Silver, Money and Credit: A Tribute to Robartus J. van der Spek on the Occasion of his 65th Birthday 9789062583393

Citation preview

SILVER, MONEY AND CREDIT A TRIBUTE TO ROBARTUS J. VAN DER SPEK ON THE OCCASION OF HIS 65TH BIRTHDAY

edited by KRISTIN I

Cuneiform documents on Parthian history: The Ral].imesu archive. Materials for the study of the standard of living. In:]. Wiesehofer (ed.), Das Partherreich und seine Zeugnisse. The

ressources > sous presse.

Arsacid Empire: Sources and Documentation. Beitrage des internationalen Colloquiums, Eutin (27.-30. juni 1996) (Historia Einzelschriften 122). Stuttgart, pp. 205-258.

Darius Ill, Alexander the Great and Babylonian scholarship. In: W Henkelman & A. Kuhrt (eds.), A Persian Perspective. Essays in Memory of Heleen Sancisi- Weerdenburg (Achaemenid History XIII). Leiden, pp. 289-346.

Review of: R.Wallenfels, Uruk. Hellenistic Seal Impressions in the Yale Babylonian Collection l Cuneiform Tablets (Ausgrabungen in Uruk-Warka. Endberichte Band 19), Mainz am Rhein 1994. MO 44/45, pp. 427-428.

2004

1999

Palace, temple and market in Seleucid Babylonia. In: V. Chankowski & F. Duyrat (eds.), Le Roi et l'Economie. Autonomies locales et Structures royales dans l'Economie de /'Empire seleucide. Actes des Rencontres de Lille (23 ]uin 2003) et d 'Orlians (29-30 ]anvier 2004)

Godsdienst en oorlog in de oudheid en nu. In de Marge. Tijdschrift voor Levensbeschouwing en Wetenschap 8/2, pp. 2-9.

(Topoi Supplement 6). Paris, pp. 303-332.

2000 The effect of war on the prices of barley and agricultural land in Hellenistic Babylonia. In: J. Andreau, P. Briant & R. Descat (eds.), Economie antique. La Guerre dans les Economies antiques (Entretiens d'Archeologie et d'Histoire 5). Saint-Bertrand-de-Comminges, pp. 293-313. The Seleucid state and the economy. In: E. Lo Cascio & D. Rathbone (eds.), Production and Public Powers in Antiquity (Cambridge Philological Society Supplementary Volume no. 26). Cambridge, pp. 27-36.

& FINKEL, I.L.: Babylonian Chronicles of the Hellenistic Period. http://www.livius.org/ babylonia.html

2005 Commodity prices from Babylon (385-61 BC). http://www.iisg.nl/hpw Ethnic segregation in Hellenistic Babylon. In: WH. van Soldt (ed.), in cooperation with R. Kalvelagen & D. Katz, Ethnicity in Ancient Mesopotamia. Papers Read at the 48th Rencontre Assyriologique Internationale, Leiden, 1-4 July 2002. Leiden, pp. 393-408. lheopais Babylon. Een multiculturele stad in de Hellenistische tijd. Lampas 38, pp. 198-214.

XXIV

BIBLIOGRAPHY OF

R.j.

BIBLIOGRAPHY OF

VAN DER SPEK

Review of: H. Hunger, A.J. Sachst & ].M. Steele, Astronomical Diaries and Related Texts from Babylonia. Vol. V: Lunar and Planetary Texts, Wien 2001. BiOr 62, pp. 546-553.

R.J. VAN

DER SPEK

XXV

2010 Seleukiden, Seleukidenreich (Seleucids, Seleucid empire). In: M.P. Streck, S. Ecklin & S. Pfaffinger (eds.), Reallexikon der Assyriologie und vorderasiatischen Archiiologie, Band 12. Berlin, pp. 369-383.

Review of:].-]. Glassner, B.R. Foster (ed.), Mesopotamian Chronicles, Atlanta 2004. Review ofBiblical Literature. http://www. bookreviews.org

Alexander de Grote in het Midden-Oosten. Kleio 51/5, pp. 4-7.

2006

Alexander in spijkerschrift. A/lard Pierson Mededelingen 103, p. 9. How to measure prosperity? The case ofHellenistic Babylonia.In: R. Descat (ed.),Approches de l'Economie hellinistique (Entretiens d'Archeologie et d'Histoire 7). Paris, pp. 287-310.

Review of: H. H. Schmitt & E. Vogt (eds.), Lexikon des Hellenismus, Wiesbaden 2005 . BiOr 67, pp. 244-247.

Grieken in de ogen van Babyloniers. Leidschrift2113, pp. 139-156. The size and significance of the Babylonian temples under the Successors. In: P. Briant & F. Joannes (eds.), La Transition entre /'Empire achemenide et les Royaumes hellenistiques (vers

2011 The "silverization" of the economy of the Achaemenid and Seleukid empires and early modern China. In: Z.H. Archibald, ] .K. Davies & V. Gabrielsen (eds.), 7he Economies of Hellenistic Societies, 7hird to First Centuries. Oxford, pp. 402-420.

350-300 av. j-C). Actes du Colloque organise au College de France par la« Chaire d'Histoire et Civilisation duMonde achemenide et de /'Empire d'Alexandre »et le« Reseaulnternationald'Etudes et de Recherchesachemenides » (GDR2538 CNRS), 22-23Novembre 2004. Paris, pp. 261-307. & HuNGER, H .: An Astronomical Diary concerning Artaxerxes 11 (year 42 = 363-2): Military operations in Babylonia. ARTA 2006.002, pp. 1-6. http://www.achemenet.com/ ressources/ enligne/ artal pdf/2006. 002-Hunger-Spek. pdf

Review of: A.L. Slotsky & R. Wallenfels, Tallies and Trends. 7he Late Babylonian Commodity Price Lists, Berhesda 2009. Akkadica 132, pp. 182-184.

2012 , GRIFFITH, G.T. & SHERWIN-WHITE, S.: Alexander (1 0) (RE 22) Balas. In: S. Horn blower, A. Spawforth & E. Eidinow (eds.), 7he Oxford Classica/Dictionary. Oxford (fourth edition), p. 59.

2007 The Hellenistic Near East. In: W Scheidel, I. Morris & R. Sailer (eds.), 7he Cambridge Economic History ofthe Greco-Roman World. Cambridge, pp. 409-433.

,GRIFFITH, G.T. &SHERWIN-WHITE,S.:Antiochusl (1) (RE21) Sorer. In: ibid.,pp.104-105. , GRIFFITH, G.T. & SHERWIN-WHITE, S.: Antiochus 11 (2) (RE 22) Theos. In: ibid., p. 105.

2008

, GRIFFITH, G.T. & SHERWIN-WHITE, S.: Antiochus lii (3) (RE 23) Megas. In: ibid., p. 105. & STOL, M.: 'The Antiochus Cylinder,' http:/ /www.livius.org > Mesopotamia. (ed.), with the assistance of G. HAAYER, EA.M. WrGGERMANN, M. PRINS & J. BrLBIJA, Studies in Ancient Near Eastern World View and Society Presented to Marten Stol on the Occasion ofHis 65th Birthday, I 0 November 2005, and His Retirement from the Vrije Universiteit Amsterdam. Bethesda (MD). Berossus as a Babylonian chronicler and Greek historian. In: ibid., pp. 277-318. Feeding Hellenistic Seleucia on the Tigris and Babylon. In: R. Alston & O.M. van Nijf (eds.), Feeding the Ancient City. Leuven, pp. 33-45. & BLOIS, L. DE: An Introduction to the Ancient World. Abingdon (second edition).

, GRIFFITH, G.T. & SHERWIN-WHITE, S. : Antiochus IV (4) (RE 27) Theos Epiphanes. In: ibid., p. 105. , GRIFFITH, G.T. & SHERWIN-WHITE, S.: Antiochus V (5) (RE 28) Eupator. In: ibid., p. 105. , GRIFFITH, G.T. & SHERWIN-WHITE, S.: Antiochus VI (6) (RE 29) Epiphanes Dionysius. In: ibid., pp. 105-106. , GRIFFITH , G.T. & SHERWIN-WHITE, S.: Anriochus VII (7) (RE 30) Sidetes. In: ibid., p. 106. , GRIFFITH, G.T. & SHERWIN-WHITE, S.: Antiochus (8) (RE 23) Hierax. In: ibid., p. 106. , SHERWIN-WHITE, S. & GRAY, E.W : Carrhae. In: ibid., p. 283. , GRIFFITH, G.T. & SHERWIN-WHITE, S.: Laodice (2) (RE 13). In: ibid., p. 792.

2009

& SHERWIN-WHITE, S.: Laodice (3) (RE 16). In: ibid., p. 792. Multi-ethnicity and ethnic segregation in Hellenistic Babylon. In: T. Derks & N. Roymans (eds.), Ethnic Constructs inAntiquity. 7he Role ofPower and Tradition. Amsterdam, pp. 101-115.

, DROWER, M .S. & SHERWIN-WHITE, S.: Seleuceia (1) on Tigris. In: ibid., p. 1340. ']oNES, A.H.M., SEYRIG, H. & SHERWIN-WHITE, S.: Seleuceia (2) in Pieria. In: ibid., p.

1340-1341.

XXVI

BIBLIOGRAPHY O F

R.).

VAN DER SPEK

& SHERWIN-WHITE, S.: Seleucids. In: ibid., p. 1341.

, GRIFFITH, G.T. & SHERWIN-WHITE, S.: Seleucus (1) (RE 2) Nicator. In: ibid., p. 13411342. , GRIFFITH, G.T. & SHERWIN-WHITE, S.: Seleucus (2) (RE 4) Callinicus. In: ibid., p. 1342. & SHERWIN-WHITE, S.: Seleucus (3) (RE 5) Ceraunus. In: ibid., p. 1342.

, GRIFFITH, G .T. & SHERWIN-WHITE, S.: Seleucus (4) (RE 6) Philopator. In: ibid., p. 1342. & SHERWIN-WHITE, S.: Stratonice (RE 8) . In: ibid., p. 1406.

BIBLIOGRAPHY OF

R.).

VAN DER SPEK

XXVII

The volatility of prices of barley and dates in Babylon in the third and second centuries BC. In: H.D. Baker & M. Jursa (eds.), Documentary Sources in Ancient Near Eastern and GrecoRoman Economic History: Methodology and Practice. Oxford, pp. 234-259. & LEEUWEN, B. VAN: QuantifYing the integration of the Babylonian economy in the Mediterranean world using a new corpus of price data, 400-50 BC. In: F. de Callatay & A Wilson (eds.) , Long-Term Quantification in Ancient History. Bari, pp. 79-101. & WALLENFELS, R.: 'Copy of record of entitlement and exemptions to formerly royal lands. ' (text no. 148 = 'Lehmann Text'; land grant by Antiochus 11) In: I. Spar & M. Jursa (eds.?, 1he

Economy, Near East (Hellenistic). In: R. Bagnall et al. (eds.), 1he Encyclopedia ofAncient History. Chichester etc., pp. 2285-2288.

Ebabbar Temple Archive and Other Textsfrom the Fourth to the FirstMillennium B. C (Cuneiform

Review of: L. Capdetrey, Le Pouvoir seleucide. Territoire, Administration, Finances d'un Royaume hellenistique (312-129 avantJC), Rennes 2007. Klio 94, pp. 527-529.

'7k ben een boodschapper van Nanaia!" Een Babylonische profeet als teken des tijds (I 33 v. Christus). Rede uitgesproken bij zijn aficheid als hoogleraar Oudheidkunde, in het bijzonder de Mediterrane en West-Aziatische Geschiedenis, aan de Faculteit der Letteren op I 0 oktober 2014.

2013 Alexander in spijkerschrift. In: D. Burgersdijk, W Henkelman & W Waal (eds.), Alexander en Darius. De Macedoniiir in de Spiegel van het Nabije Oosten. Hilversum, pp. 81-99. Review of: S. Dalley, 1he Mystery ofthe Hanging Garden ofBabylon. An Elusive World Wonder Traced, Oxford. Tijdschrift voor Geschiedenis 126, pp. 580-581.

Texts in the Metropolitan Museum of Art IV). New York & Winona Lake, pp. 213-227.

Vrije Universiteit. Amsterdam. Review of: B. Jacobs & R. Rollinger (eds.), Der Achamenidenhojl1he Achaemenid Court, Wiesbaden 2010. Bi0r71, pp. 492-496. Forthcoming & VAN LEEUWEN, B. (eds.): Coins, Currencies and Crises. On Money and Trust from c. 2000

2014 Cyrus the Great, exiles and foreign gods. A comparison of Assyrian and Persian policies on subject nations. In: W Henkelman, C. ]ones, M. Kozuh & C. Woods (eds.), Extraction and Control.· Studies in Honor ofMatthew W Stolper (Studies in Ancient Oriental Civilizations 68). Chicago, pp. 233-264. Factor markets in Hellenistic and Parthian Babylonia (331 BC-224 AD).]ESHO 57/1, pp. 203-230.

BC-c. AD 2000.

, Bwrs, L. DE & WIESEHOFER, J.: Einfohrung in die Alte Welt. Stuttgart. , Bwrs, L. DE: An Introduction to the Ancient World. London & New York (third revised edition). Coming to terms with the Persian Empire. Concluding remarks and responses. In: J.M. Silverman & C. Waerzeggers, Political Memory In and After the Persian Empire. Atlanta (GA). De Cyruscilinder als een icoon van antieke en moderne propaganda. Lampas.

Seleukos, self-appointed general (strategos) of Asia (311-305 B.C.). In: H . Hauben & A. Meeus (eds.), 1he Age ofthe Successors (323-276 BC) (Studia Hellenistica 53). Leuven, pp. 323-342. , LEEUWEN, B. VAN & ZANDEN,].L. VAN (eds.): A History ofMarket Performance from Ancient Babylonia to the Modern World. London (2015]. , LEEUWEN, B. VAN & ZANDEN, ].L. VAN: An introduction: Markets from ancient Babylonia to the modern world. In: ibid., pp. 1-16. , LEEUWEN, B. VAN & FoLDVARI, P.: Growing silver and changing prices: The development of the money stock in ancient Babylonia and medieval England. In: ibid., pp. 489-505. , LEEUWEN, B. VAN & ZANDEN, J.L. VAN: Markets from Babylon to Belfast: Some concluding remarks. In: ibid., pp. 529-544.

Review of: J. Haubold, G .B. Lanfranchi, R. Rollinger & J. Steel (eds.), 1he World ofBerossus, Wiesbaden 2013. Zeitschrift for altorientalische und biblische Rechtsgeschichte. Review of: A. Kuhrt, 1he Persian Empire. A Corpus of Sources from the Achaemenid Period. Vol. I & 11, London & New York 2007. ARTA. Review of: P.J. Kosmin, 1he Land ofthe Elephant Kings. Space, Territory, and Ideology in the SeleucidEmpire, Cambridge (MA.) & London 2014. Mnemosyne.

TABULA GRATULATORIA ]oris AARTs

Jorrit KELDER

Gerard BoTER

David KERTAI

Gert Jan BuRGERS

Kristin KLEBER

Bas VAN BAVEL

Diana I.)

-.....]

-----" c .. - - .. ___

j

138

WF.M. HENKELMAN and M.L. FoLMER

to his credit, he simultaneously recognized that the loss "in the dramatic impact of [the] document" is balanced by having an administrative document on a routine movement of a general or satrap of the Persian empire. It is indeed a felicitous turn that the debate can now concentrate fully on the administrative value of this unique document, which actually tallies well with texts in the Persepolis Fortification archive (509-493 BCE) dealing with travels of the king's court, that of several royal women and a satrap, Karkis. 9 All tally sticks published in ADAB except for one (0 18 is undated) are dated to the third year of a Oarius, probably the last Achaemenid king, Oarius Ill (c£ §2.1 below). This means that the 23 dated documents among those published by Naveh and Shaked fall into two or three categories: those from the early years of Artaxerxes Ill (Al-A4), those from the last years of Achaemenid control (01-017 and, presumably, Cl) and, finally, a single text from year seven of Alexander (C4). If this last document actually dates to 330 instead of 324 BCE (c£ n.3 above), there would be only two categories, the first dating to 353-348, the second dating to 337-330 BCE. In archival terms, the chronological division (as stated here) is quite understandable. Most leather documents are clearly drafts; they were kept for two possible reasons. First, the occurrence of palimpsests shows that the valuable material was re-used; some of the documents in ADAB may have been intended as scrap 'paper.' 10 Secondly, the drafts of more important documents served as archival copies; this must certainly be true for Al-A4, which are official letters between the highest officials. By contrast, the split tallies probably had, as we will argue, a relatively short validity and would be replaced regularly (perhaps annually). Their actual existence (and intactness) in itself suggests a deviation of normal procedure, perhaps a glitch in the system or an administrative reset of some kind (c£ §2.1). From this perspective their low date is probably not a coincidence. Yet, whatever may have caused the disruption, it apparently did not obliterate the administrative structures set up in Bactria - such is shown by C4, an important witness to continuity in a period of transition. 11 Note that this document is a list of rations, hence of a category typically kept for a shorter period than official correspondence and often found in the youngest layers of an archive. 12 To come back to Bessus: classical sources indicate that he was satrap of Oarius Ill in Bactria at the time of Gaugamela, where he commanded the Bactrians, Sogdians and the

have to be given up. The matter therefore hinges on the interpretation of the name. 9

For a discussion of these Fortification texts see Henkelman 2010; compare the remarks of Briant 2009, 149, and Tuplin 2010, 123 on Cl. For recent surveys on the Fortification archive see: Garrison/Root 2001, 1-60; Briant 2002, 422-48, 938-43; Kuhrt 2007, 763-814 (selection of commented translations); Henkelman 2008, 65-179; id. 2013; id forthc. 1; Garrison forthc. 1. A recently identified Fortification text, Fort. 1348103, dates to 20/II/35 Dar.I, i.e. May 487 BCE.

10

C£ ADAB, 16£ and Folmer forrhc.

11

On this aspect, see the reflections of, among others, Briant 2009, 147-151, 168; id. 2010, 178-180; id. forthc.; and Kuhrt 2014, 124. 12

Document Cl, a list of provisions and probably dated to 337/336 BCE (c£ above), fits the same scenario.

YouR TALLY rs FuLL!

139

Indians bordering Bactria. 13 Assuming that his appointment had not been made on the spot, he must have had the highest authority in Bactria for at least a short time, hence perhaps also in year 3 of Darius Ill, when the split tallies were issued. With due caution (for the interpretation of the name bys as Bessus is not entirely certain), one may add document Cl to his dossier, implying that he was already active in Bactria and in office by 337/336 (year 2 of Arses/Artaxerxes IV). If correct, document Cl probably relates to an inspection tour that the Achaemenid satrap in Bactria undertook ex officio and for which he needed vast quantities of provisions for the splendour (his satrapal 'table' - c£ above), piety (sacrifices to local gods) and magnificence (distributions) befitting the representative of the King of Kings. As for the inspection tour as such: it is paralleled by the journeys of Parnakka, the leading official in the Persian heartland under Darius 1. 14 Bessus is, however, not the only satrap or leading administrator active in Bactria in the period covered by ADAB. First, C4, from year seven of Alexander, by its mere existence implies a local hierarchy headed by an official answerable to the central authorities. 15 Secondly, in the official correspondence from the earlier part of the archive, we find another team of administrators: Axvamazda ( 'flmzd) and Bagavanta (bgwnt). 16 The designation of the first is not mentioned, but is plausibly assumed to have been satrap; Bagavanta was clearly his subordinate and is introduced as p/Jt', "governor," of a place called /Jlmy. Shaked read this name as Khulmi, and convincingly identified it with modern Khulm/Ta$-Kurghan, some 80 km west of Balkh (ancient Bactra), which he conjectured to be the seat of Axvamazda and the probable origin of

1 3

Gaugamela: Arr. Anab. III.8.3; c£ Curt. IV.l2.6. See also Curt. V.9 .8 (praefectus regionis), Diod. XVII. 73.2, 4 and 74. 1, where he is described as having been appointed by Darius Ill and being known to everyone in the region because of his administration (on this passage and its significance, see Briant 2002, 871). 14

Parnakka may have been the satrap of Parsa, or at least have held a position of equal prestige (see Henkelman forthc. 1). 15

After Bessus, Alexander made Artabazus sarrap of Bactria (Arr. Anab. III.29.1), then replaced him, at his own request, first by Clirus (Curt. VIII.1.19), and after the latter's death, by Amyntas son ofNicolaus (Curt. VIII.2.14; c£ Arr. Anab. IV.l7.3). Under whose authority C4 was written, depends on the interpretation of the date formula (cf. n.3 above). Robin Lane Fox finds support in document C4 for his thesis that only the 'lower hierarchy' of Achaemenid administration survived in Alexander's short-lived empire (Lane Fox 2007, 297, referring to his earlier remarks on p. 291). This, however, is an argument from silence as C4 documents a series of responsible administrators, bur simply does not refer to the officers at the absolute top of the whole system. This silence is not eloquent, because a ration list of this type (as opposed to official correspondence) would not mention the leading ranks of the system. What is, by contrast, rather meaningful is that the individuals mentioned are part of an intricate hierarchy with parallels in the Achaemenid world, as Tuplin demonstrates in a recent study (forthc.). Oddly enough, one of the elements of Lane Fox's demonstration of essential change at the top levels of administration is Peucestas' sacrificial banquet at Persepolis (2007, 291), which turns out to have been a showcase of continuity of Achaemenid royal traditions under Alexander (Henkelman 2008, 237, 289, 385-416, 419-21, 425, 549f. and id. 20lla). 16 The normalization of the names of Axvamazda and Bagavanta is based on the constructed Iranian forms as given by Tavernier forthc. 1 §§ 2.4.1-2. The change of leading administrators may in fact be one of the factors behind the chronological distribution of the dated documents.

