet ‘whoever would make … name for this very Mother’ or NPhr. ιος νι σα του μανκα 101. Phrygian 1829 κακουν αδδακετ ‘whoever inflicts harm upon this very monument’, ιος νι σεμουν κνουμανει κακουν αδδακετ ‘whoever inflicts harm upon this grave’. On the whole, word order in NPhr. seems to be less strict, possibly because many inscriptions are metrical, or at least go back to a metrical original (cf. Lubotsky 1998). In OPhr. inscriptions, we also encounter OSV order with topicalization, e.g., sin-t imenan kaliya titedat--- ‘this very monument Kaliya has established’ (B-05), materan areyastin bonok akenanogavọṣ vrekun tedatoy ‘Bonok, the high priest, has established Mother Areyasti as an image’ (W-01); cf. further si-bevdos adi[---] kạṿarmọyo imroy edaes etovesniyo (B-01), ạkinanogavaṇ tiyes moḍroṿanak avarạ (M-04). Attributives follow their heads, cf. OPhr. materan areyastin (W-01a ), matar kubeleya ibeya (B-01), but pronouns usually precede them, cf. OPhr. si-keneman ‘this niche’ (M01b), avtay materey ‘to the Mother herself’ (W-01b), σεμουν κνουμανει ‘to this grave’ (incidentally, this consideration may be used as an argument for considering OPhr. -vin in ovevin onoman W-01b as a pronoun, presumably meaning ‘(his) own name’, rather than an adjective; ove- may be a conjunction ‘or’). An exception is κακουν κιν ‘whatever harm’ (No. 100), for which cf. the Gk. postposed enclitic τις ‘someone’. In the NPhr. protases with μανκα, the pronoun and the noun are often separated, probably for metrical reasons, cf. ιος νι σαι κακουν αδδακεμ μανκαι (No. 35; for more examples see 6.2) as opposed to the regular ιος νι σα του μανκα κακουν αδδακετ (No. 82). Clitics (particles, conjunctions, enclitic pronouns) obey Wackernagel’s Law and appear after the first accented word of the sentence, e.g., the particle ni in OPhr. ios ni ạḳenan egeseti (P-04a) or NPhr. ιος νι σεμον κνουμανει κακον αδ̣δακετ (No. 3). An interesting pattern is found with the sentence conjunction κε ‘and’. It normally stands in second position, even if the sentence begins with a preposition, e.g. … τιτετικμενος ας τιαν ε̣ιτου, με κε οι τοτοσσειτι Βας βεκος (No. 99) ‘… let him be condemned by Zeus, and Bat will deprive him of his bread’; … τιττετικμενος ατ Τι αδειτου, ακ κε οι βεκος ακκαλος τιδρε<γ>ρουν ειτου (No. 76) ‘‘… let him be condemned by Zeus, and let his bread be uneatable’. Here, the prepositions με and ακ (= αδ) are followed by the sentence conjunction κε and then by a clitic pronoun of the 3 rd person. If, however, prepositions are construed with a noun rather than with a clitic, the conjunction κε stands after the noun, cf. … Βα[ς] ιοι βεκος μεβερε[τ], α̣τ Τιη κε τιτετικμ[ε]νος ειτου (No. 86) ‘… Bat will deprive him of his bread, and let him be condemned by Zeus’; … γεγαριτμενο<ς> ειτου, πουρ ουανακταν κε ουρανιον ιστεικετ (~ Gk. ἐκδείκνυμι) Διουνσιν (No. 88) ‘… let him be devoted, and he will be exposed to the heavenly king Dionysos’. When used as a word conjunction, κε appears either after each member (X κε Y κε: δεως κε ζεμελως κε), or after the second word only (X Y κε: δεως ζεμελως κε) (cf. Brixhe 1978b: 1 ff.). Incidentally, asyndetic δεως ζεμελως is also attested several times. Preverbs generally stand immediately before the verb, but tmesis is also attested. For instance, in με κε οι τοτοσσειτι Βας βεκος (No. 99) ‘Bat will deprive him of his bread’, με and τοτοσσειτι are separated, in contrast with β̣ε<κ>ος ιοι με-τοτοσσειτι σαρναν (No. 18; to be read thus with A. Hämmig, p.c., rather than as τοτοσσ’ ευγισαρναν with Haas 1966: 100). A slightly different case is the apodosis ‘let him be condemned by Zeus (and by gods)’, e.g., ατ Τιε τιτετικμενος ειτου (No. 94), α̣τ Τιη κε τιτετικμ[ε]νος ειτου (No. 86), ατ Τιη κε δεως κε τιττετικμενος ειτου (No. 62), even ατ Τιε ειτου (No. 56), where αδ was felt by the speakers to belong to the verb, as follows from many occurrences of the formulaic (τιττετικμενος) ατ Τιε αδ-ειτου with preverb repetition. 