Old Assyrian Literacy: Formulating a Method for Graphic Analysis and Some Initial Results

Subartu XXXV

KIM 1Proceedings of the 1st Kültepe International Meeting

Kültepe 19-23 September, 2013

© BREPOLS PUBLISHERS

THIS DOCUMENT MAY BE PRINTED FOR PRIVATE USE ONLY. IT MAY NOT BE DISTRIBUTED WITHOUT PERMISSION OF THE PUBLISHER.

Subar tuKültepe International Meetings (KIM)

Editorial Board

Marc Lebeau, M. Conceição Lopes, Lucio Milano,Adelheid Otto, Walther Sallaberger, Véronique Van der Stede

With the support of the following institutions:Università Ca’ Foscari di Venezia, Université Libre de Bruxelles,

Universidade de Coimbra, Ludwig-Maximilians-Universität München

Editorial Board KIMLevent Atici, Gojko Barjamovic, Fikri Kulakoğlu,

Joseph Lehner, Cécile Michel

Subartu — a peer-reviewed series —

is edited by the European Centre for Upper Mesopotamian Studies

Order forms to be mailed to:Brepols Publishers, Begijnhof 67, B-2300 Turnhout, Belgium

FIKRI KULAKOĞLU & CÉCILE MICHEL

(editors)

PROCEEDINGS OF THE 1ST KÜLTEPE INTERNATIONAL MEETING

KÜLTEPE, 19-23 SEPTEMBER, 2013

Studies dedicated to Kutlu Emre

KIM 1

(KULTEPE INTERNATIONAL MEETINGS 1)

FH



Fikri KULAKOGLU & Cécile MICHELProceedings of the 1st Kültepe International MeetingKültepe, 19-23 September, 2013Studies Dedicated To Kutlu Emre(=Subartu XXXV, KIM 1)A4, sewn, xiv+249 pagesContents: This first volume of the Kültepe International Meetings gathers interdisciplinary studies dedicated to Kültepe, ancient Kaneš (central Anatolia), and its environment, mainly during the Bronze Age.All rights reserved. No part of this publication may be reproduced, stored in aretrieval system, or transmitted, in any form or by any means,electronic, mechanical, photocopying, recording, or otherwise,without the prior permission of the publisher.

© 2015, Brepols Publishers n.v., Turnhout, Belgium

ISBN 978-2-503-55545-4D/2015/0095/103

Printed in the EU on acid-free paper

In memory of Kutlu Emre who dedicated 59 years of her life to Kültepe



vii

Table of Contents

Fikri KULAKOğLU Introduction ix Cécile MICHEL

Kutlu EMRE Tahsin Özgüç: The Nestor of Anatolian Archaeology 1

Part 1: Kültepe Bronze Age Archaeology, Chronology and Population

Fikri KULAKOğLU Current Research at Kültepe 9

Mogens Trolle LARSEN The Relative Chronology of the Old Assyrian Period and its Consequences 23

Thomas Klitgaard HERTEL Paternal Estates in Old Assyrian Society 29

Part 2: Kültepe Archives Belonging to Assyrians

Jan Gerrit DERCKSEN The Archive of Ali-ahum (I). The documents Excavated in N-O-P/20 in 1950 47

Hakan EROL The Archives of Šu-Ištar son of Aššur-bāni (Kt 92/k 264-1008) 59

Klaas R. VEENHOF The Archive of Elamma son of Iddin-Suen and his Family 73

Cécile MICHEL Women in the Family of Ali-ahum son of Iddin-Suen (1993 Kültepe archive) 85

Murat ÇAYIR A Letter of Ennam-Aššur son of Šalim-ahum Discovered in 2001 at Kültepe 95

Veysel DONBAZ The Remaining Unpublished Kültepe Tablets from the Hrozný Excavations in 1925 Housed in Istanbul 101

Part 3: Writing

Guido KRYSZAT Old Assyrian Writing and the Secret of the Kültepe Eponym List A 111

Edward STRATFORD Old Assyrian Literacy: Formulating a Method for Graphic Analysis and Some Initial Results 117

Part 4: Anatolian Settlements and Daily Life

Alessio PALMISANO Simulating Past Human Landscapes: Models of Settlement Mark ALTAWEEL Hierarchy in Central Anatolia during the Old Assyrian Colony Period 131

Xiaowen SHI Village Life in Ancient Anatolia: the case of Talwahšušara 147

Güzel ÖZTÜRK Representations of Religious Practice at Kültepe: Alabaster Idols of Early Bronze Age 155



viii

Table of Contents

Part 5: Metals and Minerals

Ergun KAPTAN Ancient Stone Materials Used for Ore Enrichment in Anatolia 173

Evren YAZGAN Cassiterite (Tin) Mineralization Related with Erciyes Volcanic Activities and the Mode of Formation of the Hematite- Cassiterite-Yazganite-Tridymite Paragenesis and its Implication for Bronze Alloys 183

Joseph W. LEHNER, Continuity of Tin Bronze Consumption Evren YAZGAN, during the Late 3rd Millennium BC at Kültepe 195Ernst PERNICKA Fikri KULAKOğLU

Levent ERCANLI The Archaeometallurgical Study on Metal Artifacts of Kültepe in Assyrian Trade Colony Period 219

Part 6: Kültepe after the Bronze Age

Gojko BARJAMOVIC Kültepe after Kaneš 233

Mehmet UğURYOL Recent Practices for the Conservation of Adobe Ruins of Kültepe 243Drahşan UğURYOL

ix

IntroductionFikri Kulakoğlu & Cécile Michel

The first “Cappadocian tablets” were discovered in the second half of 19th century, and were sold on the Kayseri and Istanbul bazars. In 1881, T. G. Pinches published two of these tablets preserved respectively in the British Museum in London and in the Bibliothèque Nationale in Paris.1 The texts were dated to the begin-ning of the 2nd millennium; they were written in Akkadian, but originated from central Anatolia. Scholars were thus a little skeptical about their authenticity. On January 8, 1888, J. Oppert wrote a letter to G. Mas-péro concerning these texts.2 G. Maspéro had asked for his expertise concerning a cuneiform text found in Egypt; J. Oppert explained that he had first thought that it was a fake too well preserved, and compared this tablet with the so-called Cappadocian tablets, for which he also first doubted about their authenticity:

Mes soupçons s’étaient surtout éveillés par la ressemblance de la tablette envoyée par vous avec les fameuses tablettes dites cappadociennes qui proviennent toutes de Césarée et que je crois toutes fabriquées ; du moins jusqu’à présent. Il existe dans tout l’Orient des ateliers de fausses antiquités qu’il faut payer plus cher, parce que, ils en font payer la main-d’œuvre. Les caractères sont un mélange de styles assyrien et babylonien (…) J’incline (à présent) vers l’authenticité.

The “Cappadocian” tablets, written in an Assyrian dialect, were found in a place called Kültepe, 21 km North-East of modern Kayseri, at the foothill of the Erciyes Mountain. In 1894, P. Jensen suggested that these tablets would come from an ancient town called Kaneš, a name repeatedly cited.3 In 1924, B. Landsberger confirmed the hypothesis of P. Jensen, and identified Kültepe as the ancient city of Kaneš.4 Assyrian merchants settled there during the 19th century BC and left many archives. In 1893 and 1894, E. Chantre began excavating on the top of Kültepe’s höyük. He did not find tablets but bought sixteen samples to the neighboring peasants;5 these were edited by V. Scheil who had previously published such a tablet.6 German archaeologists also explored the mound at the beginning of the 20th century, but could not find tablets.7

In 1925, B. Hrozný, well-known for deciphering the Hittite language, directed a Czechoslovakian expe-dition at Kültepe. Working again on the mound with a lot of workers, he dug a large area in the middle of Waršama’s palace, but did not find any tablet there. He bought some six hundred tablets to the local farmers before learning that the tablets were found in the surroundings fields, an area corresponding to the merchant district in the lower city, later called “kārum”. He excavated in the area (later referred to as F-J/7-12) and found four hundred more tablets.8 The site was then abandoned during more than twenty years, a period during which the surrounding villagers took out some stones of Waršama’s palace to build their houses and the soil of the mound to fertilize their fields. The site was plundered by illicit diggers who sold hundreds of cuneiform tablets to western museums and private collections.

In 1948, official exploitation of the site started, with the support of the Turkish Historical Society and under the scientific direction of Tahsin Özgüç from Ankara University. Since this date, archaeological missions have taken place every year, except 1952.9 Kutlu Emre joined the excavations in 1955 as a student of Professor Özgüç. In 1998, a small symposium gathering archaeologists and philologists working on Kültepe material was organized by Tahsin Özgüç to celebrate the fiftieth anniversary of the beginning of Turkish excavations. After the death of Tahsin Özgüç in October 2005, the excavations went on under the direction of Fikri Kulakoğlu, Professor at Ankara University.10 A day meeting in memory of Tahsin Bey was organized in May 2006 to which participated several archaeologists and philologists.