YouR TALLY rs FuLL!

WF.M. HENKELMAN and M.L. FoLMER

140

the documents. Bactra itself is referred to in the texts as Baxtri (bbtry). Other toponyms with suggested interpretations and identifications by Shaked include: kf kw my mytnk nbfpy wmw

Kis Kiimi Mait(h)anaka Nikhsapaya ('Xenippa,' Curt. VIII.2.14) Varnu (Gk. 'Aopvos; Sogd. oapvo)

Kis(s)/Sahr-e Sabz, c. 80 km SSW of Samarqand mediaeval Kiim, c. 60 km ESE of Pangikant Chasma Maidanak?, c.60 km SSE ofMazar-e Sarif Nakhsab/Qarsi, c.140 SW ofSamarqand near Qundiiz?, c.180 km E ofBalkhY

Whereas the details of the suggested topography require further study, the general outlines have found general acceptance: the documents reflect an Achaemenid administration in Bactria, presumably centred on Bactra/Balkh, with important dependencies at Khulmi and Varnu, and with a responsibility stretching across the Oxus, into Sogdia, where it had a number of strongholds. 18 The area is vast, but the parallel case of the Persepolis Fortification archive shows that such larger regions could be controlled by a central administration and that places even far apart could be nodal points in the same institutional network. Assuming that the split tallies are from the same source as the leather documents, they potentially pertain to this same wide network, though not necessarily to its centre. The split tallies from Achaemenid Bactria were first presented by Naveh and Shaked in a 2003 paper. Ever since, most scholarly discussion has concentrated on the leather documents in the Khalili collection. Though these documents add a wealth of information on Achaemenid administration systems and their persistence down to the very end of the empire and beyond, we feel that the tallies merit more attention than they have hiterto received. In this, we follow Briant's call for a review of the ADAB materials in comparison with other pertinent sources, which he deemed "a prerequisite for improving Achaemenid studies" (Briant forthc.). We note that the same scholar was early to recognize the specific importance of the tallies, if only because they belong to the rare category of documents dating to the last Achaemenid king. 19

141

Naveh and Shaked introduced the Achaemenid Bactrian tallies in a discussion on the Mishnaic and Talmudic legal terms i:;)illl, "break," and io/i?, "knot," both of which, as they argue, could be used metaphorically for "written receipt," especially in situations of indebtedness such as loans and taxation (11lf.). 20 They furthermore argued that the origins of the two concepts lie, in case of the 'knot,' in the widespread use of knotted memory cords (such as the Inca quipu; c£ §4 below) and, in case of the 'break,' in split wooden tally sticks and, more precisely, in either the splitting of the stick or the breaking of the tallies after termination of the agreement (112, 116; see also n.20). Thirdly, they proposed that the inscriptions together with the notched tally marks on the Achaemenid Bactrian tallies point to indebtedness (112-114) and fourthly they briefly compare these split tally sticks with similar but later objects from Bactria, and with finds from Khotan, Sogdia and Yemen, as well as the tokens referred in the book of To bit (114-116). Essentially the same comments, but in more concise form, are given inADAB (31-33, 233; also Shaked 2004, 18-21, 54), which also has the editions, translations, comments and excellent photographs of the tallies (230-258). What prompted the present study is Naveh and Shaked's explanation of the use of split tally sticks in Achaemenid Bactria. Most of these eighteen documents (D 1-D 18 in ADAB) have an Aramaic inscription that the editors render as "with ('m) PN I> from (mn) PN 2, in the year 3 of Darius the King" and take as indicating a debtor (PN 2) transferring to a creditor (PN 1) a certain commodity, the quantity of which is indicated by the notches on the edges of the split tally halves. As to the nature of this commodity involved, they wisely confess uncertainty, but venture the opinion that it might have been "a standard type of commodity, perhaps flour, so that no specification was necessary'' (ADAB, 31; not in Naveh/Shaked 2003). The reason to think of flour is not motivated, but clearly inspired by 02, which has an additional, largely enigmatic statement in which the word bwr, "white," occurs. This term, as Naveh and Shaked comment (ADAB, 237) "usually indicates white flour, which is considered to be of good quality." From this, it becomes understandable that they thought, at least in their statement in ADAB (p. 32), of the split tallies as records of rations issued to workers or soldiers: For our inscribed tallies from Bactria, it is possible to suggest the following scenario: The person whose name follows the preposition "with" was indebted for the payment that he got (probably in provisions); his creditor is the person whose name comes after "from," and who provided certain items the quantity of which is indicated by the number of the incisions. The indebtedness may in this case consist of rations allotted to workers or soldiers by the treasury. Since there are only three providers in this collection, and about fifteen recipients, it seems likely that the tallies reflect a situation in which three government clerks issued goods to

17

On these and other topographical identifications, see especially Shaked 2003; also id. 13(, 17, 31 and ADAB, 16-22, 24(, 81, 97-99, 106, 113, 121, 180f., 207, 225, 259-60. The identification of 'mynkn (Amainakana?) as mediaeval Minak (Sogdia) is problematic because it may in fact be an anthroponym, as Tuplin forthc. argues. 18

Compare the studies of Rapin on the geography of the region in which the identifications of Khulmi, Kis(s) and Nikhsapaya are accepted and archaeological sites are suggested for the last two (Rapin 2004, 148 n.28, 149 n.30, 160 n.64; id. 2013, 50, 70). On the Achaemenid layers exposed in the excavations at the Bala Hissar (Balkh) and at Kafir Qala (Chasma Safa), as well as Achaemenid sites (including a circular fortification of c. 350 m in diameter north of Bargah -to be compared with Briant's remarks [forthc.J) found in various surveys in the Balkh hinterland, see the extensive reports by Bernard/Besenval/Marquis 2006, Besenval/Marquis 2007, eid. 2008, Besenval/Marquis/Fouache 2009. On the administrative connection of Sogdia and Bactria, see Jacobs 1994, 208-227; Briant 2002, 743-752, 1026f. 19

Briant 2003, 63; compare id. 2010, 178(

Both terms can be used to designate a "receipt" in later language phases (Rabbinic Hebrew: Jastrow 1903, 1529, s.v. 1~i11.i, "receipt, renunciation"; Jewish Babylonian Aramaic: D]BA, 1193, s.v. Kl;JT;I #2, "receipt"- in Sokoloff's opinion, "the concept of 'breaking' as an indication of the invalidation of a document was borrowed from Babylonian culture in which documents were written on clay tablets"). Other terms which have survived in Middle Aramaic from similar practices with knots and strings are 1o/i(, "knot" (Hebrew; lfev 45:25; Hev 46:9) and "1nji, "knot" (Aramaic; see D]A, 78, s.v. "1nji, "receipt" [literally "tie," a calque from Hebrew: with dissimilation of the emphatic consonant o > n]; the noun 1~p, "knot" is common Aramaic;

20

cf. already Dan. 5: 12.16).

workers or soldiers on account of their wages. These 18 tallies were kept in the government offi~e: but each worker or soldier presumably held the other half of the wooden stick. If addmonal pa~m~nts were made, the two halves might have been fitted together, and further notches were InCised on the edge of the stick. The half stick held by the receiver of rations could be ~sed for checking the registration kept by the authorities, and both halves may have been left m the hands of the receiver when the full amount had been disbursed. Th~ugh _there

is little to disagree with in the overall presentation of the Achaemenid Bactrian spin ~alltes, w~ f~el t~at the technical interpretation cited here is not completely satisfactory. A_ review of e.Xl~tlng lttera~ure on the uses of split tally sticks suggests that a credit situation in w1dest sense IS mdeed theu natural habitat. If rations were provided to workers or soldiers "on a~coun~ of their wages," it is, however, difficult to see how that could have constituted a situation of mdeb.te_d_ness. Possibly for the same reason the editors initially seem to have considered another poss:,blltty: that the ~allies were a kind of vouchers stating, by the tally marks on one upper e~~e, th_e to~al quantity of the obligation'' and, by the tally marks on the lower edge, perhaps a parual dJ_sbursement of rations." 21 If we understand this correctly, it would imply t~at workers or soldiers would present their half tally each time they claimed a further allocatiOn f~om the total ration reserved on their account. Yet, in this scenario the institution would tec~mcally be the debtor, which would go against the structure of the statement written on the ~allt~s. ~oreover, one cannot help but wonder why, in the scenario proposed, an Achaemenid mstttutJOn would treat the allocation of wages (in kind or in silver) as a credit situation. In the following we will argue that the payment of rations is not the most likely context for _the Achaemenid Bactrian tallies. We will do so on the basis of a survey of the contexts in whiCh comparable split tally sticks tend to be used (§1), a critical examination of the format and tally marks(§~), and a re-analysis of the inscriptions (§3). Our main goal is not, however, to offer a~ alternative interpretation for the recurring phrase and a different suggestion for the ~ommod1ty - and at any rate our proposals remain tentative as well - but to underline the 1mpo~tance of the existence of these early split tally sticks. Not only do they raise important questions _as to Ac~aem~nid administration in general (§4), bur they may turn out to be crucial elem~nts m the ~1scuss1on about the cultural radiation of the Achaemenid empire and more ~speCially t~at of Its chancelleries (§5). In this sense we hope to contribute to the study of tallies m later penods and to the manifold roles they have played in formal and informal contexts from Achaemenid Bactria well into the twentieth century.

21

YouR TALLY

WF.M. HENKELMAN and M.L. FOLMER

142

Naveh/Shaked 200:., 114 n.6. The scenario offered in this earlier publication is the same as that in ADAB,

~~7r for the phrase his debtor is the person whose name comes after ]1.)" [emphasis added; WH] (ibid P·

r. SPLIT TALLY SncKs:

A

IS

FuLL!

143

SELECTIVE SuRVEY

1.1. Split or unsplit: tally stick formats

Whereas the simple notched stick (Fr. baton de taille, Dutch kerfitok, Swedish karvstock, Germ. kerbholz, kerbstock, Russ. doski, birka) may go back as far as the Upper Palaeolithic, the revolutionary innovation derived from it, and known today as the split tally or split tally stick, is of much more recent date. Just how recent is hard to establish, but it may well be that the Persians were among the earliest users - and perhaps the inventors - of a device that would come to play such an important role in the economic and social life in the late antique, mediaeval and early modern worlds. Split tally sticks are known to have been used, among many other regions and eras, in such diverse contexts as Han China, Viking-Age Scandinavia, early-mod22 ern Switzerland, and high and late mediaeval England. The archaeological trail of the simple notched stick starts as early as 30,000 BCE with the appearance of animal bones incised with sets of marks. The functions of such devices - examples of which are known from Palaeolithic and Mesolithic Europe and Mrica - is predictably disputed, but whether they served social activities (keeping count of hunted animals) or astronomical observation (phases of the moon), it is clear that their essential role was in facilitating counting and organizing. 23 In principle, this is also the role of notched wooden sticks that appear at considerably later date (though this may well be a matter of preservation). A series of attractively-carved short tally sticks from the fourteenth-thirteenth century BCE, 24 excavated at Kitzbiihel (Tyrol), already show a developed numeral system (1, 11, /, \, \/, /\, X). In the Aegean world, a tally stick with simpler marks was found in a twelfth-century layer at

Surveys on rallies (simple and split), few as they are, are usually limited to a particular (modern) region; most are not interdisciplinary and focus either on archaeological finds, on ethnology, or the history of accounting. See esp. the apposite remarks of Kuchenbuch (2003, 469f., 475f.) on the lack of holistic approaches. The British material is the richest in terms of physical evidence and documentation on its (institutional) embeddedness Qenkinson 1911; id. 1925), bur this has, as Hemardinquer aptly remarked, led to a relative disregard for continental material and traditions (Hemardinquer 1964, 940). The studies by Roman Kovalev (see, among others, Kovalev 2003 and 2007) provide an attempt at a synthesis on at least a European level, though their thrust is on mediaeval rallies. Gmiir 1917 focuses largely on Switzerland, Grandell 1982 and id 1989 mostly on Scandinavia, Wyffels 1988 on Flanders. More integrated perspectives, yet focusing largely on Germany and northern Europe, are offered by Kuchenbuch 1999, id. 2003, and Wedell2011, 183-304. See also Brunner 1912; Hemardinquer 1963; Robert 1965; Stone 1975; Ifrah 1981,

22

92-97; Baxter 1994. 23 Much material is illustrated and discussed in Marshak 1991, esp. 21-56, 81-108, 147-68. 24 On rhe Kelchalm near Kirzbiihel (a copper mining site) and its date, see Pichler et al. 2009; on the incised wooden sticks see Wedell 2011, 194-197, with older references. Since the sticks are rather short and have few rally marks on them, it could be that they were a kind of Losholzer, used to divide certain rights, turns in communal labour (etc.) by lot (see Gmiir 1917, 72-76).

}

j

. .

!44

Kition. 25 After these oldest finds, one has to wait until the mediaeval period in northern Europe. Simple tally sticks were found at the Viking sites ofToftanes (Faroe Islands), Elisenhof/ Tonnig and Olversum (Schleswig), dating to the eighth, ninth and early tenth century CE. They are followed by more such devices from tenth and early eleventh century sites such as Ryurik gorodiSce (Kievan Rus'), GroB Raden (Mecklenburg), and Colletiere (lsere, France). Truly large amounts are known from Novgorod (from the mid-tenth century onwards) and Bryggen (the harbour area ofBergen, Norway; 1150-1350 CE). 26 Many more recent examples (eighteenth-twentieth century) are known, often in ethnographic collections; here and there the use of tallies seems to have persisted even today or at any rate until quite recently. 27 Among the impressive array of functions of simply tally sticks we note their use in measuring land, in counting crops harvested or animals hunted, in controlling the volume of a ship's load, and in documenting tax receipts. Telling examples are a recent Albanian tally stick used for counting the number ofloaves baked, an early-twentieth century one used by a Lower Austrian midwife for keeping track of male and female babies she helped deliver, a Swiss mole tally keeping account of the number of animals killed by individual shepherds, or a fourteenth-century tally from Helgeandsholmen (Stockholm) used for counting bundles of hides delivered to a tannery. 28 Developed forms occur at an early date (Novgorod, etc.). One of such more complicated formats is that of a single tally stick rectangular in section, with notches on all four edges, and suitable for keeping more than one 'account.' An example is a thirteenth century cargo tally found aboard a Hanseatic cog that perished off Kollerup in the Jutish Jammerbugt. A 1627 tax tally from Halsingland (Sweden) has the shape of a thin sword; separate entries with personal marks identifY individual tax payments. 29 25 A recent find from Tiryns, from around 1200 BCE and inscribed in Ugaritic cuneiform, has also been identified as tally stick, though this case is clearly less convincing (compare Gmiir 1917, 67 on measuring rods and related objects). See on both finds Cohen/Maran/Vetters 2010, esp. 13-17. 26

There is no extensive treatment of all the early notched sticks, despite their evident importance in mediaeval European society. Kovalev (in Noonan/Kovalev 2000, 132 n.41) states that hundreds of mediaeval European rally sticks (mostly unsplit) exist, yet mostly unpublished. For references to the notched sticks mentioned here as well as other finds, see Grandelll982, 41-43; id. 1989, 5-14; Muller 1996, 173f (without distinction between split and simple tallies); Mille/Colardelle/Verdel 1993, 247 (Colletiere); Kovalev 2003; id. 2007; Wedell2011, 197-212; Stumman Hansen 2013, 99, 127 (Toftanes and other Faroe sires). 27

In the preface to the 1991 reprint of Gmiir 1917, the author's son, Rudolf Gmiir, mentions the continued use of wooden rallies in the Ginanzral and the Lotschental in Valais, Switzerland, at the time of writing (cf Gmiir 1917, 63). Hemardinquer (1964, 941) states the same for some rural areas of France in the 1960s.

28

Loaves: Baxter 1994, 199; midwife: Haibiick 1964, 129, pl. 2; moles: Stebler 1907, 180f, Gmiir 1917, 111; tannery: Grandell 1989, 55-62. Kovalev 2002 discusses two remarkable tallies (16'h-17'h cent.) from Canada, both connected with checking cargo. 29

YouR TALLY rs FuLL!