1830 XVI. Languages of fragmentary attestation 6.2. Agreement In NPhr., we witness progressing case syncretism in ā-stems, probably triggered by the change of final *-āi to -ā and thus by a merger of nom. and dat.sg., cf. the following examples of the protasis ‘whoever will inflict harm upon this stele’: No. 35. ιος νι σαι κακουν αδδακεμ μανκαι with “correct” endings vs. No. 69. ιος σαι κακον αδδακετ μανκαν and No. 60. ιος νι σαν κακουν αδ[δα]κε μανκαι, which show an accusative ending instead of a dative. If the subject of the sentence is “A and B”, the predicate adjective agrees in gender and number with the first member. For instance, in the apodosis of No. 33. αυτος κε ουα κ εροκα γεγαριτμενος ας Βαταν τευτους ‘he himself and his progeny (?) will be condemned by Bat’, γεγαριτμενος agrees with αυτος. Similarly, in No. 12. ζειρα κε οι πειες κε τιττετικμενα ατ Τιε αδειττνου, we see that nom.pl. n. τιττετικμενα agrees in gender with ζειρα. 7. References Brixhe, Claude 1978a Études néo-phrygiennes I. Verbum 1/1: 3−21. Brixhe, Claude 1978b Études néo-phrygiennes II. Verbum 1/2: 1−22. Brixhe, Claude 1990 Comparaison et langues faiblement documentées: l’exemple du phrygien et de ses voyelles longues. In: La reconstruction des laryngales. (Bibliothèque de la Faculté de Philosophie et Lettres de l’Université de Liège, fascicule CCLIII). Liège: Les Belles Lettres, 59−99. Brixhe, Claude 1999 Prolégomènes au corpus néo-phrygien. Bulletin de la Société de Linguistique 94: 285− 315. Brixhe, Claude 2002 Corpus des inscriptions paléo-phrygiennes. Supplément I. Kadmos 41: 1−102. Brixhe, Claude 2004 Corpus des inscriptions paléo-phrygiennes. Supplément II. Kadmos 43: 1−130. Brixhe, Claude and Thomas Drew-Bear 1997 Huit inscriptions néo-phrygiennes. In: Gusmani et al. (eds.), 71−114. Brixhe, Claude and Thomas Drew-Bear 2010 Inscription phrygienne hellénistique de Prymnessos. Kadmos 49: 161−168. Brixhe, Claude and Michel Lejeune 1984 Corpus des inscriptions paléo-phrygiennes. 2 vols. Paris: Recherche sur les civilisations. Brixhe, Claude and Günter Neumann 1985 Découverte du plus long texte néo-phrygien: l’inscription de Gezler Köyü. Kadmos 24: 161−184. Brixhe, Claude and Marc Waelkens 1981 Un nouveau document néo-phrygien au musée d’Afyon. Kadmos 20: 66−75. Drew-Bear, Thomas, Alexander Lubotsky, and Mevlüt Üyümez 2008 Three New Phrygian inscriptions. Kadmos 47: 109−116. Gorbachov, Yaroslav 2005 The origin of the Phrygian aorist of the type edaes. In: Karlene Jones-Bley, Martin E. Huld, Angela Della Volpe, and Miriam Robbins Dexter (eds.), Proceedings of the Six- 101. Phrygian 1831 teenth Annual UCLA Indo-European Conference, Los Angeles, November 5−6, 2004. (Journal of Indo-European Studies Monograph Series No. 50). Washington, DC: Institute for the Study of Man, 191−217. Gusmani, Roberto, Mirjo Salvini, and Pietro Vannicelli (eds.) 1997 Frigi e frigio, Atti del 10 Simposio Internazionale, Roma, 16−17 ottobre 1995. Rome: Consiglio nazionale delle ricerche. Haas, Otto 1966 Die phrygischen Sprachdenkmäler. Sofia: Akadémie bulgare des sciences. Laminger-Pascher, Gertrud 1984 Beiträge zu den griechischen Inschriften Lycaoniens. (Österreichische Akademie der Wissenschaften, Phil.-hist. Klasse, Denkschriften 173). Vienna: Osterreichische Akademie der Wissenschaften. Lejeune, Michel 1969 Discussions sur l’alphabet phrygien. Studi Micenei ed Egeo-Anatolici 10: 19−47. Lubotsky, Alexander 1988 The Old Phrygian Areyastis-inscription. Kadmos 27: 9−26. Lubotsky, Alexander 1993 Word boundaries in the Old Phrygian Germanos inscription. Epigraphica Anatolica 21: 93−98. Lubotsky, Alexander 1997 New Phrygian inscription No. 48: Palaeographic and linguistic comments. In: Gusmani, Salvini, and Vannicelli (eds.), 115−130. Lubotsky, Alexander 1998 New Phrygian metrics and the δεως ζεμελως formula. In: Jay Jasanoff, H. Craig Melchert, and Lisi Oliver (eds.), Mír curad. Studies in honor of Calvert Watkins. Innsbruck: Institut für Sprachwissenschaft der Universität, 413−421. Lubotsky, Alexander 2004 The Phrygian Zeus and the problem of the “Lautverschiebung”. Historische Sprachforschung 117: 229−237. Mitchell, Stephen 1993 Anatolia. Land, Men and Gods in Asia Minor, I. The Celts in Anatolia and the Impact of Roman Rule. Oxford: The Clarendon Press. Rix, Helmut 1992 Historische Grammatik des Griechischen. Darmstadt: Wissenschaftliche Buchgesellschaft. Vine, Brent 2010 Old Phrygian iman. In: Ronald Kim, Norbert Oettinger, Elisabeth Rieken, and Michael Weiss (eds.), Ex Anatolia Lux: Anatolian and Indo-European Studies in Honor of H. Craig Melchert. Ann Arbor: Beech Stave, 343−355. Orsat Ligorio and Alexander Lubotsky, Leiden (The Netherlands)