1 Pinches 1881a; 1881b.2 This unpublished letter is preserved at the Bibliothèque de l’Institut (Paris, France). We address our warmest thanks to Mireille Pastoureau, the former director of the library, who allowed Cécile Michel and Brigitte Lion to work on Jules Oppert’s correspondence.3 Jensen 1894.4 Landsberger 1924.5 Chantre 1898, 92-115.6 Scheil 1896; 1898.7 Grothe 1912; Winckler 1906.8 Hrozný 1925.9 Özgüç 1950; 1953; 1959; 1986; 1999; 2003.10 Kulakoğlu – Kangal 2010; Kulakoğlu 2011.



x

Fikri Kulakoğlu & Cécile Michel

Kutlu Emre participated to the excavations until 2013 and she kindly accepted to open the 1st Kül-tepe International Meeting with a lecture dedicated to the scientific carrier of her former Professor, T. Özgüç.11 Beside the many years she spent at Kültepe, Kutlu Hanım joined to the Acemhöyük, Altıntepe, Maşathöyük, Kululu and Kazankaya excavations under the directorship of Nimet Özgüç and Tahsin Özgüç. She also contributed to the Kamid el-Loz excavations in Lebanon while she was studying in Germany. Kutlu Emre conducted research and excavations throughout all Central Anatolia. Hittite sites of Yanar-lar12 in Afyonkarahisar, Hanözü13 in Tokat, Sultanhanı14 and Yassıdağ15 mounds and Karakuyu Dam16 in Kayseri were excavated by her. She also directed the archaeological survey named “Eastern Cappadocia Survey Project” (Kayseri Province). One of her major contributions to the Near Eastern archaeology is her monograph – and several articles17 – on lead figurines and their stone moulds; it is the reference book for the art and archaeology of Anatolia and beyond.18 The experience she has gained at Kültepe produced key volumes which enlighten the art and history of the Hittites and their predecessors.19 She also published on Anatolian pottery20 and imported ware21 from Mesopotamia and Syria. This volume, since the beginning, was intended to be offered to Kutlu Hanım who had dedicated almost sixty years of her life to Kültepe, and who was the only one who knew every single parcel of the mound and of the lower city by heart. Unfortu-nately, as the book was already submitted, Kutlu Emre passed away on Christmas day (December 25, 2014) in Ankara; this first volume of the Kültepe International Meetings is dedicated to her memory.

Fig. 1: Kutlu Emre at Kültepe.

Kültepe has always been a very prolific site, providing many artefacts and data for a wide range of dis-ciplines. Moreover, almost every year the archaeologists find new Old Assyrian tablets. The total number of

11 We have taken upon ourselves the responsibility of publishing Kutlu Hanım’s lecture in this volume without asking for her permission.12 Emre 1978.13 Emre 1992a.14 Emre 1973.15 Emre 1971a.16 Emre 1993b.17 Emre 1969; 1994b; 2005.18 Emre 1971b.19 Emre 1993a; Emre – Çınaroğlu 1993; Emre 1992b.20 Emre 1963; 1968; 1989; 1992c; 1996.21 Emre 1994a; 1995; 1999.

Introduction

xi

texts discovered in Kültepe, about 22,500, for the great majority in the lower town, places this site among the richest supplier of cuneiform texts from the whole Near East.22

A third workshop gathering Kültepe specialists took place during the Chicago ASOR annual meeting in 2012; it allowed discussions between colleagues working on the Kültepe material from various points of views.23 Fascinating discussions went on during the evenings, showing a strong desire to create the best conditions for further regular scientific exchanges. This motivated the organization of a first “Kültepe International Meeting” (KIM). We wished to create a synergy, to build up collective projects, and thus to have regular meetings taking place at Kültepe. The aim of these conferences is to bring together all the researchers working on Kültepe and its surrounding area, from a wide range of disciplines, in order to set up interdisciplinary collaborations on any question related to Kültepe-Kaneš researches. To make this dialogue possible, a general theme on material culture was addressed, and participants were also invited to present reports on their work in progress.

Every evening, general discussions took place involving all participants. On the first day, the theme addressed was linked to the archaeological aspects. The Early Bronze Age building currently excavated on the mound was discovered in 1986; systematic excavations of this building started in 2009. It is, up to now, the largest official building known in Anatolia for the 3rd millennium BC.24 Excavations have also been going on, since 2006, on the Eastern part of the lower town;25 a survey of the lower town is planned to try to find its limits. Numerous graves from levels II and Ib have been unearthed since 2006 and are systemat-ically studied by an anthropologist (H. Üstündağ). The fire that destroyed the citadel might not have taken place at the same period as in the lower town. Only one house of the lower town had a skeleton of someone trapped by the fire, killed by the fall of a wall, all the other inhabitants had time to leave their homes. The end of the lower town can be dated to 1680, then we observe a microclimate change, the environment was under water and the traffic stopped.26

The second general discussion was linked to the publication of Kültepe tablets. Complete archives are now being deciphered and studied by less than fifteen Turkish and foreign philologists. Six volumes have been published at the Türk Tarih Kurumu since 2010 (archives discovered in 1988, 1992, 1994 and 2001), three more are in press (archives excavated in 1950, 1991, 1994) and many others in preparation. In the volumes’ introduction, the archaeological context is described, and some plans of the houses are provided. K. Emre and F. Kulakoğlu wrote a detailed analysis of the archaeological context of the 2001 archive dated to level Ib (Emre – Kulakoğlu in press). There has been a strong wish to publish sealings on the envelopes together with the tablets, something which will be done for the first time for the 1991 archive. A volume on 1994 sealings is also on preparation. Such a work on archaeological archives allow addressing new topics, as the structure of families over several generations, and to link prosopographical studies together with the relative chronology now provided with complete lists of eponyms.

A final discussion was dedicated to plans for follow-up research and meeting activities. It has been collectively decided that a Kültepe International Meeting will take place at Kültepe every two years at the end of July, centered on Kültepe and its surrounding area. The international organizing team will change every time and will be constituted by two or three archaeologists and philologists, including the director of the excavations. Every meeting should be published at the time of the next one.

The first Kültepe International Meeting gathered on the site about sixty participants from three conti-nents and a dozen of countries. Thirty papers dealing with a wide range of topics were presented during the three days of the meeting. The fourth day, participants enjoyed a tour to Cappadocia offered by Kayseri Metropolitan Municipality. The twenty-one articles published in this volume are arranged in six different sections concerning the archaeology, chronology and population of Kültepe, the study of archives belon-ging to Assyrian merchants, the question of literacy, the Anatolian settlements and daily life, metals and minerals, and Kültepe after Kaneš. We are very grateful to all the authors for their contributions to this volume.

22 Michel 2003; 2006; 2011.23 Atıcı et al. 2014.24 Kulakoğlu – Öztürk 2015.25 We now avoid to identify the kārum with the excavated part of Kültepe lower town, Michel 2014.26 Kulakoğlu 2014.



xii


Fig. 2: Participants to the 1st Kültepe International Meeting, Kültepe, September 2013.

AcknowledgementsThis first Kültepe International Meeting (KIM) would not have been possible without the initiative or

the help of several persons and institutions to whom we would like to address our warmest thanks:

Tahsin Özgüç and Kutlu Emre for their tremendous work revealing Kültepe since the mid-twentieth century, the Turkish Ministery of Culture and Tourism which supports Kültepe excavations since the begin-ning, the General Director of Cultural Assets and Museums Abdullah Kocapınar, the Governor of Kayseri Orhan Düzgün, the Vice Governor Haluk Tunçsu, the President of Kayseri Metropolitan Municipality Mehmet Özhaseki for their important moral, material and financial assistance, the Kültepe team, all the participants of the first Kültepe International Meeting (KIM), the Turkish and French Institutions who sponsored the conference (Ankara University, Faculty of Letters, Turkish Historical Society and CNRS research unit “Archéologies et Sciences de l’Antiquité”), the members of the KIM series editorial board, as well as J. G. Dercksen and K. R. Veenhof, for their help in proofreading some manuscripts, and last but not least, Brepols publisher and Marc Lebeau, editor of SUBARTU series, who encouraged and accepted the creation of the new KIM sub-series.

BibliographyAtici, L. – Kulakoğlu, F. – Barjamovic, G. – Fairbairn, A.

2014 Current Research at Kültepe-Kanesh. An Interdisciplinary and Integrative approach to Trade Networks, Internationalism and Identity, Journal of Cuneiform Studies Suppl. 4, Atlanta.