WF.M. HENKELMAN and M.L. FoLMER

Kollerup: Ellmers 1995, 214-215; cf Wedell2011, 202f Cargo rallies may have been used in pairs during the loading or offioading of commodities: one kept by the merchant entrusting or receiving cargo, one by a mmber of the ship's crew; both counted the items going in or out of the vessel and at the end compared the totals on their tallies (so Grandelll988, 66f). At Colletiere (Isere, France) a number of rectangular sticks with notches on more than edge was excavated; they date between 1003 and 1034 CE (see Mille, Colardelle/ Verdel 1993, 247). There are also early pentagonal rallies for keeping five parallel accounts (Kovalev 2007, 190f on Novgorod), whereas more recent examples, the eye-catching milk rallies or Stialas from Graubiinden

I45

Though the simply tally stick's potential for complex accounting cannot be denied,_the ibilities to authenticate or identifY, or function as a legally enforceable record remamed ~::red. The unsplit notched tally, for all its flexibility and all the good it did for human life, in most cases essentially remained what it had been from the start: an aide-memoire. Before we continue to describe the split tally stick as a fundamentally different device, a few words about tally marks are in place. These are variously known as notch;~ (Eng.), encoches (Fr.), Kerben or Krapfen (Germ.), kerven, krappen or schrappen (D~tch), etc. Fo_rmed by cutting into the wood with a knife or sharp object, they were often hrghly stan~ardrzed. Notches could be varied in shape (superficial vs. deep; straight vs. oblique) and combmed to form special symbols for, e.g., decades in a decimal system and thus theoretically lie at ~e base of more complicated numeral systems. This may be the case with Roman numerals, whrch seem to continue, in developed and abbreviated form, Etruscan (and probably even older) tally marksY In any case, elaborate notation systems occur from at least the thirteenth ~entury BCE (cf. above on Kitzbiihel) and are again found on some, but certainly not all, medraeval examples. The split tally is also incised with notches, but cut lengthwise into two halves C?~· _6). Though highly developed forms exist, the split will always run through the notches, drvrdm_g them equally. Alterations can only be made if both halves of the tally are put together; fraud rs practically excluded as one-side alterations will be noticed upon joining the ~o halves. Often, the names of the parties involved will be written on one (usually the debtors)_ or_ both halve~, sometimes other information is given as well, but none of this is essential. In pnncrple, the spht rally marks are enough to guarantee any agreement (deposit, credit, etc.) between two parties, also if they are illiterate and/ or speak different languages. Split tallies were, moreove~, tr~nsfe~a­ ble as the notches would do the work by themselves. A supporting role in guaranteemg rdenuty could be played by the grain in the wood. Though tally sticks were normally split with a knife (cf figs. 16a-b below), may have as many as eight carved surfaces (Stebler 1907, 183-187; Brunner 1912, 342; Gmiir 1917, 92-95, pl. 20; compare Grandell 1989, 67-78). Halsingland tax tal~y: Grandell 1982, 65-68, id. 1989, 12f Hemardinquer (1969) describes the use of tax tallies in France, which may go back to rhe 13th cent. or earlier; a text he cites (from 1651) speaks of a rally of over 1,60 m (ibid. 5llf; compare Stebler 1907, 195). Such single tax tallies were notched in_ the presence of t_he tax p:yer a~d t~e collector. A related phenomenon is that of the Kehrtesseln, simple tallies used for rotatmg corvee obligations and community services in Switzerland (Stebler 1907, 175-180; Gmiir 1917, 80-85). Comparable tallies are known from Kievan Rus' (Kovalev 2003, 168); many different kinds are listed by Grandell1982, 117-130. 30 See Beets 1926, 2050-2055; Wyffels 1988, 26; Kuchenbuch 1999, 319; Wedell2011, 189-193. 31 Though there are various hypotheses on the origin of Roman numerical notation, the rally mark t~eory (which goes back as far as the 16th cent.) has the advantage of being the most eleg~nt an~ sysremanc, as Keyser has shown in a rich and well-documented paper (Keyser 1988). See also Chnsomal1s 2010 , 93-98. Ifrah 1981, 139-59 traces rhe origins of this notation back to a pre-Etruscan system of,rall~ marks used among Italic and adjacent peoples. Among the parallels he lists are Dalmatian shepherds rallies that were still in use in rhe 20th cent. and rhat operate a system of marks similar to the oldest known Etruscan numerical notation. Many post-Roman rally mark systems show affinity with Roman numerals, often in ways that suggest a common ancestor rather than direct dependence. See, e.g., Gmiir 1917, 52-56 on Swiss~ol t~ly marks, Kuchenbuch 2003, 478-480 on Austrian (Styrian} ones and Grandell 1989, 79-89 on Scandmav1an examples. See also Grandell 1982, 37f

146

or a ~aw, they s~metimes were only partly cut, after which the tally halves were broken apart to provide for an Irregular and unique break surface between them.32 It is important to distinguish not only between the simple and split tally stick, but also ~etween the split tally stick and the split token. The latter, in English also referred to as tally, IS what the Greeks called a symbolon. 33 It could be any object (wood, stone, ceramic, a sheet of papyrus) broken, torn or split into two parts and used for identification and authorization between two parties who could match their respective halves whenever necessary. 34 One such context was that of oral or written messages, which could be endorsed by one half of a split token; this would be sent along with the message to a recipient holding the other half. In nineteenth-century Russia, Cossacks on border duty still used this system (Kovalev 2003, 170f). Split tokens could also serve in more intricate systems of authentication and state control and acquire highly developed formats, as they did in Warring States-period China. Nevertheless, even such developed tokens remain fundamentally different from the notched, split tally stick, as the latter not only identified and authorized the parties involved, but also endorsed the terms of their agreement (credit, deposit, etc.) by means of the split tally marks.35 One could therefore say, with Baxter, that the split tally stick combines the advantages of two ancient de32

Breaking: Haibiick 1964, 127; wood grain: Grandell 1977, 103 (comparing it to watermarks in paper money), id. 1982, 28. See also Lovett 1897b, 161. 33

Throughout this article, we reserve the word tally for notched tally sticks, split and unsplit.

34

The story of Glaucus son of Epicycles (Hdt. VI.86) turns around the use of the symbolon for verification purposes. Baxter 1994, 200 argues that a plain symbolon would not be of much use and that the object used may have been a split rally stick or an object with the same function (his category 3), bur this is not evident. It is .int.eresting, though, that Herodotus mentions the use of a symbolon as a voucher for silver deposited by a Mdesian at Sparta in a story reminiscent of that of Gylippus the Spartan (c£ §5.2). Stone 1975, 50 erroneously ascribes the use of the symbolon by early bankers to Babylonia (the reference he cites pertains to Greece). 35

YouR TALLY rs FuLL!

WF.M. HENKELMAN and M.L. FoLMER

The various terms for (split) tokens in early China (jie, fu and their derivatives) are usually translated as in Eng!ish as '(~plit) tally,' which may account for their occasional inclusion in debates about the origins of the split tally stick (see, e.g., Grandell 1989, 2-4; Goetzmann/Williams 2005, 114£). Another reason for the confusion may be the use of actual notched, split bamboo sticks observed by Marco Polo in late 13'h-cent. CE '~ardan~an' (in Yunnan province, southwestern China), on which see §5.7 below. For the split tokens lfu) m Warnng States and early imperial China, see the excellent overview by Von Falkenhausen (2005). Famous among such tokens are the split bronze tiger figurines (hufu) that authenticated communications ~rom t~e ruler to a local commander (one half accompanied the oral or written order; the recipient matched it to his own). Some hufu had inlaid gold or silver inscriptions stipulating the terms of their use. The so-called E Jin Qi tallies, inscribed bronze replicas of bamboo sticks divided in five parts, are described by Vo~ Falkenhausen as "primarily tokens of privilege" (2005, 92) that exempted merchants in the centrally-organized trade fro~ the ~egular tariffs; they probably had to be joined as a complete set at the end of trading season. Though snpulatmg the date of first issue (323 BCE), validity and the tonnage allowed, they are not contracts in the strict sense. Also, the Chinese split tokens of both types are precious objects and they appear ~o. ~ave been re-us~d repeatedly- as one can with tokens, but not commonly with split rally sticks, once the mmal agreement IS terminated. Interestingly, the ritual compendium Zhou li (2"d cent. BCE) asserts that "All .those who circulate and c.ommunicate, within the Realm under Heaven must have a tally, which is provided through a [bureaucratic] procedure (cited and discussed by Von Falkenhausen 2005, 83f.). In that sense, the tokens are more akin to the ubiquitous and probably obligatorily used (private yet somehow authenticated) seals in the Achaemenid royal road system.

147

vices: the memory function of the simple tally or notched stick and the unique identification provided by the symbolon.

36

There are various formats of the split tally, but all adhere to same principle of proof and counterproof and are therefore essentially different from the simp~e, unsplit ta~,ly. :'on Kostanecki might be carried away with the subject when he defines the spht tally as the Qum~­ essenz aller Kerbholzphanomene," but he is right in drawing a sharp distinction between t~Is and simple tallies.37 We would like to add that, whereas the origins of the simple tally sock are lost in the depths of time, the ingenious device that is the split tally has every likelihood of being an invention made at a particular moment at a particular place, from where it spread as far as Iceland and China. The most basic and probably original form of the split tally involves two parts of equal shape, usually the two halves resulting from splitting a wooden stick equally and o.ve~ its entire length. The stick may still have its bark (although it is normally removed) and retam Its ~atur~ shape; in this case it is roughly circular in section. Some of the Novgorod tallies are of this baste format, though (as usual) only one of the two halves is preserved. AB in the case of the. Achaemenid Bactrian tallies, the format is recognizable by 1) the fact that the notches are cut m to the 38 edges and 2) the two different faces, one flat (the split surface), one convex. Split tally sticks may also be cut out of blocks of wood, adhere to a more regular format and be rectangular in section. Also, the way the two halves interlock may be a great deal more intricate, as in the case of two interlocking 'tuning forks' (fig. 5). A tally split this way has the advantage ofhavi~g four sides in which notches may be carved (in each case incising both halves). Grandell discu~ses a 1922 example used between a skipper delivering wood on credit to a representative of Abo Akademi in Turku (Finland). 39 One advantage of the fork tally is that both halves can be stuck tightly together for matching and adding additional notches. This may also be the background of yet another format, first and foremost known from the mediaeval Exchequer tallies (fig. 7), but also from closely-related private English tallies and continental parallels (cf. §§ 1.3, 5.8). All these were made by striking a deep diagonal cut halfway into a regularly-shaped stick, at about one-fourth Cf. Baxter 1994, 197. Robert (1956, 76) defines the split rally as "an acknowledgement made in two ~ar~~· and in such away that, when re-joined, it constituted proof of the parties' responsibility in the transacnon.

36

von Kostanecki 1900, 136. Though the prose of this and other passages in the same work is at times quite impenetrable, Gmiir's disdainful remarks (1917, 60f. and elsewhere) are not justified, if only because he himself never seemed to have realized the sharp distinction between simple and split tally. 38 Wendell, who overlooks the importance of the convex/flat surfaces, is unjustly sceptical as to the identi-

37

fication of the objects as split tallies (2011, 272£, 276). 39 Grandell 1982, 95; id. 1989, 17-8; cf. Baxrer 1994, 203. Grandell (1989, 91-99; cf. 1982, 30-32, 95) also discusses a context in which the four sides of fork-tallies were used to record credit and debit in both daler and ore. Schempf 1989, 21 illustrates another ingenious design, from 19th-cent. Styria, involving two equal tally halves that could be connected with dovetail joints. Note that tally cutting was sometimes se~n as a serious craft, not only in the case of the Exchequer's Tally Cutter, but also, e.g., in the city of Ghent m the 15'h cent. (Wyffels 1988, 26).

\

YouR TALLY

WF.M. HENKELMAN and M .L. FoLMER

IS

149

FuLu

Fig. 8: Complex 'split' tally stick: Alpscheit with slots for Einlagetessel, Tellialp, Blatten (Switzerland), 1695 CE; after Gmiir 1917, pi. 27.

Fig. 5: A fork tally as used in Turku, Finland, early 20'h century CE; after Grandelll989, 17.

Fig. 9: Split tally sticks (tailles), Vaud!Waadt (Switzerland), early 20rh cent. CE; after Gmiir 1917, pi. 23.4.

Fig. 6: Tally stick, not entirely split, Troitsky (Kievan Rus'), later lOth or early 11 rh cent. CE; after Kovalev 2007, 198 fig. 128.

E:±±-4111HIIIJIIImlllllllmiWUBHJIIIIlBmiF - --1 ; =J

,_ -~--:~ ::_-----.,_ ===---_ ::_- -·. ~~;;..--__~ ---==-=--=~ -- --=--se~ o-..::::::.: L--:--------

Fig. 7: Split tallies from the English Exchequer, 13th century CE; after Jenkinson 1911, pi. 49.1.

~- : ·

.

.

Fig. 10: Two split tally sticks (hop tallies), Kent (England), late 19th cent. CE; after Lovett 1897a, 38 figs. 1-la.

Fig. 12: An Alpvogis bundle of Krapfentesseln, Tellialp, Wiler (Switzerland), 19'h-20rh cent. CE; after Gmiir 1917, pi. 31.3.

Fig. 11: A split tally board (Krapfentessel with smaller Einlagetesse[) indicating 43,4 Kuhrechte, the tally's number and the Hauszeichen; Gletscheralp, Lotschental (Switzerland), 19'h cent. CE; after Stebler 1907, 205 fig. 46.

~.

.

WF.M. HENKELMAN and M.L. FoLMER

or one-third distance from the base (left end). The stick is then split from the incision to the 40 top (right). The result is a shorter piece of wood fitting rather tightly into a longer one, which has a socket or stump at the end. 41 Apart from the benefit of easily and exactly matching both halves, such asymmetrical tallies also had the advantage of having two recognizably different parts, which could be assigned different functions (held by, e.g., creditor and debtor). Asymmetric split tallies can also take other shapes and are well attested in other cultures. 42 In English, the larger half of an asymmetric split tally is known as stock, the smaller as foil or leaf43 Other languages put less stress on the difference in shape between the two parts of the split tally (Swiss German Doppeltessel), and more on the difference in function: talea or tallia vs. contratalea (Med. Lat.), stoc vs. wederstoc (M.Dutch), kerf vs. tegenkerf or conterkerf (Flem. Dutch), taille or souche vs. contretaille or echantillon (Fr.), Stock vs. Gegenstock or Einsatz (Germ.), Krapfen vs. Einlegetessel or Beitessel (Swiss Germ.). An interesting outlier is the pair Manndlvs . Weibl (Austr. Germ.). 44 Split tallies were crucial in various institutional contexts in England from mediaeval times. Their ubiquitous use at the Exchequer led to the Bank of England's custom of handing private investors a stock in acknowledgement for the money they had invested (the foil was kept at the bank). Such investments thus became known as bank stock, whence, it is generally assumed, the modern economic usage of the term stock and its derivatives, such as stockholder and stock exchange, which was first used for the London Stock Exchange (1773), followed by that of New York (1792). 4 5 40

In a 1670 description from Exchequer officials, quoted by Jenkinson 1913, 34, the tallies are prepared by the Tally Cutter from blocks of hazel, and then handed to the Deputy Chamberlains of the Receipt. The senior of these "holdeth the Stick in his left hand upon a Block laid upon a Table and a Cleaver in his right hand upon the Stick. And the Junio' Chamberlaine with a Mallet in his right hand first Cutteth the Stick halfe through upon one of the written sides Close to the writing and afterwards Cleaveth the Stick long waies (... )" [emphasis original]. Similarly Madox 1769, vol. 2, 258. 41

In some types of Austrian split tallies, the two halves could be held together by wooden pins (Haibock 1964, 128, pis. 1-2).

42

One interesting developed form, the triple tally, has been postulated by Grandell as part of Swedish ironworks bookkeeping: one of the three parts of such a tally would be held by a charcoal-burner, a carrier and the recipient at the smithy respectively (1982, 29f., 94; id. 1989, 14f.). Schempf 1989, 21 cites evidence for their use among snow sweepers in Vienna; Brunner 1912, 340 fig.5 illustrates an example from that very city (cf. ibid 345). Gmiir describes various complex forms, including the Alpscheit (stock) with multiple Einlagetessel, small cubes with the function offoils (Gmiir 1917, 121-123, pis. 27-28; here fig. 8). One of the objects he introduces and illustrates is a carefully-carved, 130 cm long stick weighing 3.3 kg and having 70 small cubes fitting into it, each with notches continuing onto the stock. Stebler (1907, 201f.) mentions another example, almost 2 m long, and with 165 Einlagetessel, from the year 1784. Such objects were used for recording grazing rights (Kuhrechte) and involved an entire community. 43

So already in the 1670 document quoted above Oenkinson 1913, 34) and n.40 above.

44

See Jenkinson 1911, 374; id 1925, 291f.; Brunner 1912, 343; Gmiir 1917, 127; Beets 1926, 2250; Wyffels 1988, 22; Schempf 1989, 22; Wedell2011, 185f. 45

Philadelphia had the oldest stock exchange in the United States (1790), but it was not named so until 1875. On the purported derivation of current terminology from the use of split tallies, see Jenkinson 1911, 368f.; id 1924, 292; Robert 1956, 81f.; Stone 1975, 54-57; Baxter 1994, 215.

YouR TALLY rs FuLL!

1.2. Come, Mr Tally Mon, tally me banana

As we will argue in this section, the use of split tallies seems largely confined to credit situations or, as Haibock more misanthropically puts it, everywhere where a certain degree of mistrust was appropriate (Haibock 1964, 128). Kovalev even uses the term 'credit tally' as equivalent for 'split tally stick' (Kovalev 2005, 871). We believe that this should also be the preferred interpretative frame for the Achaemenid Bactrian documents of the same format. In fact, had the Bactrian tallies been discovered some 150 or 200 years earlier, no-one would have doubted that that their function had to be sought in this direction. In the early nineteenth century, tallies - split and unsplit - were still very much part of everyday life in many parts of the world. Modern languages abound in tally-related expressions that bear witness to this former importance. Some of these expressions refer- metaphorically- to the counting (aide memoire) function of the (simple) tally, such as "einiges auf dem Kerbholz haben" (to have a bad record) or "mijn levenskerfstok geeft u dra zes kruisen met vijf krappen te lezen" (my lifetime tally will soon show six crosses and five notches). 46 In most cases, however, credit, and more often lack thereof, is referred to. Modern Persian "cubkhatet por sode" (your tally is full!) speaks for itself, as does Swiss-German "'s got Alles bl-n-im uf's Cherb-Holz" (with him everything is on credit). Compare also the examples cited in the encyclopaedic Woordenboek der Nederlandsche Taal s.v. kerfitok, such as "de kerfstok wordt van ijzer" (the tally has turned to iron, no more credit is given) or "iemand den kerfstok schenken" (pass one's debt's to someone else). Dutch "iets op eigen houtje doen" (lit. to act on one's own piece of wood) probably also belongs here and seems to suggest making notches on only one half of the split tally (i.e. acting on one's own account, without consultation). 47 Just how ubiquitous (split) tallies once were is also underlined by the fact that they were recognized as legal records in many local and state laws as well as in legal practice in European history from mediaeval times. They could, for instance, be adduced as proof against a debtor who had failed to meet the terms of a credit arrangement and their status as equivalent to merchants' books was explicitly acknowledged in seventeenth-century Holland. 48 Napoleon's 1804 Code Civil still upheld the tally's ancient role as legal proof for credit transactions, which led to its incorporation in several European law codes and its survival in legal terminology well into the twentieth century and in some cases until today. 49 46

The expression is slightly adapted, for the festive occasion, from Beets 1926, 2255.

47

Persian: M. Rouhani (pers. comm.); Swiss German: Szadrowsky 1935, 233; Dutch: Beets 1926, 22502255, Stoet 1923-1925, 373f. s.v. Op zijn eigen houtje. 48 Holland: Beets 1926, 2251. For their legal status in the Low Countries at large in earlier centuries see the evidence collected by Wyffels 1988, esp. 37f. For Germany see Kuchenbuch 1999, 308 with n.4 and Wedell2011, 289-292; for Switzerland see Gmiir 1917, 141-150, 153f. Some English contexts are cited by Baxter 1994. "Les tailles correlatives a leurs echantillons font foi entre les personnes qui sont dans !'usage de constater ainsi les fournitures qu'elles font et rec;:oivent en detail" (§1333, Code Civil des Fran(ais, Paris 1804, p. 322). 49

YouR TALLY rs FuLL!

WF.M. HENKELMAN and M.L. FOLMER

Given their widespread use, it is not surprising that the academic study of split tallies dates as far back as the dissertations of Johannes Jacobus Fuchs (De Chartis et bacillis incisis, von Kerb-Zeddeln und HOlzern, Strasbourg 1668) and Joachim Christoph Stisser (De bacillis jissis vulgo Kerb-Stocken, Frankfurt a.d. Oder 1676). The second treatise stresses the division of the two halves of a tally between a creditor and a debtor as an essential characteristic of the phenomenon (Stisser o.c. p. 16; also Fuchs o.c. p. 1lf.). The real impetus for understanding tallies in their societal context came from late-nineteenth and twentieth-century ethnologists who gave first-hand descriptions of the material as it was still in active use. Among these, the names ofMax Gmiir (1871-1923) and Axel Grandell (1899-1993) deserve special mention. The first, professor of Law at Bern University, followed steep paths to remote alpine valleys to collect tallies and related objects and gathered invaluable data about their use from informants and from his own observation. The result of these efforts was his charming Schweizerische Bauernmarken und Holzurkunden (Gmiir 1917). Our second tallyman, Axel Grandell, was born in the Swedish-Finnish community of Nagu (now Pargas, Finland). He had similarly been exposed to the living use of the tally stick, which probably lasted longer in Scandinavia than in any other part of Europe. In the two books he devoted to it (as professor of Economics at Abo Akademi University), he demonstrates his sensitivity to the frailty of a waning world and champions the tally's role as "en av mansklighetens aldsta kulturbarare" (one of mankind's oldest culture bearers). 50 Gmiir noted from personal observation that the split tallies had, in comparison with the unsplit ones, a much stronger emphasis on control and proof and were, therefore, more often used between people of different background or different status, such as customer and butcher, or lessee and landholder. A telling example is that of the grape harvest in Vaud/Waadt, which Gmiir knew through various informants, whose reports he summarizes as follows: 51 Der Gutsherr sender zur Weinlese als seinen Vertreter den sogenannten pardisseur (in Ligerz Teiler genannt), gewohnlich ein Lehrer aus der Umgebung. Dieser macht fur jeden vigneron eine taille, wobei er die Uinge einigermassen nach dem zu erwartenden Ertrag und der Dauer der Lese bemisst. Die vignerons arbeiten nicht im Taglohn, erhalten jedoch freie Wohnung, Pflanzland und 5 Rp. pro Liter Wein. Fur jeden vigneron ist ein Pressoir da; jedesmal, wenn 300 Liter aus dem Pressoir in die Butte gelaufen sind, macht der pardisseur mit der Sage oder dem Messer eine Kerbe uber den zusammengelegten Halften der taille, von welchen er gewohnlich das langere Stuck in Handen behalt, wahrend das andere, mit dem Namen des Bauern versehen, an diesen zuruckgeht. Die Tage der Lese werden durch Striche auf den Seiten markiert. Ist die Lese vorbei, so werden die Zahlen der Holzer auf Papier abgeschrieben und von beiden Teilen unterzeichnet; hierauf sind die Holzer wertlos geworden und man verbrennt sie gelegentlich. The paragraph, which actually relates to a specific use of the split tally, subsequently found its way into the Civil Codes of Belgium, Austria, Italy, rhe Netherlands and parts of francophone Switzerland (Brunner 1912, 347; Gmur 1917, 144£; Arnould 1966, 98; Wyffels 1988, 22, 37). 50 Grandell 1982 (monograph) and 1989 (collected papers); the citation is from Grandell 1982, 146. 51 Gmur 1917, 106£; see also Wedell2011, 222. Schempf 1989, 20 illustrates a l8'h-cent. Austrian print, on which a tallyman is shown cutting a notch upon the presentation of a heavy load of grapes. See also Kuchenbuch 2003, 478£, commenting on a comparable Austrian (Styrian) case.

I

53

A few elements are particularly striking in this description. One is that the owner or his representative may hold several tally halves, each worker has only one. Secondly, the notches do not directly represent the salary eventually to be paid, but units of the harvested commodity. Thirdly, the number of days worked is also recorded on the tally in a separate series of incisio~s: notches on one and the same tally can therefore express various things. Fourthly, the tallies document and guarantee a credit relation (salary to be paid after the harvest); once it is terminated, the tallies are destroyed. Gmiir also notes that the tallies are cut in anticipation of the space needed to record the harvest, but this point should not be overemphasized. He himself illustrates three tallies from the same estate with a varying number of notches (12, 19, 16); the tallies are almost exactly of the same size and each case the notches occupy less than half of the available space (Gmiir 1917, pl. 23.4; here fig. 9). The example discussed here concerns postponed payment ofsalary. This seems to have been a common context for the use of split tallies, be it at Jamaica's song-inspiring banana plantations or at hop farms in Kent and Worcestershire, where tallies were in regular use until the 1940s (fig. 10). In the latter case, the harvester and the tallyman would match their respective half tallies each time the hop-picker had a full bin to present; the tallyman would add the appropriate number of notches for the bushels of hops turned in. A notch could represent various amounts, such as one for each five bushels; the conversion to money, at the end of the season, 2 depended on the quality and the market value of the hop. 5 A second regular context was postponed payment for deliveries. We have already cited the example of the skipper bringing wood to Abo Akademi in Finland and using a fork-tally as a record of the loads for which payment was outstanding. That example is from 1922, but much older cases are known as well. These range from the tallies of a blacksmith delivering nails to Egmond Abbey in Holland around 1390, via a tally for wax deliveries to the church of Haguenau (Alsace) in 1595, to the bundle of tallies carried on a metal keyring by eighteenthand nineteenth-century French baker boys distributing bread to a series of clients, each with their own counter tally (foil). In the Bellinzona region of italophone Switzerland even dung deliveries were recorded on split tallies. 53 In a variation of the second scenario -postponed payment at acquisition - the smaller half of a split tally (foil) was presented by a customer at, e.g., the baker's or at the butcher's. For each See Lovett 1897a (the most important report); Jenkinson 1911, 368, pi. 48.2-4; Perch 2009. This use of the split rally is already attested in the 14th-cent. Low Countries (Wyffels 1988, 32, who implausibly interprets the relevant passage as referring to a simple tally). 53 Egmond: Dijkman 2011, 260; compare Beets 1926, 2254. Haguenau: Arnould 1966, 99 (also noting other uses, some persisting at the time he wrote). French baker boys: Hemardinquer 1963, 143; Ifrah 1981, 93; Vernus-Moutin 1991,73 n.40; Baxter 1994,228. Bellinzona: Gmur 1917, 105. Wyffels 1988,33 notes the use of rallies, in the Low Countries, for soap, beer, wine, bread, barrels, clover and grass, oil, lamb and sheep skins, nails, stones, candles, (broad)cloth and, unavoidably, herring. The w~rd tallyman ~eant, in th,e second half of the 20th cent. in England and Scotland, a pedlar selling goods on hire-purchase (never-never) basis; it may derive from the use of tallies to register the customer's debt (so Stone 1975, 56f.). Compare Scheuermeier 1943, 12 n.l on such vendors in Tuscany, still using split tally sticks as late as 1924.

52

154

YouR TALLY

WF.M. HENKELMAN and M.L. FoLMER

unit of the commodities he or she received, a vendor would add a notch to both tally halves held together. The sum had to be paid at certain intervals. 54 Such postponed payment was also practised in other contexts. Wyffels (1988 bijlage C) illustrates a tally stock once held by a tavern keeper in Antwerp; it notes a debt of 124 pots of wine against one Selvest van Wad, who had had his first round on October 4th, 1582. But split tallies could also record much more complicated accounts than those of Flemish barflies. A very sophisticated system of split tallies was reportedly in use on Laos markets in the nineteenth century. 55 A third and special sub-category of the postponed-payment tallies were those issued as IOUs by royal purveyors for commodities such as grain and forage requisitioned by the English court. In such institutional postponed payments, the crown officials acted as debtor and held the foil, whereas the creditor had the stock which vouched for the repayment(s) due to him. 56 Paying taxes too, could involve the use of a split tally, either in the assessment of tax obligations or because advance tax payments had been made. Haibock describes an elaborate ritual performed in Zwettl (Lower Austria) to establish the tithe due to the landholder (before 1848). In it, the baskets of corn were presented in a long row and counted by alternatingly sticking a blue or red Zehentsteck in the ground after each ten containers and adding a notch on both halves of the split tally. 57 Another example is that of Rumanian Wallachia, where, until the sixteenth century, tax payers were handed one half of a split tally in anticipation of the final settlement of their account (cf the Exchequer tallies described in §1.3). 58 And though Marc Bloch at one point expressed doubt about the traditional etymology connecting taille in the sense of "taxation" with taille, "(split) tally," Jean-Jacques Hemardinquer and others subsequently elaborately documented the use of the device in French taxation since at least the fifteenth century. 59 54

Brunner 1912, 339 notes that the use of (split) tallies had the additional advantage that children could be sent on an errand, without having to hand them money for it.

55 See Ifrah 1981, 94. Brunner 1912, 343f. cites the example of a Bavarian farrier, who carved notches on a split tally for the number of hooves he shoed. Kuchenbuch 1999, 304-306 elaborately describes his Alsacian colleague, who operated split tallies with separate accounts for new and re-used horseshoes. Lovett (1897b), finally, offers a fine report on split tallies used in baker's shops in Brittany. 56 See Jenkinson 1925, 310-315, with a list of such tallies and transcriptions of their marks and inscriptions on pp. 329-51 (compare also ibid. 304-306). Commodities mentioned are grain, malt, peas, wine, various minerals, wood, sheep, and 'bacon' (pigs). The survival of the private tallies (foils) as such is interesting because it either implies that the administration kept them as records even after the creditors had been satisfied, or that some creditors never redeemed their stocks, perhaps for reasons of presrige.

IS

FuLL!

155

Temporary assignment (with usufruct) or deposit of commodities or animals was a fourth rype of transaction that could be regulated by tally. In certain parts of Switzerland, for ex~m­ ple, assignment of grazing rights on communally owned summer pasture was done by a kind of split tally known as Krapfentessel (fig. 11). Units of pasture land were measured out by the (theoretical) size of the herd fed on it and were known as Kuhrechte. A refined notation system was used to record these on the split tallies, the stocks of which (sometimes as many as 200) were held by the Alpvogt (fig. 12). Individual holders of grazing rights would annually gather and match their foils to the respective stocks in a public ceremony shortly before or on the day the animals were driven to their summer pasture. 60 Tallies could, finally, record proper loans, mostly in coined silver or bullion, but also in seed or other commodities. In Novgorod and other trading emporia in Kievan Rus' (cf. §5.8), merchants regularly used split tallies for this purpose, i.e. in the context of long-distance trade; they would normally enter the loans in their books as well, but the tally served as the prime record between the two parties. Also in thirteenth- and fourteenth-century England money-lenders regularly used split tallies for loans and so did some of their continental colleagues. The twelfth-century Flemish investor William Cade sometimes issued tallies together with the silver he lent out; Henry 11 was among his clients. The English Crown itself regularly gave tallies for the credit in silver it took from private parties; such tallies were issued at the Exchequer, to which we shall presently turn. 61

1.3. The English Exchequer An exceptional source on the cutting, format, handling and legal status of split tallies is the Dialogus de scaccario (Dialogue on the Exchequer) written by royal treasurer Richard fitzNigel (fitzNeal) during the decade or so following 1176. 62 It describes the workings of the crown's financial department, the Exchequer, in which the talea or tallia was a central element, so much that Richard reminisced that its old name had been The Tallies (Dialogus 1.1; cf. the estimation ofJenkinson 1925, 290). The English crown received a greater part of its revenues through sheriffs, tax farmers who collected various taxes from their counties. For this right, they paid a fixed rent ('farm') to the

57

nus-Moutin subsequently found new evidence from the southern Dauphine in high mediaeval rimes (1991) and showed that of 131 communities in the 1474-76 census, 44 used tally sricks to declare their taxable income. Note that not all cases cited by Hemardinquer and Vernus-Moutin concern split tally sticks.

58

60

See Haibock 1964, 127-129. Haibock cites his grandfather, who recalled the great bonfire of tallies, after the abolishment of the land tenure system in 1848 had rurned them redundant.

See Hemardinquer 1963, 146, ciring a Rumanian publication by D. Mioc. See also Wyffels 1988, 34f. on the use of split tallies in tax collecting in the Low Countries, Hemardinquer 1964, 941 on related cases from 14'h and 15th cent. Spain, and Noonan/Kovalev 2000, 133 and Kovalev 2003, 170 on Kievan Rus'. 59

Hemardinquer first presented his finds in an Annales article of 1963, intended as a contribution to the discussion provoked by Bloch (supplement: Hemardinquer 1964). Much new material was included in his 1969 survey, which settled the question of the etymology in favour of the tally explanation. lsabelle Ver-

See Stebler 1907, 208f. and Gmiir 1917, 126-131. The use of split tallies in this context goes back to at least the 17'h cent. (cf. Kuchenbuch 2003, 488, with reference). 61 English money-lenders: Jenkinson 1925, 313; Cade: ibid. 304, with references; tallies of receipt issued by the crown: id. 1911, 370f.; Stone 1975, 54f. 62 See Amt in Amt/Church 2007, xiv-xx on the date of the work and the identification of Richard as its author.

157

WF.M. HENKELMAN and M.L. FoLMER

YouR TALLY IS FuLL!

crown, the rest was their profit. 63 The sheriffs would normally deposit (proffer) an advance in silver coin to their accounts at the (lower) Exchequer directly after Easter. Rents paid in silver were subject to material assessment: for some weighing the money was considered adequate, but county farms (rents on estates) had, at the time of the Dialogus, to be 'blanched' in an 'assay.' This meant that a sample of the silver paid was melted, purified, and then weighed against the Exchequer's standard pound. The initial deposit would then be discounted by the degree of impurity; the sheriff would (eventually) have to supplement the difference to meet the requirements ofhis 'white rent' (jirma blanca, albajirma). 64

The procedures described above required a standard set of notches indicating the deposited amounts on the split tallies. At this point, Richard is at his best (Dialogus 1.5, transl. Amt in Amt/Church 2007, 35):

The actual audit took place at a later date in the year, directly after Michaelmas, when the Exchequer would be in session again. In the settling of the accounts, such matters as credit carried over from previous years, lands granted away (hence no longer subject to direct taxation), stipends for the poor or incidental expenditures by the Sheriff on behalf of the crown wouJd be considered. Such variables, and the fact that the sheriff regularly paid only half of the fixed rents at Easter, clearly defines the initial payments essentially as provisional deposits for which the sheriff needed proof. This was provided in the form of a split tally- usually made of hazelthat stated (in writing) the term and the regnal year and (in notches on both edges) the amount paid in advance (fig. 7). Initially, a separate tally was issued for every single (part of the) deposit made by the sheriff. The Exchequer's tally cutter handed the tally stock to the sheriff, whereas the foil was deposited in the treasury. Then, at Michaelmas, the sheriff returned the stock for the audit, at which it was joined with the foil for control. 65 The last stage was when, after the record had been settled, the marshal would deposit the stock in the treasury with the fore! (bag) containing the silver paid by that sheriff. 66 63

For the working of the Exchequer see, e.g., M adox 1769; Jenkinson 1911; id. 1925; Robert 1956; Baxter 1994; Arnt in Amt/Church 2007, xx-xxvii; }ones 2008 (all with further references). For a brief synopsis on the tallies see Chatfield 1996. Noonan/Kovalev 2000, 131 n.38 note that the Norse kings of the 14th cent. similarly used split tallies in their system of tax collecting.

The tallies are cut as follows. An amount of£ 1,000 is shown on the top side with an incision the width of one's palm, £ 100 with the width of one's thumb, £ 20 with the width of the little finger,£ 1 with the width of a swelling grain of barley, a shilling even smaller, bur so that a little notch is made by the two knife-cuts and some wood is cut our of the space. A penny is indicated by a cut that removes no wood. On that part of the tally where the thousand is cut, you put no other number, unless you need to cut half a thousand, which is done by removing just half the wood in a cut the same length as a thousand, but lower down. And you do the same if you are going to cut£ 100 and there is no thousand there, and the same for£ 20, and the same for the 20s., which we call a pound. But if several thousands or hundreds or twenties of pounds need to be cut on the front edge of the same tally- that is, the one that faces you direcdy with the writing showing- and the smaller number on the other side. Bur the greater number is always cut on the obverse, at the top, the lesser number- that is, the pence- on the reverse.

This passage and other relevant parts of the Dialogus de Scaccario are of particular interest in that they describe the split tally as a highly standardized financial instrument, cut and inscribed according to ever-unchanging rules, used in an institutional environment according to precise and fixed protocols, and for all these reasons highly efficient. One could elaborate much more on the developed functions, such as that of 'assignment' or use of the tally as a transferable cheque payable to bearer that the Exchequer tally quickly adopted, but for our present purposes their basic efficiency in institutional context is what matters. 67 Although we are not in the position to accuse the Achaemenid Bactrian scribes of bureaucratic fetishism comparable to that of their Exchequer colleagues, we do believe that they similarly adhered to standard formats and patterns as they were part of an institutional system.

64

Dialogus 1.3, I.5-6, II.27 (pp. 14-16, 34, 54-60, 188 Amt). The elaborate procedure was regulated by a small (talea) memoranda, documenting the deposit before the assay, and, following it, by a talea combustionis. Whereas the former was destroyed, the latter was attached to a regular tally as proof that the silver was blanched. 65

This is how we understand the following passage, hie [= factor talearum; WH] taleas de thesauro contra vicecomitem ( . .. ) ministrat et, cum oportuerit, secundum quod ratio computationis exegerit, mutat vel minuit vel addit in talea, apposita eidem contratalea vicecomitis. We believe apposita in the last stipulation should be taken more literally than Amt suggests in her translation "comparing it to the sheriff's counter tally" (Amt in Amt/ Church 2007, 33). What is apparently meant is that, during the audit, the final amount paid by the sheriff, after all deductions and additions, is also documented on both halves of the tally put together (apposita), by adding or erasing the appropriate notches. That they were joined is also mentioned in a 1670 report from the Exchequer Qenkinson 1913, 34). The result is an up-to-date receipt. We thus modifY Amt's translation and read the passage as "He [the tally-cutter] delivers from the treasury the tallies that concern the sheriff ( . . . ) and, depending on what the account requires, he changes or subtracts or adds to the tally as necessary, having joined the sheriff's counter tally to it." 66

See Dialogus I.5 (pp. 24, 32 Amt). The text describes the sheriff as a debtor, but at the same time it is he who acts as 'stock holder' between Easter and Michaelmas. In other contexts of split tally use, the stock is commonly reserved for the creditor. English royal commissioners would, for example, issue tally stocks for

commodities they acquired from private parties and keep the foils themselves; the stocks functioned as cheques payable to bearer (cf. Jenkinson 1925, 314£ and §1.2) . The protocol for the Exchequer tallies may accordingly be understood by taking the sheriff's role, in anticipation of the audit, as that of a creditor, i.e. of silver proffered and hence available for immediate use by he crown. T he tally regulated the advance deposit, not the taxation as such: a debtor to the crown ex officio, the sheriff's payment made him de facto creditor to the crown. To say that the Exchequer tallies were essentially receipts (Robert 1956, 78f., followed by Wedell 2001, 237) tends to obscure this dynamic aspect. The term may better be reserved for rhe time after the audit, when the returned stocks remained with the Exchequer and, for all means and purposes, changed into a administrative receipts used by the treasury. It is not clear what happened to rhe foils: were rhey destroyed, handed to the sheriff, or is the fact that hardly any have survived due to pure accident, as Jenkinson holds? C£ Jenkinson 1925,294, 314£; id 1911,374,379. 67 See the literature cited in n.63 above, notably Robert 1956, on the development of the Exchequer tallies into cheques and bank notes.

WF.M. HENKELMAN and M.L. FOLMER 2. ACHAEMENID BACTRIAN TALLIES: FORMAT AND TALLY MARKS

Moritz Wedell has rightly observed that, based on the notches alone, one usually cannot decide what the rally 'says' (Wedell 2011, 235). This is basically true for all rallies, be it from archaeological excavations or in ethnographic collections, that lack sufficient context (for this very reason field observations, like those by Gmiir, are essential). Sir Hilary Jenkinson, speaking of the creators of private English tallies mimicking the format of the Exchequer tallies, wrote that (1925, 320) there is a reasonable probability that by a shilling notch and a penny cut he will usually mean to indicate 13, it is only on rare occasions that a meticulous official will think it necessary to make it clear to the outside world that he means thirteen pence and not thirteen piglets.

Indeed, as Kuchenbuch (2003, 489) philosophically remarks, Kerbholzer sind immer in 'Gesellschaft': ohne das Wissen und die Erinnerung der Beteiligten (den Kreis der praktisch Eingeweihten) sind die schriftlosen Holzer 'tot,' ohne penible Beschriftung bzw. begleitendes Schriftgut kann ihre numerische Botschaft weder raumliche noch zeitliche Veranderungen 'iiberstehen.'

If the tallies are additionally inscribed, one may more easily understand their meaning, but even here there are severe limitations, as the texts tend to be terse and omit information that was obvious to the holders. Needless to say, these observations hold a fortiori true for the Achaemenid Bactrian corpus, where location, date, administrative and archival contexts sadly are all subject to varying degrees of doubt and where one is left to conjecture who the holders were and why they used split tallies. If wooden tallies generally tend to work as fragments from a lost world, the Achaemenid Bactrian ones are shades of such fragments.

YouR TALLY IS FuLL!