Chantre, E.

1898 Mission en Cappadoce, Paris.

Emre, K.

1963 “The Pottery of the Assyrian Colony Preiod According to the Building Levels of the Kaniş Karumu”, Anadolu/Anatolia 7, 87-99.

1968 “Acemhöyük Seramiği”, Anadolu/Anatolia 10, 99–153.

1969 “Eine neue Gussforms aus Kültepe”, Zeitschrift für Assyriologie 60, 134–142.

1971a “The Excavations 1971 and 1972 at Sultanhan Höyük”, Anadolu 1971, 119–138.

1971b Anadolu Kurşun Figürinleri ve Taş Kalıpları/Anatolian Lead Figurines and Their Stone Moulds. Ankara.

Introduction

xiii

1973 “Sultanhanı Höyüğünde 1971-1972 yıllarında yapılan Kazılar/The Excavations 1971-1972 at Sultanhanı Höyük”, Anadolu/Anatolia 15, 87–138; pl. I-XII.

1978 Yanarlar. Afyon Yöresinde Bir Hitit Mezarlığı/A Cemetery Near Afyon, Ankara, TTKY. V/22.

1989 “Pottery of Levels III and IV at the Karum of Kanesh”, in: K. Emre – B. Hrouda – M. Mellink – N. Özgüç (eds), Anatolia and the Ancient Near East. Studies in Honor of Tahsin Özgüç, Ankara, 111–128.

1992a “Hanözü: Masat Çervesinde bir Hitit Yerlesmesi”, in: H. Otten – H. Ertem – E. Akurgal – A. Süel (eds), Hittite and Other Anatolian and Near Eastern Studies in Honour of Sedat Alp, Ana-dolu Medeniyetlerini Araştırma ve Tanıtma Vakfı Yayınları, Sayi 1, Ankara, 137–158.

1992b “Two Imported Bottle Shaped Jars from Karum Kanish”, in: B. Hrouda – S. Kroll – P. Z. Spanos (eds), Von Uruk nach Tuttul. Eine Festschrift für Eva Strommenger: Studien und Aufsaetze von Kollegen und Freunden, Münchener Vorderasiatische Studien 12, München, 51–56.

1992c “Çorum Müzesinden Bir grup Hitit Seramiği”, Anadolu Medeniyetleri Müzesi 1991 Yılı Müze Konferansları, Ankara, 103–114.

1993a “A Group of Hittite Statuettes from Alaca Höyük”, Istanbuler Mitteilungen 43, 235–244.

1993b “The Hittite Dam of Karakuyu”, in: H. I. H. Prince Takahito Mikasa (ed.), Essays on Anatolian Archaeology, Bulletin of the Middle Eastern Culture Center in Japan 7, Wiesbaden, 1–42.

1993c “New Lead Figurines and Moulds from Kültepe and Kızılhamza”, in: M. Mellink – E. Porada – T. Özgüç (eds), Aspects of Art and Iconography: Anatolia and its Neighbors. Studies in Honor of Nimet Özgüç, Ankara, 169–177.

1994a “A Type of Syrian Pottery from Kültepe/Kaniş”, in: P. Calmeyer – K. Hecker – L. Jakob-Rost – C. B. F. Walker (eds), Beitraege zur Altorientalischen Archaeologie und Altertumskunde. Festschrift für Barthel Hrouda zum 65. Geburtstag, Wiesbaden, 91–96.

1994b “A New Mould from Kültepe”, in: Festschrift für R. Mayer-Opificius, Altertumskunde des Alten Orients 4, Münster, 71–77.

1995 “Pilgrim-Flasks from Level I of the Karum Kanis”, in: Essays on Ancient Anatolia and its Surrounding Civizilations, Bulletin of the Middle Eastern Culture Center in Japan 8, 173–200.

1996 “Kantharoi from Kültepe/Kanish”, Bulletin of the Middle Eastern Culture Center in Japan 10, 1–11.

1999 “Syrian Bottles from the Karum of Kanish”, Bulletin of the Middle Eastern Culture Center in Japan 11, 39–50.

2005 “Kültepe/Kaniş Karum’unda 1993-2001 Yılları Arasında Keşfedilen Yeni Kurşun Figürinler ve Kalıplar”, in: Refik Duru’ya Armağan, İstanbul.

Emre, K. – Çınaroğlu, A.

1993 “A Group of Metal Hittite Vessels from Kınık-Kastamonu”, in: M. Mellink – E. Porada – T. Özgüç (eds), Aspects of Art and Iconography: Anatolia and its Neighbors. Studies in Honor of Nimet Özgüç, Ankara, 675–713.

Hrozný, B.

1927 “Rapport préliminaire sur les fouilles tchécoslovaques de Kültepe”, Syria 8, 1–12.

Grothe, H.

1912 Meine Vorderasienexpedition, 1906-1907, vols I-II, Leipzig.

Jensen, P.

1894 “Die kappadocischen Keilschrifttafelchen”, Zeitschrift für Assyriologie 9, 62–81.

Kulakoğlu, F.

2011 “Kültepe-Kaneš: a second-Millennium BC Trading Center on the Central Plateau”, in: S. Steadman – G. McMahon (eds), The Oxford Handbook of Ancient Anatolia, Oxford, 1012–1030.

2014 “Kanesh after the Assyrian Colony Period: Current Research at Kültepe and the Question of the End of the Bronze Age Settlement”, in: Atici et al. 2014, 85–94.



xiv


Kulakoğlu, F. – Kangal, S.

2010 Anadolu’nun Önsözü Kültepe Kaniş-Karumu. Asurlular İstanbul’da/Anatolia’s Prologue Kültepe Kanesh Karum. Assyrians in Istanbul, Kayseri.

Lansberger, B.

1924 “Über die Völker Vorderasiens im dritten Jahrtausend”, Zeitschrift für Assyriologie 35, 213–244.

Michel, C.

2003 Old Assyrian Bibliography of Cuneiform Texts, Bullae, Seals and the Results of the Excavations at Assur, Kültepe/Kanis, Acemhöyük, Alishar and Boğazköy, Old Assyrian Archives Studies 1, Publications de l’Institut historique-archéologique néerlandais de Stamboul 97, Leiden.

2006 “Old Assyrian Bibliography 1 (February 2003 – July 2006)”, Archiv für Orientforschung 51, 436–449.

2011 “Old Assyrian Bibliography 2 (August 2006 – April 2009)”, Archiv für Orientforschung 52, 416–437.

2014 “Considerations on the Assyrian settlement at Kaneš”, in: Atici et al. 2014, 69–84.

Özgüç, T.

1950 Türk Tarih Kurumu Tarafından Yapılan Kültepe Kazısı Raporu 1948 (Ausgrabungen im Kültepe), Türk Tarih Kurumu Yayınlarından V/10, Ankara.

1953 Kültepe Kazısı Raporu 1949 (Ausgrabungen im Kültepe), Türk Tarih Kurumu Yayınlarından V/12, Ankara.

1959 Kültepe-Kaniş, New Researches at the Center of the Assyrian Trade Colonies, Türk Tarih Kurumu Yayınlarından V/19, Ankara.

1986 Kültepe-Kaniş II, Türk Tarih Kurumu Yayınlarından V/41, Ankara.

1999 Kültepe-Kaniş/Neša Sarayları ve Mabetleri. The Palaces and Temples of Kültepe-Kaniš/Neša, Türk Tarih Kurumu Yayınlarından V/46, Ankara.

Pinches, T. G.

1881a “Tablet from Cappadocia, now in the British Museum”, Proceedings of the Society of Biblical Archaeology November, 11–18.

1881b “Tablet from Cappadocia in the Bibliothèque Nationale”, Proceedings of the Society of Biblical Archaeology December, 28–33.

Scheil, V.

1896 Tablette cappadocienne, Recueil de travaux relatifs à la philologie et à l’archéologie égyptiennes et assyriennes 18, 74–75.

1898 “Textes cuneiforms”, in: Chantre 1898, 92–109.

Winckler, H.

1906 “Die im Sommer 1906 in Kleinasien ausgeführten Ausgrabungen”, Orientalistische Literaturzeitung 9, 621–634.

117

Old Assyrian Literacy: Formulating a Method for Graphic Analysis and Some Initial Results

Edward Stratford*

The British Museum holds eight tablets authored by Šalim-ah˘um. However, it is intuitively clear when

looking at this group of tablets that not all eight were written by the same hand. Did Šalim-ah˘um write his

own letters? If so, which letters did he write? Perhaps more importantly, what does whether or not Šalim-ah

˘um wrote his own letters imply for Old Assyrian literacy?