159

Novgorod finds from the later tenth century CE onwards, where many tallies are broken. 68 Significantly, the Irish insurgents taking Carrick Castle (Kildare, Ireland) in 1294 CE, burnt the rolls and tallies relating to the county, thus unilaterally annulling outstanding debts (see Hemardinquer 1963, 147). In more regular circumstances, such as the grape harvest in Waadt, the parties involved would together burn their tallies after the account had been settled (c£ §1.2 above). Destroying, in this case breaking, tallies is also known from the Bactrian split tallies with Bactrian script dating to the fourth or fifth century CE that were recently published by Sims-Williams. This last group is particularly important because it gives the closest comparison in time and space and because the general format is very similar to that of the tallies under discussion (c£ §5.4 below). 69 The fact that all eighteen tallies published by Naveh and Shaked are unbroken is unlikely to result from chance preservations and more plausibly indicates that the agreements (or debts) they recorded remained in force and that the tallies were unredeemed. This perspective agrees well with the narrow time-frame of the documents: all but one are dated to "year 3 of Darius" (cf. §3.1; D 18 omits the date formula altogether). The circumstance that the use of a dating formula is nearly ubiquitous and at the same time only one regnal year is mentioned probably implies 1) that such tallies were regularly issued and had to be distinguishable by regnal year, 2) that their validity was limited in time. If one accepts the palaeographic argument (ADAB, 31), the name of Darius can only refer to the third and final ruler of that name, and the date therefore must be 333/32 BCE. If the regular validity of the tallies was, say, one year, they ought to have been reviewed sometime in 332/31 (year 4), after which most would be destroyed, perhaps a few retained because of unresolved problems, and, presumably, a new series would have been issued, dated to the fourth regnal year. That this did not happen does not automatically have to be directly related to the Macedonian invasion (which at any rate had not reached Central Asia by then). One has only

68

Nevertheless, starting from a series of material observations, via comparison with better-known tally traditions (§§2.1-8), and ending with a re-examination of the Aramaic texts (§3), we believe that we can at least indicate a possible context and a possible function (§4).

2.1. Intactness and date At the start it may be noted that none of the eighteen tallies is broken, viz the halves resulting from splitting a tally stick are as such intact. This suggests to us that they still retained legal force at the time the archive was abandoned. In other contexts, the halves of a split tally stick that lost their validity, either because the underlying agreement had changed, had been annulled or, most commonly, because its terms had been satisfied, were usually broken, burnt, or otherwise destroyed. The breaking of split tallies is already mentioned in the Dialogus de Scaccario (confringitur !.5, Amt/Church 2007, 34) and is materially attested among the

See Noonan/Kovalev 2000, 132. Grandell reports that some of the Bryggen tallies are also broken, but does not indicate whether, in his view, this was done purposely (Grandell1988, 66-68). See also §5.8 below on Novgorod and Bryggen. 69

Sims-Williams 2012a, 20f, "Many slips are broken in half, so that only one end survives. In some cases the break is clearly deliberate, being partly cut with a knife" (cf id. 20 12b pis. 118-122 for images). See also Naveh/Shaked 2003, 116, where the breaking of these same rallies is already explained as a sign that the debt was cleared. Wyffels (1988, 28, 31) cites evidence demonstrating that destroying tallies after acquittal was a regular phenomenon in the Low Countries from the high Middle Ages onwards, Vernus-Moutin (1991 , 71) does the same for the French Dauphine, and Lovett (1897b, 164) describes it for 19'h-cent. Brittany. An interesting developed use of destroying split tallies is described by Haibock (1964, 129), referring to the work of the so-called Wender (lit. averter) in pre-modern/early modern Austrian sympathetic magic. In case of skin diseases, the patient and the healer would carve a notch for each wart on the two halves of a tally held together. Then, each of them would throw his half backwards into the fire, expecting that the ailment would disappear after a while. We tentatively propose to see the ritual as a contract between patient and Wender, whereby the latter takes the role of creditor (to be satisfied for his labour), who will take rhe warts (as payment) and rhus rake control of the ailment. The destruction of the tally signifies, besides the magical 'averting,' that the transfer is complete and the account settled.

!60

WF.M.

HENKELMAN

and M.L.

YouR TALLY rs FuLL!

FoLMER

161

to remember the deposit of what is now known as the Persepolis Fortification archive in the northeastern fortification wall of Persepolis sometime after year 28 of Darius I, to realize that an apparent disruption in the bureaucratic system does not necessarily spell the end of empire.

were incised, to facilitate the eventual splitting (fig. 6).7° Whatever the technique used by the Persians, it would appear from the published photographs that their tallies, too, were notched

At the same time, the stress Alexander's campaign caused for the empire's resources, logistic structures and trade networks may have had an effect on the famous 'normal run of things' even in Bactria. What this meant for agreements recorded on the tallies obviously depends of the parties involved: if they were, e.g., landholders serving in the king's army, a mobilization would presumably have had its impact on the system. Another possibility would be that the tallies relate to (semi-)commercial investments (silver loans) and that the trade routes had become disturbed.

Another noteworthy aspect of the Achaemenid Bactrian tallies' format is that they are not perforated. Private and institutional tallies from mediaeval and later Europe often, though certainly not always, have holes that allow for bundling (by, e.g., a tallyman holding half tallies pertaining to a group of hop harvesters) or for attachment to other records. Though split tally sticks in principle constituted a complete record of any transaction, known European contexts sometimes show the parallel use of paper documentation, such as the Exchequer's receipt rolls, or merchants' books. In such cases, the tally was still the central record with binding authority, but its documentary role was a subordinated one. The physical expression of this reality would be the attachment of the tally to a paper (parchment) document by means of the hole in its end.7 1 Such was, however, certainly not always the case: not only split tally sticks in more mundane contexts like milk and bread deliveries could stand on their own, but autonomous use may also be assumed for some of the earliest contexts, such as the Novgorod and Bryggen finds. It may well be that this represents the original use of split tally sticks. At any rate, the fact that the Achaemenid Bactrian tallies are not perforated is potentially meaningful and should

Yet, whatever its historical context, the apparent problem that caused the preservation of a whole range of unbroken tallies from a single year was not a permanent one, but a glitch that was soon mended. Here one may compare the deposition of the Persepolis Fortification archive in one or two small spaces in the northeastern fortification (casemate) wall: the action may be explained from an administrative problem, a change in the hierachy, or the accession of a new king, but in any case it did not affect administrative structures and procedures as such - the Persepolis Treasury archive attests to their continuity (c£ Henkelman 2008, 162-177). Something similar may be inferred from the presence of a leather document dating to Alexander in the Khalili corpus. As has often been pointed out, it suggests continuity of administrative and bureaucratic traditions beyond the Macedonian invasion (c£ §O above). Though we can only speculate on this point, it may well be that the part of the administration that used split tally sticks also continued its operations; if so, the absence of tallies from a later date is not a function of a changed, abandoned, or dysfunctional system, but rather of one that performed as it should.

before they were divided in halves.

be included in their analysis.

2.3. Tally marks: sign types and reading direction The editors of ADAB recognized three types of notches in the Bactrian tallies- what they called broad, narrow, and angular incisions - and assumed an ascending hierarchy between them (ADAB, 231). As is clear from the photographs, however, the notches on the Bactrian tallies are of jour types (figs. 13-14). By analogy with the terminology used in the Dialogus de Scaccario (cf. §1.3 above), these may be defined as:

2.2. Uniformity of format A second important aspect is the tallies' uniformity, which is not self-evident given the great variety of formats that occurs in other contexts. The Achaemenid Bactrian tallies have the same formula in Aramaic (though additions are possible), operate the same system of notches, and seem to have the same orientation (see §2.3 below). Though their length varies for practical reasons (cf. §2.5 below), their width is usually between 18 and 24 mm (three are narrower). Also, their inner surfaces appear to be quite smooth- a requirement for the Aramaic writing, largely limited to this side -which may indicate some abrading, but is more likely to imply a developed splitting technique. Splitting a stick evenly is not as easy as it may seem, for which reason the English Exchequer had developed a complicated method that guaranteed an exact and straight split that cut midway through the notches (c£ n.40 above). Earlier still, at Novgorod, longitudinal groves were made on opposing sides of the stick, before the notches

A. shallow, straight cut which removes no wood B. deeper, straight cut with additional cut

to

remove a little wedge of wood

C. shallow, oblique cut which removes no wood D. deeper, oblique cut with additional cut to remove a little wedge of wood

The length of the tails of the notches varies; in some cases this seems to result from slightly uneven splitting, but generally it may simply indicate that length did not have a particular meaning (as it has in other tally mark systems).

See Kovalev's insightful comments (2007, 197£) on the fortunate find of 13 tallies that had been prepared be split, but were never finalized. They date between c. 950 and 1150 CE and may have belonged to the Novgorodian predecessors of the English sheriffs (i.e. tax farmers). C£ §5.8 below. 71 See, for instance, the mediaeval English private tallies attached to a 1304 Exchequer account presented by Jenkinson 1914. Jenkinson 1925, 3ll£, 319, pis. 63-7 discusses and illustrates the use of parchment

70

to

labels attached

to

(private) tallies.

WF.M.

HENKELMAN

and M.L.

YouR TALLY

rs FuLL!

FoLMER

The heads of notches of types B and D always point to the same direction, regardless whether the notch is carved in the lower or upper edge. Notches of types C-D had to be carved by two incisions in the form of a< mark, with the tip on the line of the future split. In a few cases the incisions ran further and crossed, making an x mark, which shows up as a 1\ on the preserved tally half (these are the 'angular' notches ofNaveh and Shaked). 72 The sticks were split through the tally marks and afterwards inscribed on the flat interior; only in three cases there is also writing on the convex outer side (c£ §3.2 below). These cases are important for establishing the way in which the stick should be held in order to read the tally marks. In two cases (D 1, D 17), the stick was turned along its horizontal axis; once (02) along the vertical axis, with the writing directly continuing close to the left edge. The Achaemenid leather documents in the Khalili collection are not of much help here, as the text on their reverse often is an address line, hence disconnected from the main text. Still, most of these documents are turned along their horizontal axis, as one would do with clay tablets. We assume, as working hypothesis, that the stick normally had to be turned in the same way and that 02 is an exception. If the sticks are turned along their horizontal axis, the heads of the notches of types B and D always point to the left, whether they are on the top or lower edge. This is true for all eighteen tallies in the Khalili collection, a remarkable consistency that must indicate a standard division between two parties. The officials representing the administration, presumably the creditor, apparently always kept that half that had the heads of the notches facing left (extant), the other party, presumably the debtor, always had the one with the notches facing right (not extant) .7 3 As for the direction of reading the tally marks, it is important eo remember that the original, unsplit tally, was a three-dimensional object. Series of tally marks would run from one end of the stick to the other, often in two 'lines' on opposing sides of the stick. Different halves would only emerge with the splitting of the stick, which, ideally, cut all tally marks on both sides exactly in the middle. In a system in which tally marks are cut into a stick starting from the left, with the highest values, the reading direction is to the right, with signs representing decreasing values. Once split, sign sequences will still run from left (the edge) to right, but only on one half: the other has sequences running from the right (the edge) to the left. In other words, on the first half the tally marks would be read prograde, on the other retrograde. This cannot have posed a real problem, though, since ancient users 1) could recognize the beginning (the end of the stick), 2) knew the ascending values of the signs, 3) were accustomed to additive notation, in which order and position of signs is less critical than in positional systems. 72

This is the case on D3 (Ix, lower edge), DII (2x, upper edge; lx, lower edge) and 013 (lx, upper edge). The shape occurs between more regular notches of type C. We see no reason to take it as a separate mark. 73 Such right-facing notches of types B and D, but especially of type B, strikingly resemble the cuneiform sign DIS (l).

Fi

13· Tall marks on the upper edge of split tally sticks in The Khalili Collection;

~~ove: ~T 1

(AD;B 0~); ~ddle: IAT 15 (011); below: IAT 11 (04) and IAT 9 (03); ©The Khahh FamrlyTrust.

F

14· Tally marks on the lower edge of split tally sticks in The Khalili Collection; IAT 11 (ADAB 04); below IAT 15 (011); ©The Khalili Family Trust.

rg;bov~:

We assume, by the near-universal ordering principle of numerical notation syste~s (Chrisomalis 2010, 364), that highest power of the base is stated first on rh~ Achaememd Bactrian rallies. We furthermore hypothesize that _signs oft~~ type B or 0, ;~eh are cl~a~ly intended to be more pronounced, representing htgher postttons tha~ A an b '/espe~uve{ Since, in sequences, B and D always occur right of A and C, respecuvely, V:e. e teve t at t e reading direction of the signs is retrograde on the preserved halves. On the mtssmg other halves, the reading direction would have been prograde.

An inverse parallel from the modus operandi of the English Exchequer may b~ ins~uc~~e at this point: the tally-cutter would cut the tally marks from left to right, ~er whrch: e t re~ would first be inscribed in Latin and then split Qenkinson 1925, 308). Thrs syst~m ~vou the holder of the stock, since only here tally marks and inscription woul~ ru~ 1~ t e same direction (prograde). On the foil, the reverse of the stick so to speak, the mscnpuon would, naturally, still be prograde, but the rally marks retrograde (cf. ibid. 294). In Achaemenid Bactria, the administration consistently preferred that half of the spl~t tall on which the reading direction of the tally marks wa~ the. sa_me as that of the AramaiC ins~ription. If the missing halves were also inscribed, the mscnpuons there would have run counter to the reading direction of the tally marks.

---------------------------rl WF.M. HENKELMAN and M.L. FoLMER

2.4. Tally marks (his): notation system Since none of the four types of notches ever occurs in a continuous series of more than nine, we propose to read them as markers of grades (or powers of the base) in a decimal system_74 They could represent four successive positions in a single system and be symbols for 1, 10, lOO and 1,000, but this is hardly the only possibility, and arguably not the most likely. A slight variation would be that the simplest symbol (type A) denotes a fraction (one tenth) of a standard unit. This would be especially relevant if the tallies record transactions in silver. 75 Yet, entirely different systems are possible as well, and these, too, would find parallels in other tally traditions. One alternative would be that the straight incisions (types A-B) do not count the same things as the oblique ones (C-D). The first could, e.g., represent amounts of a certain commodity, the others containers, days of harvesting, value, or some other category. The Kentish hop-tallies (c£ §1.2 above) counted the number of bins turned in, as well as the number of days worked. The location of the marks on the tally may be relevant as well. Wyffels cites examples of baker tallies expressing the value of the loaves delivered expressed in guilders on one side and in stuivers (shillings) on the other; Scheuermeier knows of Italian day labour tallies noting, on opposite sides, the units of work performed and the commodities issued in kind.7 6 In the case of the Achaemenid Bactrian tallies, it is certainly possible that the notches on the lower edge have a different meaning from those on the upper one. Obviously a combination of location and sign type for different categories should be considered as well. In that case the use of straight vs. oblique signs and their place on the lower or upper edge determined, e.g., female vs. male animals and goats vs. sheep. A closer look at the distribution of the notches provides a little clarity. The marks occur in multiple sequences on every tally, separated by pronounced blank spaces. Only types A and B and types C and D are regularly found grouped together; the few cases where other combinations seem to occur are all debatable. Both type A and type B may run up to nine notches in row; we assume that the simpler mark, type A, represents the lowest grade. Types C and D occur less frequently but seem to have a parallel relation, with D representing the highest grade. We assume that they, too, are part of a numerical system with a decimal base, though we cannot be absolutely certain in this case (the longest sequence attested is 8 for C and 7 for D). If types A-D are part of a single system, then one has to assume some kind of boundary between B and C, as there are no unambiguous sequences in which both occur. 74

Other, non-decimal systems are known in conjunction with (split) tallies, such as the quaternary system of counting bundles of furs in Scandinavian and related contexts: four such bundles (each of 10 skins) constituted one timbr (40). A tally stick with every fourth notch extended (i.e. marked) illustrates this way of counting (see Grandelll988, 67 with fig.2). The system is already attested at Novgorod, where it also seems to have been used in the context of the fur trade. Also in Kievan Rus', ternary and septenary counting bases are known. See Kovalev 2003, 167f.; id 2007, 191-194. 75 Tally mark systems in which a simple, shallow notch does not represent the basic unit, but a fraction thereof are not uncommon. See, e.g., Gmiir 1917, 53, 56. 76 Baker tallies in the Low Countries: Wyffels 1988, 27 (also mentioning an example of two rows of notches indicating whole and half barrels); compare the example cited by Brunner 1912, 339. Day labour tallies in early 20'h-cent. Italy: Scheuermeier 1943, 12 n.l (noting that payments in bread, oil, meat and other commodities were re-calculated to theoretical units of bread). Vernus-Mourin describes tax tallies from the Dauphine whereby separate edges are used for gros, half gros, quarter gros and patac (1991, 69-71).

YouR TALLY IS FuLL!

The following table lists the sequences of types of notches on the upper and lower ed?:s of each tally (adding up B.AAAAAA as 1.6BJ with totals per edge, followed by grand totals.

tally upper edge

total

lower edge

total

grand total

1.9sA; 0.4nc

9.9aA; 5.2oc

Dl

1.6BA; 3A; 4.8aA; 1.3aA; 4.8nc

8.0aA; 4.8 0 c

9/ 8; 4A; 6A; 4c

D2

la; 6A; 1.7aA; 4.7aA; 1.8B/9; 4.7 1oc

9.8aA; 4.7 1oc

4A; 6A;

D3

4.9BA; 4A

5.3BA;

3c; 1.3BA; 1s

2.3BA; 0.3nc

7.68 A; 0.3nc

D4

1.68 A; 6A; 3s; 9.1 8A; 3.5BA; 7o

17.8aA; 7.0oc

7A; 18 ; 1.28A; 4.1 BA; 1.5aA; 4.lnc

8.5BA; 4.1 DC

26.3aA; 11.1 nc

D5

7A80; lA

0.8BA;

SA; 6A; 1.4BA; 5.4aA; 1.3BA; 5.4oc

9 .2aA; 5.4nc

lO.OaA; 5.4nc

D6

3.3BA; 1.3BA; 1.2BA; 48 ; 1.3sA [. · .]

10.98A [... ]

no notches (surface damaged)

0;0

10.9aA [. .. ]

D7

no notches

0;0

2n; 3s8I

3.0 8A; 2.0nc

3.0aA; 2.0 0 c

0;0

3.3BA; 2n-3A82

3.3BA; 2n-3A

3.3BA; 2n-3A 10.5 8 A; 0.5nc

10.1 8 A;

15.48A; 3.2nc

D8 D9

no notches

0

0

h;

1.6aA; 1.6nc 3.3sA; 1.6oc

13.1BA; 6.3 1oc

DlO 3A; Ss; 2.5oc; 7c

5.3 8A; 3.2DC

lA; 2c; 3.7BA 2A; 2.8 8 A8\ 6a; l.lBA

3.8 8A; 0.2 0 c

Dll SA; 2c(-)9.3sA

9.88A; 0.2DC

68 ; 2c(-)3.7sA

9.7aA; 0.2nc

Dl2 lA; 6.5aA; 1.6nc; 1.4nc

6.68A; 3.QDC

1

3A; 3.2 sA; 3.8BA; 5C?

7.3 sA; 0.5nc? 13.91aA; 3.5DC?

Dl3 SA; 4A; lc(-)8.laA

9.0 8A; O.loc

2.4BA

2.4aA;

1.38A; 0.3oc

SA; 6c; 1.2aA; 6c84

2.0 8A; 1.2nc

3.3sA; 1.5nc

SA; 4A; 3c?; 9A

1.8aA; Q.3DC?

6.8aA; 3.9DC?

2.1 BA; 4.6BA; 3c

Dl4 3A; 3c; la

6.7BA; Q.3oc

Dl5 la; 2A; 1.6nc; 3.8BA; 2n 5.0sA; 3.6nc Dl6 la; 20 ; 9.2 8A; 4.3 0 c

10.28A; 6.3 0 c 3A; SA; le; 2.1BA;

Dl7 3.8BA; 3.4BA

7.2BA;

0

Dl8 3s

3.QBA;

0

7c

0

19.5sA; 0.4nc

1

0

11.48A; 0.1 0 c

2.9aA; 1.4nc

13.l aA; 7.7nc

1.8aA

1.8aA; 0

9.QBA;

3.5aA

3.5BA;

6.5aA; 0

0

0

The data presented here are adapted from ADAB, 230-258. A few minor errors in counting individual notches have been corrected, as have the corresponding totals (D3, D5 , D9). Whereas Naveh and Shak~d in most cases correctly identifY the upper edge as 'above' and the lower as 'below,' the photographs of t e convex sides are always upside down with respect to the notches, except in the cases o~ D 1 ~nd 17. Instead of the three types used by Naveh/Shaked, we have applied the four types of notches 1dennfied m §2.3. 78 ADAB, 233 reads this sequence as 9B. 79 Not 7A, pace ADAB, 237. 80 Not 6A, pace ADAB, 243. 81 These notches are not on the upper edge, pace ADAB, 246. 82 This is the only case where A-type marks follow D type marks. Is 3A and error or an imprecise carving for 3C? Note that the notches are not on the upper edge, pace ADAB, 247. 83 There is no space after the 8A sequence, 6B follows directly. 84 The upper edge here is what is "below" in ADAB, 253, the lower edge "above." 77

I?

166

One of the striking features made visible in the table is that more than one DC (or C, D) sequence per edge is rare (D10obv., D12obv., D14rev., D15obv., D16obv-), whereas it is the rule for BA (or A, B) sequences. Furthermore, the upper edge on average has more sequences and higher totals, yet there is not a single pattern fitting all tallies. There is no consistent ratio between the totals for BA and for DC signs. The numerical system as such is simple and can be characterized as cumulative-additive and having (cf. §2.3 above) a decimal base, be it that A-B and C-D may not represent, at least not on these tallies, four successive grades (or powers). The system is cumulative because the sign types or logograms are repeated until a higher position is reached (AAAAAAA = 7, B.AAA = 13); there are no separate signs (logograms) for 2, 3, 4, etc. It is additive because any total expressed is the sum of the values of each sign, regardless of their position. 85 All Aramaic texts in ADAB, be it on leather or on wood, use the common Aramaic numerical notation. The tally marks are probably not derived from this system, though they share their decimal base with it. First, the Aramaic signs are different in shape (except for Aramaic '1,' which, unsurprisingly, resembles types A-B). Secondly, Aramaic numerical notation bundles signs for legibility (Ill Ill Ill = 9), which the tally mark notation does not. Thirdly, the Aramaic system has a separate sign for 20 (20+20+20+20+10+3 = 93 (C1:48]). Most critical, however, is that Aramaic numerical notation is a hybrid system; it is, as Chrisomalis (20 10, 7074) explains, cumulative-additive up to 99, but above 100 it becomes multiplicative-additive. Thus, in document C8:3, '800' is expressed with eight vertical strokes preceding the sign for 100 (8 · 100 = 800). The common Assyro-Babylonian numerical notation, from which also the Elamite system derives, is similarly hybrid and has its turning point also at 100 (8 ME= 800; cf. ibid. 247-249). We are, then, in a situation comparable to that of the English Exchequer, which operated a cumulative-additive notation on its tallies until 1826, yet existed in a larger cultural context in which the 'western' (Hindu-Arabic) numerical notation system was increasingly used. 86 In Bactria, too, the tallies, had (kept) their own notation system, besides another system that was more closely connected to writing. The relative plainness of this tally mark system clearly should not be misconstrued as an indication of simplicity of accounting techniques or administrative protocols. Perhaps it is worth repeating that the Bactrian cumulative-additive notation is different from systems that have no symbols for higher grades (or powers) of the base. In the world of tally marks, such undifferentiated systems are not uncommon. They may include special signs for the tenth (and fifth) position in a sequence, bur only as unit marker. Examples are

85 86

YouR TALLY

WF.M. HENKELMAN and M.L. FOLMER

For different types of numerical notation see the useful typology in Chrisomalis 2010, 9-14.

But note that the cumulative additive numerical notation on the Exchequer tallies is more complicated in the sense that it was pre-decimal (Carolingian) and used a combination of different bases.

IS

FuLL!

IIIIIIIIIXIIIIIIIIIXII or IIIIVIIIIXIIIIVIIIIXII as notations for '22.' 87 Whereas such notations can be used in a variety of social and economic contexts, they will typically not relate to more complicated (ac)counting with the help of a counting board or abacus. The essence of those devices is, after all, that tokens representing lower grades can be subsumed under a single token for a higher one, hence equivalent to a notation like XXII for '22.' In this context, we refer again to the Exchequer tallies, which were used contiguously with an elaborate counting board (the checkerboard cloth or scaccarium). Whereas the Bactrian numerical notation system seems purely decimal, some other systems reflect a quinary sub-base used alongside the main decimal base (5, 10, 50, 100, 500, etc.). In such systems, there usually are separate symbols for the grades (powers) of the quinary sub-base (CCLXV = 265). 88 Not only does the notation system on the Bactrian tallies not indicate a sub-base, but the sequences also do not indicate an underlying decimal-ternary counting system. This deserves full emphasis since Naveh and Shaked thought of the tally marks as indicating, perhaps, a commodity like flour or grain. The measure of capacity used to express quantities of cereals and some other dry commodities is counted decimally, but with a ternary sub-base (1, 3, 10, 30, 89 etc.). This is true for Bactria (ADAB) and Persepolis (PFA) alike: Old Iranian

Aramaic

Elamite

*da8v(i)ya-, "tenth"

bpn (abbreviated b)

daduyaorQA

grade

quantity in liters

0.1

0.971. 2.91 l.

I

bawif

0.3

gryw (abbr. g) or s'h

kurrima or BAR

1

*rdba- (rtba-)

'rdb

irtiba

I

I

3 (10)

(expect 97

gwn (abbreviated g)

30

291

(?) *griva-, "bushel"

(cf. Skt. gorJi-, "bag")

9.7 l. 29.1 l.

l.) I.

In both Elamite and Aramaic, there is a scribal convention for the alternating use of units of the decimal and the ternary base without confusion. In Elamite (PFA), scribes could opt for notaSee, e.g., the dolly-peg-shaped simple tally stick (c.1200 CE) from Bryggen (Grandell1988, 67 fig.1). In view of the plausible origin ofEtrusco-Roman numerical notation in the use of tally marks (cf. n.31 above), it is often believed that Roman numerals like V (and also, perhaps X), originally served as boundary or unit markers. A notation like XIV would then be an abbreviation or subtraction notation (for the term, see Chrisomalis 2010, 348) for IIIIVIIIIXIIII (as known in some tally mark systems) and indicate an emancipation of the sign V from unit marker to logogram. 88 Decimal-quinary notation was not uncommon in mediaeval and later tally stick traditions; for the Roman case see Chrisomalis 2010, 95. 89 For the Aramaic terms see ADAB, 37 and Tuplin 2013, 70 (in comparison with PFA and the Ars:ima letters). In Aramaic s'h, bpn, and gwn can also be used as liquid measure (ibid.). For the Elamite terms see Hallock 1969, 31, 72-74; for reconstructed Old Persian forms see Tavernier 2007, 38, 81, 448-450. Note that griva- may be attested on the Old Persian tablet from Persepolis (Stolper/Tavernier 2007, 13). There are some further terms that could be units for (dry?) measure, but they are not fully understood. See, e.g., HinziKoch 1987 s.vv. ha-ma-zi-if, SU.lg, zi-ut.

87

168

WF.M. HENKELMAN and M.L. FoLMER

tions like "21 1 BAR," which reads as "21 (irtiba) + 1 BAR" (64 BAR= 621 l.). The notation is very short and practical, also because BAR, a logogram, is a simple sign and is written, in such combinations, without determinatives (G 15 BARMES). In Aramaic, the names of the units are likewise abbreviated, as in" '2 s 6," i.e. "2 ardab, 6 se'tl' (C3:22). Such shorthand notations can actually be seen as a first step towards a stronger multiplicative element in the numerical notation. The Achaemenid Bactrian tallies do not seem to refer to a commodity counted in units in a mixed decimal-ternary system, though this would not have been impossible or unparalleled. Nor is, to come back to the suggestion raised above, a mixed base an obstacle for the use of a counting board - in fact the use of such a device for the more complex calculations in the Persepolitan journals/accounts and the Bactrian leather documents is quite likely. 90 It cannot be said, however, that the absence of a clear indication of a mixed base in case of the tallies necessarily means that the commodity implied was not flour or grain. In the Fortification archive, the decimal-ternary system was actually used interchangeably with a purely decimal one. In texts produced by lower administrative strata (the tongue-shaped memoranda and related texts), the mixed base is rarely found, whereas it is common in the 'journals' (registers) and accounts, which were produced by higher-trained scribes in Persepolis (cf. Hallock 1969, 31, 46). 91 Moreover, not everything could be counted in the decimal-ternary system: it was reserved to units of dry (and perhaps liquid) measure, hence commodities measured out in jars, sacks, etc. of standard sizes, but not applied to itemizable things such as cheeses, sheep or people. By analogy, two basic options apply to the Bactrian case: 1) the tallies represent a lower administrative stratum, 2) the tallies refer to a commodity not counted in the decimal-ternary system. Incidentally, measuring silver weight was also not exclusively done in a decimal system, but involved yet another mix: besides a decimal base (10 shekels equal1 karfa), a sexagesimal sub-base occurs (1 mina equals 60 shekels or 6 karfa). Conversely, as the material from the Treasury archive from Persepolis shows, the basic unit, the (weight) shekel, could be divided by 2, 3, 4, 5, 6, 8, 9, 10, 12, 18, 24, and 40. In theory, these can be subsumed under 18, 24, and 40 as underlying dividers, but this is by no means the only way of understanding a clearly quite intricate system. As for its relevance for the present discussion, it is important to remember that the accountants in Persepolis operated complex fractions in the first place because they had to express the diminutive silver payments as equivalent to rations in kind according to fixed conversion rates (10 qa of barley for 0.3 or 0.33 shekel). 92 90

See the lucid demonstration of the use of a counting board with English pre-decimal currency (as practiced at the Exchequer) by Baxter 1994. 91 The same scribes not only used a mixed base, but also experimented with positional notation: in the broad left margins of their tablets, position often determined rhe power of the base (or sub-base) referred to. 92 Fractions: Cameron 1948, 37-39; Hallock 1950, 240f.; Cameron 1965, 182-187; Tavernier 2007, 450£, 455f. (etymologies of Iranian terms). The system is clearly related to that of Achaemenid Babylonia, but different in some details. Note that the \4o fraction does not occur as such, but is expressed as \4 · \!to; the Yz4 fraction only occurs on the Babylonian tablet from the Treasury archive (PT 85: giru), but seems to underlie the simultaneous existence Vs and V6, unless Ytso was the communal basis, from which, then, also 1/9 and V1s would derive. The interchangeability of the various fractional bases could be understood as another indication for the use of a counting board, more precisely one with different sets of tokens for each base. On the conversion rates see Hallock 1969, 17 and Henkelman 2005, 145-149.

YouR TALLY

IS

FuLL!

To summarize briefly the last two sections: there are four (not three) different kinds of tally marks which seem to represent grades (powers) in a decimal system, but perhaps not in a continuous series. On the extant tally halves, which consistently represent the same part of the stick, they are meant to be read retrograde, like the Aramaic inscription. Tally marks A and B seem to denote continuous powers (10 A= 1 B); they occur regularly in multiple sequences (separated by blank spaces). Tally marks C and D presumably have the same relation (10 C = 1 D), but this is less obvious, basically because the marks are less frequent and rarely occur in more than one sequence. As they occur mostly at the end of a 'line' of tally marks, they may represent a different category, such as containers. The tally marks show only superficial similarities with the Aramaic numerical notation system. Flour, barley and some other commodities could be counted, in Bactria and Persepolis alike, in a decimal system with a ternary sub-base; the accompanying numerical notation system operated a simple logogram (PFA) or abbreviation (ADAB) to indicate the measure and, hence, the base. There are no traces of this in the tally mark notation system.

2.5. Tally marks and tally length Because of their cumulative-additive numerical notation, higher numbers required much more space on the Exchequer tallies, a fact that His Majesty's Tally Cutter had to reckon with. In case of very large amounts, one had no recourse than to use lengthy sticks. Although there seems to have been a maximum length of five (!) feet, even longer tallies actually exist. The Bank of England still holds the tally stocks (never redeemed) representing its original loan to the crown (1694), one of which measures 6ft., 8 in. (c.2.60 m). 93 The length of the Achaemenid Bactrian tallies, too, shows a correlation with the number of notches. The correlation is not absolute, but strong enough to show a trend, as the following table indicates:9 4 012 tally

length

tally marks

018

7.5 cm

017

8.7cm

11 notches 6.5BA; 27 notches 9.0BA;

08

9.0 cm

11 notches 3.3 8 A; 2 0 -3A

07

9.2 cm

03

11.5cm

09

12.5 cm

5 notches 3.0BA; 2.0oc 25 notches 7.68 A; 0.3oc 29 notches 10.5 8 A; 0.5 0 c

014

12.7 cm

013

12.9 cm

totals BA and DC 0 0

30 notches 3.3BA; 1.5oc 25 notches 11.4BA; 0.1 DC

························· ···· ······· ·························

14.3 cm

48 notches 13.9?BA; 3.5DC?

010

15.5 cm

015

17.5 cm

42 notches 15.4sA; 3.2oc 44 notches 6.8BA; 3.9DC?

011

18.7 cm

016

19.0 cm

37 notches 19.5 8 A; 0.4 0 c 46 notches 13.1 8A; 7.7 0 c

··· ····· ···························· ···· ·· ······· ············ 01 19.7 cm 61 notches 9.9 8 A; 5.2oc 20.7 cm 21+ notches 10.9BA [... ] 06 02

22.7 cm

77 notches 13.1 8A; 6.3?oc

05

22.8 cm

04

25.3 cm

46 notches 10.0sA; 5.4oc 68 notches 26.3 8 A; 11.1 0 c

93

Jenkinson 1911, 378; id. 1925, 307; Robert 1956, 83£; Baxter 1994,231£ The total numbers of notches are corrected from ADAB, 230-258, where different totals are given for 03, 05 and 029. Cf. n.77 above.

94

j .

WF.M. HENKELMAN and M.L. FoLMER

We have tentatively divided the eighteen tallies into three groups. The first, up to 13 cm, has between 5 and 30 notches, the average being 20.3 notches. The second group, running up to 19 cm, is more consistent, with 37 to 48 notches (average 43.4). Only in the last one, up to 25.3 cm, there is an outlying case (D6) of a mere 21 notches on a tally of20.7 cm, but is clearly because the tally is damaged. The other tallies in this group have between 45 and 77 notches, with an average of 63. That the length of the tally is congruent with the number of notches may strike as obvious, but actually tells us something about the nature of the transactions recorded on the tallies, for it indicates that the tally cutters knew in advance the total amount to be carved in the wood. In this context, a 1670 description of the protocols operated at the Exchequer may be cited; it states that the Tally Cutter95 cutteth a well season'd and proporcionable Sticke ofHazel! wood (which he before was provided of from the Usher there) into four square sides and a suitable length.

The tally cutter at the Exchequer could easily anticipate the necessary length of each tally stick, as it was notched at a single occasion, viz the deposit of an advance payment of rent (or of a series of rents). It could be adapted during the audit, but it was not a running account like tallies for pots of wine to be paid to the tavern holder or field work to be reimbursed. In these last two scenarios tallies were incised with notches at multiple occasions; in the case of the hop-tallies from Kent and Worcestershire this was at each time the picker presented an agreed number of bundles to the tallyman. A certain anticipation of the amount of work to be done may, obviously, also have informed the length of tally cut for such purposes, but this would hardly lead to the kind of correlation shown by the Bactrian tallies. 96 In fact, the grape harvesters' tallies from Waadt, though having quite different totals of notches, have almost exactly the same length on account of the fact that it was unknown beforehand how much they would collect from the vines (cf. §1.2 above).

2.6. Sequences of tally marks The tallies issued by the Exchequer to the sheriffs upon their advance payments at Eastern indicated one single amount, even if it represented the sum of several rents. The Achaemenid Bactrian tallies, by contrast, invariably have more than one series of notches, suggesting multiple amounts. This may also be the reason that Naveh and Shaked- though they do not make their argument explicit - hold that the notches could have been made "on different occasions when allocations were made and have been added to within the same year according to need" (ADAB, 31; cf. Naveh/Shaked 2003, 114). 95

Citation after Jenkinson 1913, 33 (emphasis original). The report goes on to describe the inscription of the tallies by the scriptor talliarum. Incising the notches and writing the short statements on the tallies were apparently done by different individuals.

YouR TALLY rs FuLL!

As we have seen, the editors of the Bactrian documents also suggest that the tallies reflect a situation "in which three government clerks issued goods to workers or soldiers on account of their wages." Additional allocations could be made, until the full amount was disbursed, at which point, still according to Naveh and Shaked, both halves of the tally may have been handed to the recipient (Naveh/Shaked 2003, 114). One counter argument that immediately springs to mind is that tally marks of types C-D, which are more rare and usually add up to lower total values (of whatever they represent) than types A-B, typically come at the end, i.e. on the left side of the half tally sticks as we have them. Though they do not seem to be a summary of the preceding BA sequences, they are likely to represent a different or additional category that had to be stated separately and was usually reserved for the last position. If the tallies were running accounts, one would not expect this kind of organization. Another point is the correlation between the length of the tally sticks and the number of tally marks, as noted in the previous section: such coordination can hardly be expected from running accounts. A third argument is that the sequences of BA type tally marks have values that are very hard to interpret as partial rations. Take the case of D4 (fig. 13): upper edge 1.6 8 A; 6A; 3 8 ; 9.1 8 A; 3.5BA; 7D (total17.8BA; 7.0oc) lower edge 7A; 18 ; 1.28 A; 4.1 BA; 1.5 8 A; 4.1 oc (total8.5 8 A; 4.1 0 c; grand total26.3BA; 11.1od

If the type B tally mark represents agryw (*griva-) of9.71. and type A a tenth of that, 0.971., the totals would be rather high. For the sake of clarity we will simplifY the measures of capacity to 10 and 1 1. respectively: upperedge 16;6;3;91;35;70

(total178 I.; 7.0 0 c)

loweredge 7; 1; 12;41; 15;4.1 0 c

(total851.; 4.1 0 c; grand total2631.; 11.1 0 c)

The total of 263 1. (which excludes whatever the C-D tally marks represent), if expressing rations in grain or flour, would be enough for 263 1 1./day or about 175 1.5 1./day rations -to cite the two most commonly attested rations attested for male adults in Persepolis (PFA). In itself it would not be unreasonable to think that this represents the total amount given for days of work within one administrative year, but if so, it is hard to understand the individual allocations: why sometimes 3 1., and another time 91? Another option would be to assume that the system represents (only) the ternary sub-base of counting measures of capacity (i.e. in a seemingly decimal way: 3- 30). In the Fortification archive, there are a few documents that use bawif (c.3 1.) and irtiba ('rdb; c.30 1.) for ration scales (rather than QA = c.1 1. and BAR= c.1 0 1.); this is a handy way of expressing amounts in case the rations are 2;3 of the standard scales (see, e.g., PF 1955, PF 2038). Type A would stand for a hypotheticall/3o fraction of the gryw = c.0.3 1., type B for the Aramaic equivalent of the Persepolitan bawif of c.3 1.:

96

Lovett (1897a, 40) notes that the last notch of each man's daily work is sometimes marked in ink. From this he could read past weather conditions: five notches on a single day correspond to clear skies, one or two to very bad weather. In other words: the eventual amount of notches was as unpredictable as the English climate.

upper edge 31.8; 1.8; 9; 27.3; 10.5; 7o lower edge 6.9; 3; 3.6; 12.3; 4.5; 4.1oc

(total 80.4 I.; 7.0 0 c) (30.31.; 4.1 0 c; grand total120.71.; 11.1 0 c)

! f'""""''!

172

YouR TALLY rs FuLL!

WF.M. HENKELMAN and M.L. FoLMER

With this conversion, the fact remains that the individual allocations seem erratic, and do not agree with any of the known regular allocations in the Fortification archive. On the basis of the correlation between tally length and number of notches, we have already doubted the interpretation of the Achaemenid Bactrian tallies as records corresponding to running accounts. This doubt finds support in the observation that the C-D type tally marks usually come at the end of a series of multiple sequences of A-B type marks, again suggesting that the tallies were notched at a single occasion. Thirdly, as noted above, the amounts from the BA sequences, whether converted to gryw or to 'rdb, do not match known ration scale patterns.