For some time there has been a sound consensus among Old Assyrian specialists that the average mer-chant in the Old Assyrian trade was literate at a level that allowed them to read and produce the documenta-tion involved in their trade, and that some women were literate as well. There are a number of reasons why this seems likely. The set of signs used in Old Assyrian is smaller than the full range of signs employed in other corpora, reducing the burden involved in learnng the script. The regular travel of many merchants made it impractical to depend on scribes, particularly their own scribes in the course of trade. Moreover, it can intuitively be observed that documents by some authors exhibit clear differences in skill in the ductus. The apparent capacity for some Anatolians to write, betrayed by subjectively poorer hands, also seems to support the broad base for literacy among the Assyrian population in Anatolia.1 C. Michel has argued that the need for families to communicate and the presence of women’s letters evidence literacy training even within homes.2 Despite these and other important studies, what Powell noted over three decades ago is still true of most cuneiform corpora: “Really sound statistics for literacy prior to the nineteenth century do not exist. Every estimate for literacy […] must be partially a kind of intuitive process, i.e., putting together eve-rything one has learned and experienced and, after weighing it all in the balance, reaching a conclusion”.3

The present brief study is by no means the first time that analysis of ductus has been performed on cuneiform. However, the principal focus here is on the Old Assyrian corpus and its relation to literacy in a somewhat anomolous population. This study is meant to set up a framework for answering questions through analysis of hands such as: How many merchants can be verified to have had the ability to write? To what extent was an individual’s literacy likely to be affected by class, occupation, gender, etc.? Which merchants could write and which could not? Can it be proven that merchants tended to write their own letters? In several of the cases reviewed below, the most logical way to explain the distribution of hands is to assert that the merchants in question were writing their own letters. The present study appears to support the idea that travelling merchants needed to have some level of craft literacy. That is, they were able to read and write their own letters when necessary. However, not everyone in the trade travelled often. Letters from residents of Aššur in several cases reveal more than one hand. While the ability to distinguish two different hands in association with a single author may be increased simply when more tablets survive for that author, the presence of two different hands is limited in this study to persons in Aššur.

There are two primary operations in chirographic analysis (distinguishing hands), that are presently most useful for the Old Assyrian corpus. The first operation is analyzing documents authored by the same person to determine if more than one hand is present among the documents. The second operation is analy-zing documents with multiple authors to see if a hand associated with one of the authors can be identified. To distinguish one hand among hundreds is extremely difficult if not impossible. But executing these two operations on a series of cases corroborates that some authors may have dictated their letters, that some authors may have been unable to inscribe their own documents, but that many authors appear to be their own scribes, and that literacy was not so prestigious that higher status individuals reserved the inscription of documents for themselves. Specifically, one case reveals that in two witnessed documents, the lowest ranked witness is the most likely scribe. Chirographic analysis also has the potential to distinguish homo-nymous individuals, and can corroborate the authorship, through associated hand, of acephalous letters or anonymous notes. In this way, chirographic analysis can help sort out archival distinctions in the “old texts” (those tablets excavated from Kültepe before the commencement of the official Turkish excavations in 1948), and also support studies related to social networking and microhistorical reconstruction. This

* Brigham Young University; [email protected] Kryszat 2008, 231–238.2 Most recently, Michel 2010, 349; Michel 2011, 105–106. Kryszat 2008, 231–38.3 Powell 1981, 436.



118

Edward Stratford

approach promised to be profitable when combined with a number of other approaches, such as tendencies of orthography, material analysis, etc., but the focus in this contribution is on chirography.

Initial inquiry into this topic was performed in the summer of 2013 during a short visit to the British Museum. Subjects were chosen from the correspondents of Pūšu-kēn. Further work was done on the Nies Collection at the Yale Babylonian Collection in the summer of 2014. I would like to thank the staff at each museum for their helpfulness and generosity. I also made use of images gathered by myself at these two collections. Images of all other tablets included in this study were accessed online at the Cuneiform Digital Library Initiative. Insufficient time for autopsy on all tablets of interest at the British Museum and a few at the Yale Babylonian Collection meant that some tablets do not appear in this study.

After discussing the aims and some basic suppositions and suggested practices, I will review several cases to demonstrate the utility of chirographic analysis in the Old Assyrian trade, and discuss some of the initial findings.

1. Aims, Methods, Terminology, NotationDistinguishing individual hands in cuneiform, what I will call cuneiform chirography, draws from the basic principles of modern forensic document identification. But it is also useful to recognize chirography as related but distinct from paleography in aims and thus methods and assumptions. The Old Assyrian corpus offers good opportunities for chirographic analysis, though clear findings are not always guaranteed. After discussing methods of chirographic analysis attempted here, I will discuss terminology and notation, giving brief explanations for terms used in this piece.

1.1 Aims

The intent of chirographic analysis is to distinguish hands of individuals, or the hands involved with documents associated with individuals. Chirography (from χειρο “hand”, γραφ “writing”) is a rather outda-ted term, but was up through the 19th century used to refer to the study of handwriting and its analysis, and also to refer to the handwriting of a particular individual. In medieval Europe, a chirograph was a document written in the same hand several times on the same sheet. After each copy was cut away, sometimes in irregular pattern, the validity of the copy could be established by reuniting it with its companion copies for inspection. The emphasis in chirography was to establish the legitimacy of the document, with emphasis on the handwriting. It appropriately corresponds to a “hand” as used here without the confusion between actual hands and the product of those hands. Because chirography as a study implies the “hand” as an object of investigation, the term seems appropriate for distinguishing this endeavor from paleography, where broader trends are normally the focus of analysis.

While seemingly counterintuitive, cuneiform may actually an easier medium to distinguish hands than the medium of ink and paper handwriting. This is primarily because the best analysis of individual hands must focus on the construction of individual letterforms as opposed to overall similarities in the appearance of one hand to another on different tablets.4 Though the cuneiform script has a larger set of characters than an alphabetic script, it is not necessary to survey all characters on a tablet to compare handwriting. Moreo-ver, cuneiform characters are made up of individual wedges, and thus differ amongst themselves in one very basic way – the positions of the wedges relative to one another. Though often subtle, this variation is still sometimes more easily observed with the naked eye than the difference between the ink script hand-writing of two persons.

1.2 Methods

Chirographic analysis and the methods I advocate for the Old Assyrian corpus are an adaptation from foren-sic document identification, and are related to cuneiform paleography. However, the focus in each of these related endeavors is slightly different. In modern forensic document identification, the focus is usually on determining whether or not a document is an intentional forgery or legitimate. In cases where handwriting is involved, this includes the analysis of handwriting for the purpose of verification. But forensic analysis also involves a range of other approaches, including material analysis of ink and paper. The methods of such practices are often kept confidential in order to deter the development of techniques to overcome them. However, the basic principle of forensic document analysis is to establish whether or not there are suffici-ent characteristics in different documents that match “or correspond in their relationships to one another,

4 This focus on individual elements as opposed to overall appearance is considered a basic principle of handwriting analysis, as expressed in works cited in following notes and in Ansell 1979, 239–251. The principle of variation with a single writer, though more circumscribed in relation to other writers has been demonstrated for handwriting with analysis on a single letter: Marquis et al. 2005, 23–32.

119

Old Assyrian Literacy: Formulating a Method for Graphic Analysis

of such number and significance as to preclude the possibility of their occurrence by pure coincidence”.5 Handwriting is considered a learned pattern of behavior, and thus the aim is to detect idiosyncrasies in writing that betray artifice.6 Because forensic document analysis is either actively or provisionally for use in a court of law, presentation of the results is sometimes important. When presenting results, a frameset approach is often used. This approach will be discussed further below.

While cuneiform chirography is similar to handwriting analysis in modern forensics, its scope is distinct from the broader focus of cuneiform paleography.7 Paleography normally refers to a project that, at least in intent, comprises the complete cataloguing of all graphs in a corpus. Because the cuneiform record spans thousands of years, and many corpora are not intensively author-centered, one use of paleography is to distinguish the development of sign forms over time or among schools or scribal circles, sometimes simply to improve interpretation. The characteristics embedded in the graphs produced by members of a school, scribal circle, social group defined by a time period, etc., could be said to converge in form. Such a form has sometimes been referred to as an allograph, though others refer to an allograph as “any graph of a particular grapheme”.8 If for just a moment, the “group” sense is allowed for allograph, then the term idiograph might defined in contrast to that idea, that is, the form of graphs of the same grapheme that are characteristic of an individual hand. I would argue that in chirography, cataloguing all idiographs for a person is less neces-sary than in projects of paleography because the operation is one of direct comparison in a smaller scale. Chirography in the Old Assyrian corpus is useful for further investigating literacy and its prestige. Many characteristics that are important for paleography are already noted in distinguishing between corpora, such as the lateral layout of documents from Emar. When focus is on the shared practices of document creation, document formatting, allographs, and general aspects of ductus may be applicable. On the other hand, at least within the Old Assyrian corpus, chirographic analysis must be instead focused on identifying idiogra-phic characteristics of particular graphs.9 This analysis can reasonably be used to corroborate documents already suspected to be written by one and the same scribe, or to investigate whether or not all documents authored by the same person were written by the same scribe. It will probably always be too difficult to distinguish one hand from dozens of others. And so it also seems unnecessarily onerous to demand a full catalogue of all graphs for two different hand in order to demonstrate a distinction between them.