As can be easily observed from the images published in ADAB, the 'lines' of tally marks on the upper and lower edges give a rather regular impression. In connection with this, it may be remembered that split tallies often have an asymmetrical shape: a smaller foil fits into a larger stock with a socket or stump (c£ §§ 1.1, 1.3, 5.8). This is especially useful when notches have to be added regularly and the two tally halves have to be fitted together exactly and tightly. In the Bactrian case, there is no indication that this format was used. If the extant halves are foils (in itself implausible), they should have had a sharp, angular end to fit into the stock's socket. Since they do not, it is more than likely that the sticks were split evenly. In other words, their format forms a fourth argument against the running account thesis, for it would not have been very easy to hold the two equal halves tightly together, match the tally marks exactly and add additional signs as neatly as they actually have been incised. Rather, as argued before, the tally sticks were split after they had been notched with all the tally marks. To these arguments, a few subsidiary points may be added. Ration texts in the Fortification archive usually are communal records for entire groups of workers, with their sex and age groups specified yet documented on a single tablet. From that perspective, it would be surprising to find similar workers in Bactria with an individual paper trail. One might instead argue that the tallies were in fact communal accounts (with each BA sequence representing one individual), but if so the groups would be rather small, and their ration scales strikingly erratic. Finally, it may be argued that flour, the commodity that Naveh and Shaked suggested, probably because of the occurrence of a word for "white" in one tally inscription (c£ §3.3 below), is not regularly given to mere workers (kurtas} in Persepolis, who typically received their rations in unground barley. The few exceptions to this rule pertain to travelling groups (who had no time or means to grind their own portion) and to groups who received (small) extra rations, or groups who were otherwise special. The idea that flour was regularly distributed to workers in Achaemenid Bactria strikes us as unlikely. A more fundamental, final point is that split tally sticks, as Haibock pointedly put it (1964, 128), were used everywhere where a certain degree of mistrust was appropriate (§1.2 above). They served credit situations in the broadest sense, especially if the parties involved were not in daily contact, did not know each other well, or if there was a hierarchical difference between them. One might argue that the difference between an administrator and a worker receiving rations fits this description, yet the transactions between them required no negotiation. In fact, ration payments seem to be entirely unknown in the world of the split tally. The reason is probably very simple: (institutional) ration payments did not constitute a credit situation, and certainly not one for which proof in the form of a tally was needed.

173

~ ffA!I!I!!I!I:+AIIIII:±IIJIIII:Of+IIII:VIIII: ±Ill: 1111111111 ±A 11 1:

ii \i ~ ~ ~ i! :l i\ 1i !! !' ~

~e==~++~,,~,,v~·~ii~ii~li~ii~i:________~~

Fig. 15: Matching tally sticks of a herdsman (longer) and farmer Cab (shorter);, above: groups of animals transferred (one on Cab's stick, several on the herd~man~), below: the same for pounds of butter paid to the respective farmers; GroB Wtstermtz/ Vala5ski Bystfice (Czech Republic), 19th? cent. CE; after Domluvil1904, pl. 8.

As we have seen (§0 above), Naveh and Shaked seem to have realized that split tallies imply the roles of creditor and debtor and probably for that reas~n ini~ially considered _the possibility that the tallies were a kind of vouchers relating to. runnmg ran on accounts agamst which workers drew quantities of flour (or another commodtty). That would, however, make the institution the debtor, which disagrees with the language of the inscriptions (cf. §3.1 below). Again, one might argue that the institution considered the ration p~yments a~ part of an account that had not been set. That is not excluded: the same was essennally true m Pers~­ polis, but there (as everywhere in the ancient world) ration payments were made on t~e basts of rosters, ration lists, standing orders and the like. The control took place between dtffere~t administrative echelons and responsible officials, not between the institution and the peo_ple .tt deployed and fed and who, in other terms, were already fully in its control and for whom tt dtd not need the security offered by the split tally. There are many alternative possibilities for explaining the multiple sequences of tally marks on the Achaemenid Bactrian tallies. Keeping parallel or complex accounts on one and the same tally, be it a simple or a split one, is in fact well attes~ed, includi_ng, among other contexts, in livestock management. In nineteenth-century Moravtan Wallachta, tenant farmers with an overcapacity in pasture land would accept sheep from smallholders and entrust them to a contract (head) herdsman or baca. Domluvil describes how the tenant farmer and the shepherd kept track of the complicated flocks (1904, 207; c£ fig. 15): Der baca verzeichnete auf seinem Stocke, vrub, durch Kerbeinschnitte (ebenfalls vrub oder vrubek genannt) wie viele und was fiir Schafe er von den einzeln~n ~au~rn zur Weide ~bernahm; auf einem kiirzeren Stabe oder einem anderen Holze, zum Betsptel emer Dachschmdel, wurden diesel ben Zeichen eingekerbt und dem Bauer iibergeben. Auf di~se. Kerbholzer wu~~e auch verzeichnet, wie viel Salz dem baca fiir die Schafe iibergeben und wte vtel Butter und ~se _ Schaftopfen _ dem Bauer von der Anzahl seiner Schafe abgefiihrt wurde. lm Herbst endl~_ch, wenn die Weide beendet war, erhielt jeder Bauer seine Schafe nach dem Kerbholze abgezahlt zuriick· in diesem Verkehre also galten die Kerbholzer als ganz gute Beweismittel, denn die Zeiche~ der nebeneinander gelegten Stabe des baca und des Bauern mussten vollstandig kl~p­ pen, als waren sie mit einem Schnitte eingekerbt. Auf dieselbe Art wurde selbstverstandhch auch iiber die abgefiihrte Butter und den Schaftopfen abgerechnet.

j

YouR TALLY IS FuLL!

WF.M. HENKELMAN and M.L. FOLMER

174

!75

Domluvil goes on to explain the tally marks and tally stick formats used in this system, with separated surfaces reserved for animals, cheese and salt, separate entries for each farmer sending livestock, and distinct marks for lactating, half-lactating ("Halfschafe") and gilt animals.97 Such complicated notation systems are also described by Kuchenbuch, describing the case of a Hungarian swineherd charged with keeping the animals of several farms, and by Gmiir, who cites examples of single tallies (of type known as Alpbritsche) with a count of cows, oxen, horse, sheep and goats. 98

2.7. Physical quality There is another aspect of the Bactrian tallies on which later traditions, notably that of the Exchequer, provide some comment: the orderliness or even attractiveness of the preserved objects. Jenkinson, writing in 1911, felt that "only England systematized the tally into an official instrument cut strictly according to certain rules" and added, with great certainty and satisfaction, that we have, as a nation, a genius for systematizing customary things: upon which account the writer inclines to refuse the theory that the official tally was a Norman importation (Jenkinson 1911, 368).

Around the same time Sir Marc Aurel Stein was travelling and excavating the ruins of ancient Khotan (cf. below, §5.5). In his publications on these he consistently displayed a keen interest in the regularity, convenience and appeal of the ancient "stationery'' that he had brought to light. Of, e.g., the split Sogdian tablet found in association with the Chinese limes system (§5.3), he wrote that it struck him as

Fig. 16a: Octagonal unsplit milk tallies (Stialas de latg); Tavetsch, Graubiinden (Switzerland), 19th? cent. CE; after Gmiir 1917, pi. 21.1-3.

a device as carefully thought out as many other details of the ancient stationery in wood that I have so often had the occasion to examine (Stein 1921, 654; c£ ibid. 372, 597£ and passim).

Indeed, the systematic formats of the Khotanese materials (seventh-tenth century CE), notably the split tallies, are striking arguments against Jenkinson's intuitions about the English nation. More important, however, is that they show that the Exchequer tallies' appealing design has good precedent in the long history of the tally. This aesthetic aspect has long been commented on by ethnologists; Gmiir's 1917 monograph on the early-modern Swiss tallies, often carefully cut and adorned with Hausmarken, provides a telling case in point. 99 Some tallies developed

97 See Domluvil 1904, 207-210; Domluvil discusses and illustrates one example of a tally that has also separate symbols for counting cheese and salt. 98

Kuchenbuch 1999, 306; Gmiir 1917, 78, pi. 14.2-4 (Alpbritschen dating to 1761, 1813 and 1867). See also Gerschel1962, 157-61 (interpreting the notation system of the early mediaeval Irish Ogham script along the same lines: complex accounts for cows and sows). See also the pentagonal and other complex tallies referred to in n.29 above. 99 Gmiir repeatedly draws attention to the aesthetic quality of the objects he discusses and, from his background as professor ofLaw, singles them out for their physicality (1917, 3): "Wir haben hier ausnahmsweise

Fig. 16b: Stiala with 16 entries on eight surfaces, with Hauszeichen; Graubiinden (Switzerland), 19'h? cent. CE; after Stebler 1907, 184 fig. 17.

rather imaginative and artistic forms, such as the stialas from Graubiinden which have become much-coveted collector's items (cf. n.29 above; here fig. 16a-b). Rather less studied are the social attitudes related to such 'proud' tallies, but it may be clear that their function, or that of any tally to be sure, cannot be reduced to being expressions of social inequity, dependence or debt. Literary sources from mediaeval and later Europe add to this impression. We have ~ready referred to the uniquely designed split tallies (and the longer Zehentstecken) operated m t~e tithe assessment in early nineteenth-century Lower Austria (cf. §1.2). The solemn and surpnsingly complex ritual shows that the tallies with their carefully incised c~des became argu.ments in a performative act_ they tangibly confirmed and vouched for the fatrness ?f the credtt relation just established and simultaneously gave physical reality to tenant farmers self-esteem and rights, as well as to his place in the social hierarchy. mit Augerungen des Rechtslebens zu run, die man sinnlich wahrnehmen ~nd. mit .~-Ia~den grei.fen kann; und nicht nur in Gestalt von Tinte und Papier stehen sie vor uns, sondern m etgentumhchen .zeJChe.n u~d merkwiirdigen, oft sogar kiinstlerisch angehauchten Holzformen, welche ebenso unmodern w1e gehetmmsvoll anmuten."

'

WF.M.

HENKELMAN

and M.L.

Considerations on the ritual and performative roles of the (split) tally in local communities in Austria and elsewhere in nineteenth- and early twentieth-century Europe, hold a fortiori true for the Exchequer, an institutional environment that had to represent the English Crown and its vision of the world. The Exchequer tallies are not only highly standardized, but also unmistakably made to be elegant, which is unsurprising given the status of their holders (sheriffs and other men of high responsibility) and the elaborate ceremony of which they were the centrepieces. Tallies were essential expressions of the bond between the Crown and those who made its functioning possible. Certainly in comparison with the Receipt Role and other parchment documents simultaneously kept, their physicality is striking, as is the number of people who handled them at various points before and during the audit proceedings. 100 It would not be overstated to say that the tallies, after which the Exchequer was originally named, were the cornerstone of the whole institution, together with the chequered cloth or scaccarium, after which the Exchequer had been re-named at the time of the Dialogus de Scaccario. It is worth mentioning in this context that not only did the checkerboard cloth, on which the arithmetic was done during the audit, resemble a chessboard, but the ritual itself was explicitly likened by Richard fitzNigel to a game of chess with the Treasurer and the sheriff as opposing kings. Not without a sense of drama he even compared the ordeal, in which tallies played a very visual and weighty role, to the Last Judgement "when the books of all are opened." 101 Whereas we are not obliged to assume that the representatives of the King of Kings went as far as their later English colleagues, it is still important to keep in mind that the Bactrian split tallies also have a physical reality beyond their documentary function. As split records, they established a tangible relation between the institution and its partners, be it tenant farmers, private debtors, or contract herdsmen. It is not hard to imagine a certain ritualistic solemnity with which the institution would have sought to inspire the sealing of even small-scale agreements. That would be the case with herds audited at livestock stations before and after consignment, but also with amounts of seed or silver lent to individuals, undoubtedly framed in the pervasive ideology of gift and counter gift. Aramaic and Elamite documents on leather and clay obviously had their own reality and could be drafted and handled in intricate ways. The carefully-written documents in the Arsama correspondence, with the satrap's outstanding seal meticulously impressed on highly-refined buff-coloured clay, are a powerful reminder of the Achaemenids' attention for their imperial stationery. The reverence felt for royal letters and their seals is well known in classical sources (Hdt. III.l28.2-5, Thuc. I.l29.1, Pol. VII.21.5; see Briant 2002, 344£, 661) and a case can be made for their role as "objects of power" in communicative acts more intricate than the 'simple' exchange between a sender and a reader (Alien 2013). Yet, even such eloquent leather, clay or parchment documents did not 'perform' in the same, fUndamentally mutual way as the 100

Wedell (20 11, 278-288; also ibid. 245-259) has pointed out the performative aspects of the use of tallies and their 'code' in the very intricate Exchequer ritual. See also Kuchenbuch 1999 and 2003, for comments on the Sitz im Leben of tallies in European history. 101

to

YouR TALLY IS FuLL!

FoLMER

Dialogus de Scaccario I.1 and 1.5 (pp.10, 38 Amt). On the chequer pattern of the cloth and the allusions the chess game see Jones 2008, 470; Church 2012.

177

split tally. Arid though the status of the first group as objects desperately needs mo~e a_ttentio~, it remains at the same time true that they hardly ever reach the same level of phystcahty. Their regularity, recognisability and appeal equally bear a hint of ritualistic behaviour, ~ut are ~t t~e same time aimed at different audiences. The orderliness of the Achaemenid Bactnan tallies, m order words, gains meaning from the fact that these were instruments with which a state institution reached out beyond its proper limits.

2.8. The Achaemenid Bactrian tallies: an interim summary In the preceding sections (§§2.1-7), we have argued that the thesis of ration accoun~s, even in the sense of vouchers, is not the most attractive interpretation of the Achaememd Bactrian tallies. At the same time we have indicated possible directions in which an alternative interpretation may be sought. Before we turn to the texts on the tallies, it may be useful to retake the main arguments and present them in the context which has been discussed in the earlier survey of the documented uses of (European) split tallies (§§1.1-3). First, the institutional character of the tallies deserves repeated emphasis. The unity of their format, but also the fact that always the same half of the tally stick was kept by the administration, undoubtedly answered to the same kind of detailed protocols that ruled the very complex Persepolis Fortification archive. To this should be added the regularity in the shap~ of the tallies and the care with which they were evidently made (apparently from the same kmd of wood). Early modern and modern parallels in the use of split tallies alert us to the fact that such 'official' tallies were probably meant to be treated with a certain respect, if not handled ceremoniously- after all they indirectly represented the king's sway over the lands. We have seen that split tallies occur in a variety of circumstances, but all of these have in common that they relate to some kind of credit situation, be it a postponed payment of salary, a postponed payment for delivery or acquisition, an assessment or advance payment ~f tax obligations, a deposit or temporary assignment, or a straightforward (silver) loan. Receipt of regular rations is not a context in which split tallies are commonly found, nor does it present a circumstance in which the instrument would be particularly necessary or useful. It is therefore unsurprising that the BA tally mark sequences seem erratic from the point of view of ration payments and do not agree to known scales, whether one uses a conversion by *griva- or by artabe. Naveh and Shaked have, in their earlier 2003 study, proposed a slightly different scenario, in which the tallies correspond to running accounts of flour (or another commodity) against which the workers or soldiers could draw the quantities they needed - until their allowance was depleted (c£ §0 above). Here again, the BA sequences would present inexplicably errat_ic amounts. Moreover, it would be difficult to see why the authorities in Bactria would engage m a relation that was essentially credit-based with simple workers; in other contexts, such people were not even accounted for on individual documents. Finally, the proposed scenario would

.,...._

W.F.M.

HENKELMAN

and M.L.

YouR TALLY IS FuLL!

FoLMER

put the administration in the role of the debtor, which is unlikely given the contents of the inscriptions (c£ §3.1 below). Various factors plead against the idea of running accounts in general. The circumstance that the length of the tallies correlates undeniably with the number of tally marks suggests anticipation. The same is true for the observation that a 'line' is usually closed with a sequence of C and D marks. The neatness with which the tallies are notched is also an argument, as it would not have been very easy to achieve this after splitting, also considering the format of the Achaemenid Bactrian tallies (split evenly, without leaving a socket or stump). These more general observations would also plead against a moderately adapted version of the scenario proposed by Naveh and Shaked: one might suggest that, instead of a kind of ration credit, postponed payment of salaries was the raison detre of the tallies. If so, the BA sequences could indicate units of some commodity harvested or numbers of items manufactured by individual workers, whereas the DC sequences could be the conversion into silver to be paid (or the number of days worked, etc.). Apart from the circumstance that this would again make the tallies into records of running accounts, which seems unlikely, there is also the point that it would make the administration debtor, which is countered by the inscriptions (cf. §3.1 below). The inscriptions additionally show that some of the recipients/debtors in the Bactrian file had two or more tallies for the same year, which does not support the idea of work for hire (c£ §1.2 above). And, finally, there is the small but not unimportant detail that the Achaemenid Bactrian tallies do not have holes that would allow them to be bundled on a strap or ring held by the tallyman overseeing the harvest (vel sim.) -as is very common in other contexts, especially in cases of postponed payment of salaries. In other words, even before studying the inscriptions in detail, it is becoming clear that the meaning of the Achaemenid Bactrian tallies will, if anywhere, have to be found in the sphere of loan, deposit or assignment. With this, it has to be born in mind that the tallies are unlikely to have regulated a transaction between the institution and individuals belonging to its internal sphere, viz who were within its control and with whom it did not need to enter into credit-based relations. Rather, the literature on the use of split tally sticks (§1.2) suggests that the device was commonly used between parties of different background or status who were not otherwise closely connected. What the various types of tally marks represented is difficult to guess. We have found no arguments to support the idea that it was flour or grain, nor any that would rule out this possibility. Though the analysis of the texts yields some possible clues, it remains difficult to assess what was the object of the transactions concluded with the help of the tallies, and, by consequence, who were the people involved in it on the receiving side. The intactness and date (3.Dar.III) of the tallies invites speculation as to this regard, but will not likely produce unequivocal answers.

3.

179

AcHAEMENID BACTRIAN TALLIES: ARAMAIC INSCRIPTIONS

The Achaemenid Bactrian tallies are inscribed, like the Exchequer tallies, but unlike some of the Swiss tallies as we have noted earlier (§ 1.2). Script was a possibility, but not a requirement for the use of the split tally stick. In fact, it has often been noted that its most revolutionary aspect is precisely that it allows illiterate people to conduct credit agreements. Robert, for example, speaking about mediaeval England, defines the fact "that this tangible wooden object, with its careful scoring, was a comprehensible record to the untutored men of that distant past" as the fundamental raison d'etre of the tally (1956, 77). When split tallies speak by the means of writing, it is because their tally marks cannot be understood or function without explanation. Nineteenth-century Swiss farmers who met each other on a daily basis, had no need for such elucidations on their mutual tallies; even though literacy was relatively widespread in their world, they often would eschew writing in handling their communal arrangements, such as the partition of grazing rights per tally. Novgorod merchants, by contrast, who would engage in trade with various parties and often at various places, needed to be able to connect each of their tallies to individual transactions (cf. §§ 1.2, 5.8). In fact, they often kept books besides their tallies, which they would show as collateral, in case of legal dispute. Script moreover made their tallies transferable and allowed them to be sent over longer distances. Institutions, too, if they handled tallies on a regular basis, needed inscriptions, not only to distinguish between various accounts, but also to allow all clerks and administrators to independently 'read' the tallies. Yet, even when script is used, it does not immediately follow that all parties involved in the agreement are literate. What constituted important information for the institution, would matter less for the contract herdsman who received a half tally with the herd entrusted to him; what was essential knowledge for the professional money-lender, was painfully evident to the peasant who took a small loan from him. This potential asymmetry on the level of literacy connected to split tallies may find physical expression in the absence of an inscription on one of the tally halves (though we are rarely in the position to check this, as normally only one half survives). Given the above, inscribed tallies were no more 'exclusive' than uninscribed ones. Their essence remained the sequence of tally marks through which the split ran and which was intelligible to all. Moreover, literacy was not the only element the tally could dispense with: proficiency in a communal language was another. When dealing with the Achaemenid administration, with its bi- or trilingual and often bi-literate scribes, it is easy to overlook such factors, but for the proper understanding of the Bactrian tallies they may be vital and should be kept in mind. All eighteen split tally sticks in ADAB have writing in Aramaic script written in dark ink on the flat inner surface. Only in a few cases, the convex exterior is inscribed as well. Since it appears (c£ §2.3) that the preserved objects always represent the same half of the split stick, it cannot be established whether the lost other halves were inscribed as well. It is true that, often, only the stock, the part kept by the creditor, was inscribed (c£ Wyffels 1988, 25), but the Ex-

I8o

WF.M. HENKELMAN and M.L. FOLMER

chequer foils, to name one counter example, had the same inscription as the stock Qenkinson 1925, 294). In the following, we will offer some comments on the general terminology used in the inscriptions (§3.1), on the occurrence of the expression "second tally" (§3.2), and on possible textual evidence for silver credits (§3.3).