Because chirography depends on small but consistent differences, autopsy is always preferable, espe-cially for the initial survey of a hand or hands in a single individual’s group of documents. Nonetheless, it will always be necessary to use photographs to compare texts that cannot be physically placed next to each other, which is the case particularly for the “old texts” in the Old Assyrian corpus scattered in dozens of museums and private collections. Moreover the most efficient and reliable way to publish comparisons is through a frameset approach. While hand copies are useful for editions, because there is no published standard for producing hand copies, and because the effort required to reflect the details necessary for chirographic analysis is generally impractical, hand copies cannot be used for chirographic analysis. Some idiographs are sufficiently distinct that they are represented in some handcopies, for example the unusual UD idiograph found in one of the hands in Šalim-ah

˘um’s documents, but more often the differences are not

represented.10

Not all apparent phenomena on tablets are of equal utility in chirographic analysis. The intent is to discern an individual scribe, therefore separating the factors of producing the clay document and their effect on the appearance of the document is necessary. Broadly speaking, there are three factors that affect the appearance of the ductus on a cuneiform tablet. The materials employed affect certain aspects of the appearance. As is evident in many Hittite documents found at Hattusha, a sharp stylus and washed clay are more amenable to a fine ductus. Similarly, poor clay and a dull or low quality stylus, written hurriedly at the

5 “When any two items possess a combination of independent discriminating elements (characteristics) that are simi-lar and/or, and there are no inexplicable disparities, it may be concluded that they are the same in nature or are related to a common source (the principle of identification)”. Huber – Headrick 1999, 84.6 Davis 2007, 265. The article is an excellent discussion of the difference between handwriting identification and forensic document analysis form someone associated with the challenge of the cuneiform script, although treatment of these practices for cuneiform is not directly treated.7 One common point of departure for discussion of paleography is the statement made by Edzard (1978, 555).8 This way of defining allograph and idiograph is similar to but importantly different from the way they were descri-bed for the Cuneiform Digital Paleography Project. There, “Any graph is an allograph of a grapheme”. (www.cdp.bham.ac.uk). While this usage of the word is acceptable, there is no semantic space left for the shared characteristics of graphs of the same grapheme. Graphs of the same grapheme can pragmatically be referred by their grapheme name.9 Focus specifically on idiographic characteristics as distinct from allographic characteristics may suggest we call this activity idiography, but we will agree that it does sound too, well, stupid.10 The hand copies of J. Lewy and several others do distinguish between the different forms of the UD sign in Šalim-ah

˘um.



120

Edward Stratford

end of the day could lead the same person to produce much less elegant writing. Likewise, the training of the scribe will manifest itself in the layout of the tablet, and perhaps also at an allographic level. In addition, the scribe’s conscious management of space on the particular tablet will have an effect on the appearance of the ductus as well. However, the habitual but idiosyncratic manner of producing individual graphs will be the most consistent characteristic to differentiate a scribe’s chirography, as these idiographic characteristics will manifest fairly regularly despite differing circumstances of clay or stylus quality, or available space.

There are several reasons why chirography, which proceeds on comparing small sets of documents at a time, is a more appropriate approach at this point for the Old Assyrian corpus than paleography. First, in the Old Assyrian trade, where peregrinating merchants could be acquiring the materials for their letters, notes, and other documents from one of several dozen large cities, or from more than a hundred intermediate locations, the quality and type of clay and perhaps also stylus varied widely, and the relation between con-tent and space might have presented challenges on occasion. The scribe’s negotiation with all these factors could directly affect the relative size of signs and wedges, the layout, even the angle of the wedges. The regular appearance of slanted rulings on Old Assyrian tablets also may be diagnostic of a particular hand, but not sufficiently so. More often, the slanted rulings found in Old Assyrian letters arose from scribes’ efforts to economize space. As a scribe wrote an individual line, he or she sometimes tried to fit an extra word at the end of the line by writing smaller characters. This had an effect on vertical as well as horizontal dimensions, leaving extra room below the characters on the right side of the tablet. When the new ruling was drawn, the scribe followed the available space, rather than cutting a straight ruling. In the aggregate, this could have provided space for an extra line or two at the bottom right of that face of the tablet. Some scribes would have employed this consciously when worried about surface space, while some scribes did so habitually. Moreover, it does not seem at this point a useful exercise to distinguish different scribal schools or other such groups in the Old Assyrian corpus. Even when material circumstances do not substantially affect the general appearance of the tablet, scribal training was likely either shorter in Old Assyrian mer-chant society, or even perhaps taught in the home, making it more unlikely to discover allographs marking training from any particular school. It is even difficult to say yet whether an order of impressions of wedges is indicative of a particular school, or whether the gestalt of the sign was taught, and an individual produced the shape as they found most natural. Nonetheless these questions should be pursued in the future.

With these factors in mind for the Old Assyrian corpus, the frameset approach is well suited for chi-rographic analysis. In the frameset approach, the emphasis is on comparing individual graphs of the same grapheme or sequences of the same graphs on particular tablets, focusing on the arrangement of wedges. This is best accomplished when the authors of the documents are known, or there is some existing reason to suspect that the hands of two or more tablets are comparable. While the ductus in other periods, such as the Neo-Assyrian period, was more regularized, the gestalt of a sign in the Old Assyrian period was the most important graphemic quality. In Old Assyrian, some graphemes were sufficiently similar in form (for example, MA, KU TÚG), that the idiographs of one grapheme in one hand may match more closely the idiographs of a different grapheme in another hand. This means that the context is often more important than the form. But in addition to variations, strictly speaking, there are tendencies of rendering that can also be diagnostic for separating two hands, for example, the regular relation of signs to the ruling, consistent lean of graphs, or the general tendency of spacing the graphs. The best candidates for idiographs are simpler graphemes or elements of graphemes for which a large number of exemplars will occur on most tablets. Consistent variation in simpler graphs or elements, composed of fewer wedges, is more significant than variation is more complicated signs. For example, it is more likely that the scribe of a tablet intuitively for-med the MA grapheme in a particular way than the ILLAT grapheme, which is more complicated in form. Still, no set of graphemes is appropriate to designate ahead of time, as each act of comparison should be done with flexibility toward identifying useful idiographs.

To display a frameset table of one idiograph from one hand, I suggest showing a photo of a representa-tive graph, surrounded by smaller photos of the other graphs of that idiograph. To compare two opposing idiographs from different hands, two representative graphs can be displayed side by side, with related idio-graph graphs surrounding the open sides of each representative graphs (Fig. 1). Because producing such tables can become expensive in print, such tables are best distributed in electronic format, such as through the internet, and only a few examples are included here. On a larger scale, the Old Assyrian Research Environment (OARE, accessed through http://oare.byu.edu or http://ochre.lib.uchicago.edu) will be able to facilitate both the comparison and distribution of chirographic analysis. The data structure of OARE, which can separate texts into individual graphs on the epigraphic level, allows sections of photos encompassing a graph to be linked to the unit representing that graph. Developments in the near future will allow texts to be sorted by author or authors, and graphs sorted by grapheme and visually arranged for comparison. This will provide for the possibility to more quickly arrange frameset tables for different hand, making opportunities for autopsy more efficient and facilitating distribution of results.

121


The observations of forensic handwriting experts are that handwriting on ink and paper changes very little over time. But it is possible that the hand of a scribe could change over time, perhaps even suddenly or gradually (revolutionary or evolutionary).11 An answer to this question is beyond the scope of the current contribution, but it may be possible to establish in cuneiform, and seems likely. At present, it appears that such a consistency holds for Old Assyrian merchants, but allowance must be given for change over time, especially when an individual’s hand is preserved from youth and through several decades. Many tablets from the case of Šalim-ah

˘um included in this initial inquiry arise from within a short period of time, demon-

strating that in the distinction between two different hands there, development over time was not the factor. However, the possibility of changes in hands over time will need to be considered further in future studies.12

1.3 Terminology

Most terms used here are familiar to cuneiformists, but within chirographic analysis some terms must be employed more specifically than is necessary in other contexts. Terminology laid out by other pro-jects, principally from the Cuneiform Digital Paleography Project (University of Birmingham and British Museum) are already useful.13

It is important to distinguish between the creation of a message or content in a document, and the wri-ting. An author is the person or persons responsible for the content of a document. A scribe is the person who actually impresses the graphs into the tablet. The term chirography may be preferable to hand in con-texts where clarity is needed, but the broadly accepted use of hand for the body of written graphs of any one scribe is acceptable and will be followed here.14 The written graphs as an object of examination on a tablet can be called ductus, though this term can also refer to larger corpora, such as the Old Assyrian ductus. In this contribution I use document and tablet interchangeably.

It is common practice to refer to the characters of the cuneiform writing system as signs, both in the sense of grapheme, a form which represents a word (logogram) or syllable (syllabogram), and in the sense of a graph, a specific impressed instance of a sign (grapheme). Grapheme and graph are used here to ensure clarity. As already discussed, idiograph denotes the form of a grapheme that is diagnostic for a particular hand. All graphs in a particular document are potentially idiographs, but it is not necessary to use all the graphs to distinguish a hand.

Every cuneiform graph is made up of one or more wedges. For chirographic purposes, five different wedge types can be distinguished in the Old Assyrian ductus, four denoted by their orientation: horizontal, vertical, downward diagonal, and upward diagonal. Each of these four wedge forms appears as a pyramid in negative with three faces (the surfaces of the impression), three outer edges (where the faces meet the surface of the tablet), three inner edges (where the faces meet one another), and one vertex (where the three inner edges converge). Each of the four wedges is distinguished by the direction which the tail (the long inner edge) points, leaving a head (the smallest of the faces) extending opposite the tail across the vertex. The fifth wedge type, the winckelhacken (or corner wedge), has an appearance somewhat like a boomerang. Like the other wedges, the winckelhacken has three faces, three inner and outer edges, but no head. Instead, it has an upper tail and lower tail, and opposite the two tails, a nose.

Particularly in the Old Assyrian corpus, graphs are hung from rulings, which normally extend the full width of the tablet, are laid down as the text progresses, and may change in angle to follow the bottoms of graphs.

Wedges can exhibit different attitudes as a result of their production. For example, one of the four basic wedge types may be inclined right or left, or upwards or downwards, so that one of the faces along the tail are much larger than the other.

The location of a wedge may be described as leading (in front, i.e. on the left) or trailing (at the end, i.e. on the right) or in relation to another wedge. Many graphemes consist of recognizable elements that may be shared with other graphemes, or even of themselves comprise a grapheme. When such an element is represented by a grapheme, it can be referred to by the name of the grapheme, but designated as an element, such as the ŠE element in the IM grapheme.

11 Ellison 2002.12 For a portion of the Šalim-ah

˘um tablets used in his case study, see Stratford 2014.

13 www.cdp.bham.ac.uk accessed April 2014.14 Most scribes have three hands: two attached to their arms (physical hands) and one hand or chirography that is the sum total of their written graphs. (Some scribes may only have two hands, one attached to their arm. But I can think of no scribe worth their salt who has only one hand).



122

Edward Stratford

1.4 Notation

The hand associated with an individual’s authored documents must be recognized as potentially distinct from the hand of the author himself, especially when authored documents from an individual exhibit more than one hand. As will be discussed below, Šalim-ah

˘um’s documents are a case where two different hands

are exhibited. As such, I will represent a hand, as distinct from the author, by the following notation: χ<Personal Name>A, where the χ stands for chirography, and the subscripted capital designates the various hands. Thus in the case of the two different hands in Šalim-ah

˘um’s documents, the first will be designated

χŠalim-ahumA and the second χŠalim-ah˘umB. Where it is necessary to add the patronym to the name to

avoid ambiguity, a hand can be designated χŠalim-ah˘um (s. Dān-Aššur)A. This is the approriate way to

designate it in an index, whereas a discussion that contextualizes the hand sufficiently does not necessitate extended notation. In cases where confidence is high enough to assign a hand to an individual, it will be designated in the same way, but without the subscripted capitals. In this study, the subscript is retained except in case where there is speecific evidence that the scribe and author are the same. This shoudl be con-sidered a best practice. If the hand of a particular text is being considered, then it is appropriate to express it as χ<Text>. In such cases, a subscript is only useful if there is some reason to think that there are two hands present on the tablet.

When displaying a frameset, it is useful to uniquely identify each graph displayed in the frameset table. This can be done by reference, including line number, with a (1) or (2), etc., to denote if it is the first, second, etc., graph of that grapheme on that line. Different readings of the same grapheme should not be distinguished when numbering of graphs on the same line.

When listing tablets subjected to chirographic analysis, it is appropriate to distinguish between tablets inspected by autopsy, designated with a superscripted a, i.e. BIN 4 9ª, from tablets inspected only through available photography, designated with a degree sign, i.e., TC 2 1°.

2. Case StudiesA short series of case studies allows for an opportunity to exhibit this method and show some of the poten-tial findings. In the first case, it is demonstrated that Šalim-ah

˘um’s letters were inscribed by two different

hands. In the second case, it is shown that differentiating two homonymous persons can be aided by chiro-graphic analysis. The remaining cases contribute to a sense that literacy was widespread. This seems to be borne out because it was equally likely that persons of higher or lower social rank could act as the scribe in documents where multiple authors are involved, unless someone’s representatives constituted a separate party among the authors, in which case they were more likely to act as one of the scribes. While all of these observations need to be followed up by further analysis, the initial finding at least show some of the subtle ways that hands can be distinguished and the potential of such analysis.

2.1 Šalim-ah˘um

In this first case, chirographic analysis problematizes the question of literacy for Šalim-ah˘um son of Dān-

Aššur, frequent correspondent of Pūšu-kēn. The primary autopsy subjects for this study were the relevant tablets in the British Museum and the Yale Babylonian Collection, and clearly show two different chi-rographies. While there are a number of idiographic characteristics distinguishing the two, the different UD idiographs are most easily recognized, to the point that in several of these cases, transliterations have marked UD graphs in χŠalim-ah

˘umB with an exclamation point. This is easy enough to distinguish on pho-

tographs as well, thus all documents solely authored by Šalim-ah˘um for which I could obtain photographic

representations and make an identification are listed here.15

χŠalim-ah˘umA (BIN 6 53ª, BIN 6 82ª, BIN 6 87ª, BIN 6 100ª, BIN 6 116ª, CCT 2 1ª, CCT 2 2ª, CCT 2 3ª,

CCT 4 25bª, POAT 7°, POAT 19°, TC 1 14°, TC 1 26°, TC 2 1°, TC 2 2°, TC 2 3°, TC 2 4°, TC 2 5°, TC 3 21°, TC 3 22°, TC 3 23°, VS 26 43°, VS 26 47°, VS 26 58°).

Idiographs – UD graphs exhibit two compact, well-executed winkelhacken in upper half of line, the vertical separated from the ruling (Fig. 1). MA graphs are characterized by long top and bottom horizon-tals, the head of middle horizontal trailing heads of top, and small leading vertical mostly occluded by top horizontal. ŠA graphs sometimes exhibit upward and downward diagonals instead of winkelhacken ahead of the trailing vertical.

15 Other tablets which I did not have images of include: TTC 6, MDOG 102, 86 (Eilsberger), KTS 1 27a, KTS 1 27b, KTS 1 28, KTS 1 41a, AKT 3 72, AKT 3 73, AKT 3 74, AKT 3 75, AKT 3 76. In several cases, hand copies suggested χŠalim-ah

˘umA, however, collation is necessary for chirography. I did not have time to examine CCT 2 4b during my

visit to the British Museum.

123


General Appearance – Many examples of χŠalim-ah˘umA (esp. CCT 2 2, CCT 2 3) have a careful

appearance with a well-spaced layout, straight rulings and graphs; wedges hang directly on the ruling, high-placed winckelhacken strike the ruling just above the nose, and full-line winkelhacken rarely cut the ruling below but handsomely transect their own ruling. More generous spacing when available seems to be a characteristic of this hand (BIN 6 87, BIN 6 92).

χŠalim-ah˘umB (BIN 4 25ª, BIN 4 26ª, BIN 6 113ª, BIN 6 171ª, BIN 6 202ª, CCT 2 4aª, CCT 4 5bª, CCT 5

5aª, TC 3 20°).

Idiographs – UD graphs appear more like a malformed A grapheme with a tall vertical ahead of two semi-verticals (Fig. 1). MA graphs are well-formed and rectangular, with the leading vertical more pro-minent than in χŠalim-ah

˘umA and apparently impressed last, always clearly discernible, often leading top

horizontal. Upper and middle horizontals are similar length, upper sometimes shorter. LÁ graphs exhibit upper horizontal clearly free below ruling. NA graphs are consistently narrow, with short leading horizon-tal. The ŠE graphs and ŠE elements in IN and LI graphs exhibit top row of four to six winckelhacken, and bottom row of two winckelhacken at leading and trailing edges of the top row.

General Appearance – This hand is executed more finely and compactly than χŠalim-ah˘umA. The two

scripts are quite similar at first look in general appearance, but can be confidently distinguished, especially by the UD graphs.

A single idiograph is normally insufficient to differentiate hands, but in the case of χŠalim-ah˘umA and Šalim-

ah˘umB UD idiographs are distinct enough to be diagnostic, and are supported by the analysis of other idio-

graphs. The clear difference in the UD idiographs makes it possible to differentiate a large number of letters from images provided by the Cuneiform Digital Library Initiative. Chirographic analysis corroborates Šalim-ah

˘um’s authorship of two acephalous letters, BIN 6 171 and BIN 6 202, by association with Šalim-ah

˘umB.

There is no way to establish which chirography was Šalim-ah˘um’s, or if either was. Both hands are

fine, and the UD idiograph in χŠalim-ah˘umB cannot be viewed as somehow deficient, suggesting a less

professionally trained Šalim-ah˘um, while χŠalim-ah

˘umA indicates a professional scribe. Further analysis

may offer some clarification.

2.2 Aššur-nādā son of Aššur-idīDistinguishing hands can sometimes serve to corroborate the distinction of homonymous persons. Aššur-nādā son of Aššur-idī is well known through his letters, most gathered in a publication of his archive. BIN 4 229, also authored by an individual named Aššur-nādā but not included by Larsen in the publication of Aššur-nādā son of Aššur-idī’s archive, exhibits a very different chirography, corroborating the supposition that the two are of different authors. This proposition awaits further corroboration because many of the tablets authored by Aššur-nādā were not able to be studied for this contribution.16

16 Including: ATHE 37, ATHE 36, ATHE 35, AKT 1 17, CCT 3 7a, CCT 3 7b, CCT 3 8b, CCT 3 9, CCT 3 10, CCT 4 10a, CCT 4 34b, KTS 1 12, KTS 1 13a, KTS 1 13b, KTS 1 33a, LB 1231, RA 60, 140a, SÉ 15.

Fig. 1: Frameset of UD idiographs for χŠalim-ah˘umA and χŠalim-ah

˘umB.



124

Edward Stratford

χAššur-nādā (s. Aššur-idī)A (BIN 4 49ª, BIN 4 52ª, RC 1749 C°, RC 1749 D°, TC 1 2°, TC 1 15°, TC 2 47°, VS 26 19°, VS 26 20°).

Idiographs – IM graphs display a leading element composed of a leading vertical, followed by two or three downward diagonals which are terminated by two upward diagonals impressed from left to right. ŠE elements at the leading edge of the LI is similar with two rows, five on top and two on the bottom that are both crowded to the left. DI graphs have three top winckelhacken written from bottom to top, with the base winckelhacken position towards the trailing end.

General Appearance – There is a lack of practiced placement of wedges and in managing the space between the rulings. Graphs leans heavily into the rulings, with the verticals surpassing the ruling. Several graphs have elements comprised of diagonals where winckelhacken are more common (AM, etc.).χAššur-nādā (≠ s. Aššur-idī)A (BIN 4 229).

Idiographs – IM graphs exhibit standard DI element composition. ŠE elements in the LI graphs exhibit a full row of wedges on both top and bottom. DI graphs exhibit an even top row, with a well-centered base winckelhacken.

General Appearance –Most graphs display a downward/right leaning. The name Aššur-nādā is rendered with a preceding DINGIR (dA-šur-nada), which does not appear in any of the tablets authored by Aššur-nādā s. Aššur-idī presented in Aššur-nādā #44-71.

2.3 Kurub-Ištar and Kulumaya

Several tablets were available in the Yale Babylonian Collection authored by Kulumaya, Kurub-Ištar, and one jointly authored by them. The two hands can be distinguished, showing that the hand associated with Kulumaya is found in the jointly authored letter (BIN 4 8).

χKulumayaA (BIN 4 13ª, BIN 4 8ª).

Idiographs – KI graphs and KI elements in NA graphs are rendered narrow and tall, with internal horizontals often traversing the trailing vertical (Fig. 2). IN graphs exhibit leading element with two wink-kelhacken, one each at leading and trailing edges of top row. UB graph in BIN 4 8 executed with trailing element composed of two converging diagonals.

General Appearance – Single horizontal wedges of signs fall just below and clear of the ruling. Condi-tion of the clay seems to have influences a more irregular rendition of many graphs.

χKurub-IštarA (BIN 4 16ª, BIN 4 17ª).

Idiographs – KI graphs and KI elements in NA graphs exhibit small horizontals sequestered within the triangular body, not extending past trailing vertical (Fig. 2). IN graphs rendered with leading element composed of two rows of several winckelhacken. UB graphs are executed with trailing element composed of three winckelhacken.

General Appearance – Single horizontals, such as leading horizontals, tend to be impressed high, close to the ruling. Graphs generously spaced on the tablets.

Fig. 2: Frameset of KI element idiographs for χKulumayaA and χKurub-IštarA.

125


2.4 Hinnaya and Puzur-Aššur

h˘innaya and Puzur-Aššur both corresponded with Pūšu-kēn through a number of letters available for auto-

psy. In one letter, both appear as correspondents, h˘innaya appearing first, alongside the representatives of

Pūšu-kēn. The hand of the co-authored letter aligns neither with χh˘innayaA nor χPuzur-AššurA, suggesting

that it was inscribed by one of the representatives of Pūšu-kēn.

χh˘innayaA (BIN 4 19ª, BIN 4 23ª, BIN 4 54ª, POAT 50°).17

Idiographs – NA graphs exhibit a medium leading horizontal, with horizontals sequestered by trailing verticals. DI graphs exhibit a leading winckelhacken that is preserved downward ahead of the base wink-kelhacken (Fig. 3). TIM and TUM graphs exhibit extended triangular trailing elements, especially where space permits. NI graphs exhibit a long upward diagonal and shorter downward diagonal trailing the two short verticals.

General Appearance – The lean of graphs in different tablets vary somewhat, sometimes vertical, and sometimes slightly right.

χPuzur-AššurA (BIN 4 221ª, BIN 6 221ª, BIN 4 27ª).

Idiographs – NA graphs exhibit short leading horizontal, with horizontals that traverse trailing vertical. DI graphs exhibit a top row in winckelhacken all terminating neatly in the base winckelhacken (Fig. 3). Trailing elements of IN, NI and TIM graphs exhibit top downward diagonal angled barely downward slightly below the ruling, with the upward diagonal angled more acutely upward.

General Appearance – In general appearance, similar to χh˘innaya with similar variability in lean.

χBIN 4 21ª.

Idiographs – NA graphs have short leading horizontal, with horizontals that occasionally traverse the trailing vertical slightly. DI graphs exhibit short winckelhacken on top row that align well with base wink-kelhacken, though terminate short of base (Fig. 3). Trailing elements of TIM graphs very narrow along vertical axis.

General Appearance – Higher tolerance for space at the end of lines. There is far more room between graphs than in χh

˘innayaA.

2.5 Aššur-imittī and Šū-h˘ubur

Aššur-imittī and Šū-h˘ubur were both frequent correspondents of Pūšu-kēn. In their co-authored letters,

Aššur-imittī is listed ahead of Šū-h˘ubur. It appears that χAššur-imittīB is responsible for BIN 4 24. Whether

this is an assistant of Aššur-imittī, or Aššur-imittī himself, will require further analysis.

χAššur-imittīA (BIN 6 24ª, BIN 6 34ª, CCT 2 44aª, CCT 2 44bª).

Idiographs – NA graphs exhibit three downward diagonals, terminated by a diagonal pointing down and toward the leading edge when stylus is blunt, but also NA graphs with short leading horizontal, strong wink-kelhacken to bound leading side of KI element, and trailing vertical with slight lean (both types occur on BIN 6 24). DI graphs exhibit three winckelhacken on top row, with strong but not large winckelhacken below.

17 Tablets for which I did not have photographs or time to inspect: CCT 2 26b, KTH 18, TTC 17, KTS 1 14b.

Fig. 3: Frameset of DI idiographs for χH˘ innayaA, χPuzur-AššurA, and χBIN 4 21ª.



126

Edward Stratford

General Appearance – Graphs are generally squat and simple.

χAššur-imittīB (CCT 2 41bª, BIN 4 24ª).

Idiographs – DI graph exhibits only two winckelhacken on top row. NA graphs exhibit trailing vertical with only slight lean, with a tall winckelhacken at top left corner of KI element that often crosses the ruling. The BA graphs exhibit short lower downward diagonal.

General Appearance – Rulings slant upwards early and heavily. Groups of horizontals are aligned well. Filler wedges are plentiful, but not excessive.

χŠū-h˘uburA.

Idiographs – NA graphs consistently show a short leading horizontal, followed by a lower dominant winckelhacken, and a upper smaller winckelhacken, anywhere from 3 to 5 horizontals, and a strong trai-ling vertical with slight right lean. The horizontals protrude past the trailing vertical sometimes, but the general characteristic is one of compactness. UD graphs are consistently “classical”. MA graphs display no prominent vertical at leading edge. ŠE element of IN graphs exhibit two full rows of four to five fine winckelhacken.

General Appearance – Generally compact, verticals almost vertical or slight lean right. Tendency to use many fill wedges.

2.6 Ilī-ālum

Letters authored individually by Ilī-ālum (BIN 6 64, POAT 8, RC 1479E, VS 26 29) as well as letters authored by Ilī-ālum and two others of higher social rank, Lā-qēpum and his older brother Aššur-nādā offer an opportunity to query the prestige of literacy in relation to social rank. While the chirography of Ilī-ālum is clear and easy to discern in both individual and co-authored situations, a clear example of a χLā-qēpum is difficult to find. Admittedly, this weakens the argument that all are Ilī-ālum’s hand, or at least demands that the comparative operation is of similarity only. KTH 9 may provide an object of comparison, but there is sufficient uncertainty about the precise author of the document, and sufficient certainty about χIlī-ālum s. Aššur-idī, that χIlī-ālum s. Aššur-idī is the only one reviewed here.

χIlī-ālum s. Aššur-idī (BIN 4 51ª, BIN 4 61ª, BIN 6 64ª, POAT 8°, RC 1479E°, VS 26 29°).

Idiographs – MA graphs exhibit a stout leading vertical with a head placed ahead of, but also below the top horizontal, and the graphs are generous in their widths. DI graphs are anchored by the top winckelhak-ken, with the bottom winckelhacken in the row executed more narrowly, inclined downwards along its axis. TIM, TUM, NI graphs exhibit elongated triangular trailing elements, hung high on the ruling.

General Appearance – Ductus is ordered, compact, but well-managed. When executed with a fine sty-lus, the ductus can be very fine.

3. DiscussionWhile only an initial study, the cases reviewed above give a sense that chirography is useful for under-standing literacy, and for the context of document creation in some cases. In the cases of Šalim-ah

˘um and

Aššur-imittī, multiple hands can be associated with these authors. It is difficult at this point to be able to say whether or not the multiple hands are evidence of copies of letters made in Kanesh, or of two different hands active at the moment of creation of the first document, back in Aššur. Both are know to operate and generally stay in Aššur. Chirographic analysis seems to also be useful for prosopographical distinction. The consistency in hands among the letters authored by Aššur-nādā s. Aššur-idī and the difference between them and the letter from a different Aššur-nādā is consistent with the distinction made through reading the texts themselves. Generally speaking, the ability to find matches for hands across single authored texts supports that many of the peregrinating merchants, such as Aššur-nādā, Ilī-ālum, Kulumaya and Kurub-Ištar, both authored and wrote their own documents most of the time. This initial review offers no revision of the general sentiment that most persons involved had at least the craft literacy to read and write their own letters. This is further supported by the documents authored by multiple persons. While in the case of Kulumaya and Kurub-Ištar the higher ranked Kulumaya inscribed their joint letter, in the case of Ilī-ālum and his brother Ilī-ālum, it was the younger brother who acted as scribe. It is equally possible that the per-son with the clearer hand was selected for the job, which would be consistent with Ilī-ālum, whose hand is well executed. Kulumaya’s hand also appears more skilled. At the same time, the social proximity to the addressee may have been decisive, as in the case of the letter written by h

˘innaya, Pūšu-kēn’s representa-

tives, and Puzur-Aššur. In that case, because the hand did not match either of the named addressors, the representatives are the most likely responsible party. Finally, Aššur-imittī’s multiple hands, of which one is responsible for the joint letter with Šū-h

˘ubur, allows the decision for who wrote to either be Aššur-imittī,

127


or his representative. The variation exhibited in these various cases still supports the notion that literacy was broadly held enough so that merchants held no notions that writing a letter for a group was a privilege reserved for the ranking member of the social group.

There are a number of other potential questions that can be asked in addition to the distinctions made here. For example, it is possible to evaluate who representatives of a merchant are when they are addressors if a match can be found for the hand in likely candidates. In addition, there is the possibility to evauate the literacy of female writers, the hands involved in other types of documents, and to further explore the numebr of hands associated with permanent residents of Aššur as opposed to perigronating merchants. In this last case, it will be useful to compare hands present in letters authored by residents of Aššur with the recipients of those letters. While the hand of Pūšu-kēn could not be presented in full here, at least initial review of his hand as opposed to χŠalim-ah

˘umA and χŠalim-ah

˘umB suggests that neither hand is that of

Pūšu-kēn. In other words, he was not copying the letters in his own hand after they arrived.

The general appearance of a tablet, the layout of the graphs, can be an initial indicator of different hands. For example, the tendency for more fill wedges in χAššur-imittīB as opposed to χAššur-imittīA. The tendency to leave more spacing also seems to some extent diagnostic, as in the difference between χKulumayaA and χKurub-IštarA. It is also consistent with the differentiation between χBIN 4 21 and both χh

˘innayaA and χPuzur-AššurA. However, similarity of general layout is not consistent with differentiating

features at the level of idiographs. On grounds of layout, one might propose that χh˘innayaA and χPuzur-

AššurA are both from the same hand, but at the idiographic level, this assumption would be overturned.

*

Many other studies have dealt with this topic, and often informally, within the process of otherwise editing texts or analyzing archives. By all accounts, chirographic comparison is subjective. However, it can provide some fodder for interesting social questions and has the potential to enrich our understanding of social structures and sometimes interpersonal dynamics of certain situations, or add context to the con-tent of a letter. It appears, at this early moment, that craft literacy, the ability to read and write documents applicable to the trade was widespread, and that this skill need not be displayed in association with social rank, suggesting it was regarded as common enough among the merchants that it was not a significant sign of status. It is hoped that this brief study gives some beginning to a discussion further refining the modes of employing chirographic research on the Old Assyrian trade in a way that will further expose the manifold layers of information that are embedded in a coprus of documents created by many of the people whose names we already know. Chirography, in conjunction with other approaches offers yet another opportunity to mine the rich world provided by the archaeological and textual remains found at the site of Kültepe.

BibliographyAnsell, M.

1979 “Handwriting Classification in Forensic Science”, Visible Language 13, 239–251.

Davis, T.

2007 “The Practice of Handwriting Identification”, The Library 8, 251–276.

Edzard. D. O.

1978 “Keilschrift”, in: E. Ebeling, B. Meissner, E. Weidner, W. von Soden and D. O. Edzard (ed.), in: Reallexikon der Assyriologie und vorderasiatische Archaologie, Berlin and Leipzig, 544–568.

Ellison, J. L.

2002 A Paleographic Study of the Alphabetic Cuneiform Texts from Ras Shamra/Ugarit, Ph. D. Disser-tation, Harvard University, Department of Near Eastern Languages and Civilizations.

Huber, R. A. – Headrick, A. M.

1999 Handwriting Identification: Facts and Fundamentals, New York.



128

Edward Stratford

Kryszat, G.

2008 “The use of writing among the Anatolians”, in: J. G. Dercksen (ed.), Anatolia and the Jazira during the Old Assyrian Period, Old Assyrian Archives Studies 3, Publications de l’Institut historique- archéologique néerlandais de Stamboul 111, Leiden, 231–238.

Marquis, R. – Schmittbuhl, M. – Mazzella, W. D. – Taronie, F.

2005 “Quantification of the Shape of Handwritten Characters”, Forensic Science International 150, 23–32.

Michel, C.

2010 “Le langage figuré dans les lettres paléo-assyriennes: Expressions relatives à l’homme et à la nature”, in: L. Kogan (ed.), Proceedings of the 53e Rencontre Assyriologique Internationale, Babel und Bible 4, 347–376.

2011 “The Private Archives from Kaniš Belonging to Anatolians”, Altorientalische Forschungen 38, 94–115.

Powell, M. A.

1981 “Three Problems in the History of Cuneiform Writing: Origins, Direction of Script, Literacy”, Visible Language 15, 419–440.

Stratford, E.

2014 “‘Make them pay’: Charting the Social Topography of an Old Assyrian Caravan Cycle”, Journal of Cuneiform Studies 66, 11–38.

Old Assyrian Literacy: Formulating a Method for Graphic Analysis and Some Initial Results

Documents