3.1. General terminology The flat surface of most of the Achaemenid Bactrian tallies contains the following three elements, which - if present - appear in fixed and invariable order: 1. 'm PN 1

"with PN 1"

(always present)

2. mn PN 2

"from PN 2 "

(usually present)

3. hint 3 dryhwf mlk'

"in year 3 of king Oarius"

(usually present)

The date is lacking only once (0 18); in one other case it is broken (06). The preposition mn is clearly more dispensable, as it is omitted in five cases. If organized by allocator (presumably the creditor) and recipient (debtor), the texts show a clear distribution in its presence or absence (proper names after ADAB, 57-60): 013 'm 'dw'p

Aduapa?

012 'm nyk 07 09

'bwdy

Abudi??

Nayaka?

mn 'bwdy

Abudi??

3.0ar.

'm ptyr

Patiyara

mn 'bwdy

Abudi??

3.0ar.

'm ttk

Taitaka??

mn 'bwdy

Abudi??

3.0ar.

0

3.0ar.

'm tyry 010 'm wknw

Tir(a)ya

0

'bwdy

Abudi??

3.0ar.

Vikanava(ntf

0

'bwdy

Abudi??

3.0ar.

011 'm frtwk

Caratuka??

0

08

prtn 'bwdy Paritana +Ab.?? 3.0ar.

05

'm bgyf

Bagaica

mn gwzh

Gauza

3.0ar.

02

Bagayaza

mn gwzh

Gauza

3.0ar. "white ... " (§3.3)

04

'm bgyz 'm pftrk

XSaSraka

mn gwzh

Gauza

3.0ar.

03

'm srybr

Saryabara

mn gwzh

Gauza

3.0ar.

06

'm spyt rhby Spaita(-)rhpj mn gwzh Taitaka?? 'm ttk mn gwzh

Gauza

[3.0ar.]

Gauza

3.0ar. "second (tally)" (§3.2)

Ax•apavya??, mn 'prwn Kaufutaca??

Uparivana

3.0ar.

01

014 'm 'fJpwy

qwpptf 016 'm npwzt

Nafavazata

mn 'prwn

Uparivana

3.0ar.

015 'm ttk

Taitaka??

mn 'prwn

Uparivana

3.0ar.

017 'm Itrykr

CiSriyakara

[(illegible traces)]

018 'm pftrk

XSaSraka

0

0

3.0ar. 0

YouR TALLY rs FuLL!

I8I

The occasional absence of mn seems to be a trademark of 'bwdy or his secretary and may therefore be a scribal habit. The lack of mn was apparently no problem in understanding what the tally was about. In 018 (and perhaps 0 17), the information mn PN 2 is totally absent, perhaps suggesting that this phrase was not crucial to a proper functioning of the tally. 017 is problematical in the sense that part of the text is illegible and "seems to have been deliberately rubbed ofF' (ADAB, 257). 018, perhaps not coincidentally the shortest of the eighteen tallies (cf. §2.5 above), has no date, no mn and no allocator (bftrk also occurs in 04). Apparently these elements could, if necessary, be dispensed with, though usually they were not. To resume: 'm+ PN 1 had to be mentioned in any case, the date and PN 2 preferably too, but the preposition mn itself was not really necessary because the sense would be clear anyway.

It may be remembered that inscriptions on tallies generally tend to be very terse and formulaic, hence not, necessarily, grammatical. One could even say that their terseness is a reverse indicator of the openness of the context in which the tallies operate. Commodities are not mentioned, except, plausibly, in one case (cf. §3.3. below), nor is the term of the agreement. The dates are only by regnal year, not by month. And, finally, only four names occur with the preposition mn (PN 2). Taken together, this suggests a relatively closed and well-defined environment in which agreements with a regular term (which may have been one year, as argued in §2.1) were concluded between two parties, with, on one side, a range of different individuals and, on the other, only four people, presumably administrators. They were handling a single type of commodity, which therefore did not normally have to be specified, and they knew the procedure well enough to use terse if not elliptical language when issuing inscribed tallies to seal the agreement. Since the tallies presumably record disbursements of some commodity, it is likely that the prepositions 'm (1) and mn (2) precede the names of the parties involved in such a transaction.102 Of the two elements, the prepositional phrase mn PN 2 is easiest to understand. There cannot be much doubt that it refers to the person who hands out the commodity involved, hence the supplier, allocator or creditor (or a combination of these). The preposition often indicates spatial separation and means "away from" or "originating from"; it is frequently used to indicate the source or origin of objects and persons. This is probably how the prepositional phrase mn PN 2 on the tallies should also be understood: "from PN 2" or "originating with PN 2." 103 In Aramaic letters, the sender is indicated in the same way (mn PN 2). Several examples can be found in the corpus of Aramaic letters in ADAB: mn 'l;mzd, "from Axvamazda" (A1-A6) and mn bgyf, "from Bagaica'' (B2-B4). The phrase mn PN 2 being more or less secured, it may logically be assumed that 'm PN 1 refers to the person who received the commodities, i.e. the recipient, utilizer or debtor (or a combination of these). The phrase 'm PNJ> "with PN 1," could not be dispensed with and is mentioned first. This is in full agreement with the use of split tallies in an institutional envi102 10 3

The function of rhe two prepositions was also discussed by Naveh/Shaked 2003, 113-114. DNWSI, 650, s.v. mn 5: mng. la-b, "coming from," "from."

WF.M. HENKELMAN and M.L. FoLMER

YouR TALLY rs FuLL!

ronment: the institution will dominate the form and handling of the agreement. It knows itself and has alternative ways to organize documents when they do not state the date or the name of the responsible officer. What cannot be omitted is an identifier of the person it is conducting a transaction with; in the absence of sealings, this has to be a personal name written out in the inscription. The other piece of information, it should not be forgotten, that is indispensible, is the amount involved- the tally marks vow for that. The 'm PN 1 phrase, then, seems to convey the most crucial written information on the tally. The basic and most ascertained meaning of the preposition 'm is "with." 104 One way to explain the use of 'm in the tallies is to assume an underlying nominal clause or a verbal clause with a verb of motion, such as 'zl, "to go." The essential information (tally marks + 'm PN 1) could then be understood as:

"because," abstract "in addition," and, by semantic extension, directional "to" (with animate complements). 106 In Aramaic texts from the Achaemenid period the preposition can also be used to indicate an indebted person ("incumbent upon"). This abstract meaning is related to the locative meaning "on" of the preposition, one of its central meanings. All relevant sources are from fifth-century BCE Egypt (Elephantine). The earliest example is found in a text from Elephantine datable to the early fifth century BCE (TAD B4.1). According to Porten and 107 Yardeni this text may be a fragmentary receipt or a cancellation of debt:

!82

TADB4.1 :2 ytly 'Lyk ksp wb.tn wi'rn[ wkntn w'mr w]ktn I have (a claim) on you (for) silver and wheat and barley [and emmer and wool and )linen

xx (items/units) went ( 'zl) with ('m) PN 1

TADB4.1:4

xx (items/units) are with ('m) PN 1

ytl(y ']lykm md'm mkl

It is well possible that such a clause appears in an elliptic manner in the tallies: the verb is absent and the subject (the commodities) is not explicitly mentioned (but see §3.3 below). This is very reminiscent of the internal and external address formulae of letters, which - in most of the cases - consist of two prepositional phrases as well. 10 5

[I) have (a claim) [o)n you (for) some food

Similar use of 'lis found in two deeds of obligation from Elephantine (end fifth century BCE): TAD B4.5:3£ (407 BCE)

'yty lk 'Ly ksp ...

In the same manner, mn PN 2 can be understood from an underlying nominal or passive verbal clause:

you have (a claim) on me (for) silver. ..

xx (items/units) were given/supplied/disbursed (yhyb) by (mn) PN 2

TAD B4.6:3 (400 BCE)

xx (items/units) originate with (mn) PN 2

'yty lky 'Ly ksp I 2 you have (a claim) on me (for) silver, 2 sh(ekels)

Though, simple as this may seem, the suggested meaning of the preposition 'm is not without complications and requires further investigation.

In a loan of silver from Elephantine ( TAD B4.2), 'I is found with again a very similar meaning:

Since we explore here the possibility that the tallies are records sealing a (credit) transaction, we should examine whether 'm can have the specialized meaning of 'incumbent upon.' Is it possible that the preposition is used to indicate the debtor in a credit agreement? In other Aramaic texts from the Achaemenid period, 'm is not attested in this meaning: instead, the preposition '/, "on, upon," is found. The use of the preposition 'm in contexts comparable to ours is only known from Aramaic texts datable to the period of Middle Aramaic, a fact also noted by Naveh and Shaked (Naveh/Shaked 2003, 113-114; ADAB, 233).

TADB4.2:2

wyrbh 'Ly ksp interest will accrue from me (at the rate of) silver TADB4.2:8

whn /' Ilmt lk kl kspk wmrbyth ... y'qp kspk wmrbyth zy yJt'r 'Ly wyhwh rbh 'Ly yr[l lyrb 'd ywm zy 'flmnhy lk and if I do not pay you all your silver and its interest .. . your silver and its interest which remains from me will double and be accruing (interest) from me month by month until the

The meanings of the preposition '/ in Aramaic texts from the Achaemenid period are legion. Among the more frequent ones, one finds directional "against," locative "on," causal DNW.C'l . h" . h ," "near"; 3 . "wit . h" . h regard J. , 867 - 869 , s.v. 'm4 : 1• "w1t , "rage th er w1"th" ; 2 . "wit , "towards, '"'wit to"; 4. "with," "incumbent upon." The meaning "incumbent upon" is attested in texts from the period of Middle Aramaic onwards Qewish Aramaic and Nabataean Aramaic).

day that I pay it to you

104

105

We do not address here the variety in the use of prepositions and the order of addressee and addressor. On the address formulae in Aramaic letters from Bactria, see Folmer forthc. For a full treatment of formulae in letters, see Schwiderski 2000.

106

This is not an exhaustive enumeration of the semantic range of the preposition '/. See for more details

DNWSI, 844-850. s.v. '/.Often '/assumes meanings primarily linked with other prepositions, such as benefactive "for" (normally expressed by [). The preposition '/in the meaning "to" is another example. In older Aramaic this meaning was primarily expressed by '/ .See Folmer 1995, 589-629. 107

TAD B, 102. See also Porten 2011, 257, n.15 .

·..J

ut

YouR TALLY IS FuLL!

WF.M. HENKELMAN and M .L. FoLMER

Literally, the Aramaic of the last passage says: "it will increase on me" (1.2, wyrbh 'ly and 1.8, wyhwh rbh 'ly). In this case, as in all the preceding cases mentioned, the preposition 'l gives expression to the obligation of payment, which rests on the shoulders of the debtor. The above-mentioned document TAD B4.2 from Elephantine is also of relevance to another aspect of our discussion. The document was written by the debtor Gemariah bar Ahio himself. In line 16 of this document, he is mentioned as the scribe: 108 ktb spr' gmryh br 'fJyw '[pm Jhdy' zy '[ zpr' z nh

texts from the Judean desert (early Jewish Palestinian Aramaic) and in Nabataean Aramaic. The expression used is yty + l with pron. sf. + 'm with pron. sf. (in negated clauses L' yty + l with pron. sf. + 'm with pron. sf.). In these clauses, the pronoun combined with l indicates the creditor and the pronoun combined with 'm refers to the debtor (in the Nabataean inscriptions, however, the debtor precedes the creditor). The literal translation is "there is to PN 1 with PN 2 ," meaning "PN 1 has a claim on PN 2 " or "PN 2 owes PN 1." 11 2 1he following examples are from (I) early Jewish Palestinian Aramaic and (II) Nabataean Aramaic. In early Jewish Palestinian Aramaic, the formula is best preserved in Na{Jal $e'elim no. 13, but it can be reconstructed with some plausibility in several other texts:

Gemariah son of Ahio wrote the document upon the instruction of the witnesses who are on this document

I. Evidence from early Jewish Palestinian Aramaic In the same document it is stated that the creditor is obliged to write a receipt for every (monthly) repayment of the debtor:

~e 13:3-6 (waiver of claims; 134/135 CE; text Yardeni A, 134; transl. Yardeni B, 57)

3. l' yty[ ly] 4. 'nh Jlm1yn brt yhwsp qbJn 5. mn 'yngdh 'mk 'nt 'l'zr br &nny[h] 6. dy hwyt b'lh mn qdmtdnn I] do not have, I, Shelam~ion daughter ofYehosef qbJn from 'Ein Gedi, with you(= you don't

TAD B4.2:6-7 wtktb ly nbz '[ kl 7 ksp wmrby zy 'hwh mJlm lk

owe me), you 'El'azar son ofHanania[h], who had been my husband from before

and you shall write me a receipt (nbz 109) for all the silver and interest which I shall be paying you

~e

13:8 (text Yardeni A, 134; transl. Yardeni B, 58)

8. i['] yty ly 'mk '[nt] 9. 'l'zr 'i fbt kl md'm

In conclusion, from other Aramaic texts from the Achaemenid period it is not self-evident that the preposition 'm was used to indicate the debtor, since all the existent evidence from this period points to the use of 'l in this specific situation. However, the relevant documents from Elephantine are 50-150 years older than the documents from Bactria and the language may obviously have changed, even considerably. Prepositions are words that can easily undergo semantic change 110 and meanings may be added by semantic extension. In the case of 'l, another factor may have been of importance: in the Achaemenid period, 'L had taken over many meanings of other prepositions (in particular of 'l, "to"). It is imaginable that the use of 'm rather than '/in the sense of "incumbent upon," as likely for the tallies from fourth-century Bactria, resulted from a process of disambiguation of the preposition 'f.

I do not have with you(= you don't owe me), y[ou], 'El'azar, concerning anything ~e 24a: 1 (fragment of a deed; late first or early second century CE; text Yardeni A, 80; transl.

Yardeni B, 37)

] yty ly ywm' dnh '[mk I have/I [do not]have this day with y[ou ... (=you owe me/you do not owe me) 4Q344:2 (fragment of a loan bill; first century CE?; text Yardeni A, 21 ; transl. Yardeni B, 20)

] 'my 'nh 'l'zr br hsp[ with me, I, 'El'azar son of hosef[ ('El'azar is both writer and one of the parties, the creditor, in the deed. 113)

In the western parts of the former Achaemenid empire, we see that in texts from the middle of the first century CE onwards, it is the preposition 'm, and not '/, which is predominantly used to indicate the debtor of a loan or some other liable person. 111 In Nabataean Aramaic, the two prepositions seem two have co-existed in similar environments, which is not unimportant for the evaluation of the use of 'm in the Bactrian tallies. There is evidence in Middle Aramaic

~e 12:6 (receipt for dates; 131 CE; text Yardeni A, 76; transl. Yardeni B, 36)

6. dy .. 'mk mn bb lwy 7. 'bwk (reading uncertain) which you owe (?), from .. . (of?) Levi your father (The particle dy should be construed with tmryn, "dates," mentioned in the preceding line. 114 )

108

See also Porten 2011, 256.

109

This word appears as Aramaic loan in the Neo-Assyrian and Neo-Babylonian dialects of Akkadian (nibzu "document," "receipt"; see Kaufman 1974, 77). The word is found in several dialects of Aramaic (Samaritan Aramaic, Mandaic, Jewish Palestinian Aramaic and Christian Palestinian Aramaic; in all these dialects with the meaning "lot") and it may be related to Biblical Aramaic ;,~p~ (Dan. 2:6) (Kaufman 1974, 77). See also HALOT, 1742: "Dokument, Urkunde." On the possibility of TAD B4.1 being such a receipt, see above. 110 See Luraghi 2003 and Bortone 2011 for prepositions in Greek. 111

This was also noted by Naveh/Shaked 2003, 113, " 'my in the sense of'I owe." '

112

See also DJA, 71, s.v. Ol1 b: "to owe" (in combination with the preposition'(). His signature as one of the contract partners is found in 1.6: 'l'zr br hwsp 'l npJh ktb[h, "'El'azar son of Yehosef, for himself, he wrote [it] ." See also CottonNardeni 1997, 290-291. For a different interpretation see Yardeni B, 20 (Yardeni combines ktb with the following) . 11 4 Same interpretation in DJA , 71 (dmy tmrn .. . dy 'mk, "the price of the dates which you owe me"). 113

r86

WF.M.

HENKELMAN

and M.L.

FoLMER

YOUR TALLY IS FULL!

Mur 18:3 115 (loan bill; 55 CE; textYardeniA, 17; transl. Yardeni B, 19)

Tomb inscription from Hegra (H36; 31132 CE; text and transl. Healey 1993)

3. 'my 'nh zkryh br yhwfmn mn ...... 4. ytb bkslwn ksp zwzyn .. .

7. wmn y'bd k'yr dnh pyty 8. 'mh qns ldwfr' 'lh m[r'n' ks]p sl'yn pmf m'h prty 9. wlmr'n' kwt and anyone who does other than this will be liable for a fine to Dushara the god of our lord in the sum of five hundred Haretite se/a's and to our lord for the same amount

(you have) with me(= I owe you), I, Zekariah son ofYehoi).anan, from ... [... ], residing in Kesalon, silver, zuzin

It is interesting that in this last text, a loan of silver is concerned. Tomb inscription from Hegra (H34; 71/72 CE; text and transl. Healey 1993)

6. [wmn dy y] 'bd k'yr dy '1'7. pyty '!why kpl dmy 'tr' dnh 8. klh wl'nt dwfr' wmnwtw [and whoever d]oes other than what is written above shall be liable for double the price of this whole burial place and for the curse of Dushara and Manotu

?e 8a:5 (deed of sale of house; 135 CE; text Yardeni A, 71; transl. Yardeni B, 34)

wrfh l' yty lk {lk} 'my bgw drth dk and you

«yow>

have no right(s) with me in that courtyard 116

This text can be paraphrased as 'you are not authorized to claim from me that courtyard.' 11 7

B. Examples with the preposition '/:

?e 8a:9 (c£ above; text Yardeni A, 71; transl. Yardeni B, 35)

Tomb inscription from Hegra (Healey 1993, H34; 71/72 CE; text and transl. Healey 1993)

9. wrfh l'l 0. yty lk 'my bgw drth dyly wl' m 'l wl ['] mpq 'ly

10. wmn y'bd 11. k'yrdy '!' dy yty '!why b!J'h 12. ldwfr' wmnwtw ksp sl'yn 'lp bd prty 13. wlmr'n' rb'l mlk nbfW kwt

and you have no right with me within my courtyard, and no (rights of) entering or going out upon me 118

Il Evidence from Nabataean Aramaic A. Examples with the preposition 'm: 11 9

!jev 1 (dowry settlement?; promissory note?; 94 CE; text Yardeni A, 271; transl. Yardeni B, 87) 12.ywm' hw 13. yty lky 'nty mtysy brt kmnw br 'mrw 'my 'nh 14. mqymw b'lh br 'wt'lhy br blp'lhy dy 'mr b'yn' fprw[ dy 15. bm'wb [ks]p sl'yn m'h bdh wpmfyn that day you, mtysy, daughter of Kmnw son of 'mrw, you have with me, (= I owe you), I, Mqymw, her husband, son of 'wt'lhy son of !jlp'lhy, who resides in 'yn' fprw[ which is] in Mo' ab, [silv]er sl'yn, one hundred and fifty Kam Kam tomb inscription from Hegra (1 BCE/CE; text and transl. Healey 2009, 68, no. 7 120)

7. wmn dy l'y'bd kdy '!' ktyb pyty 'mh 8. ldwfr' whblw wlmnwtw Jmdyn 5 and whoever does not act according to what is wricren above shall be liable to Dushara and Hubalu and to Manotu in the sum of 5 shamads

and whoever does other than what is written above will be liable for a fine to Dushara and Manotu in the sum of one thousand Haretite se/a's and to our lord Rabel, King of the Nabataeans, for the same amount

The above-mentioned texts in Jewish Palestinian Aramaic and Nabataean Aramaic demonstrate that during the period of Middle Aramaic (200 BCE-200 CE) the preposition 'm was used to indicate the debtor/obligator, not only in legal documents in which financial and other claims are recorded Qewish Palestinian Aramaic and Nabataean Aramaic), but also in Nabataean tomb inscriptions, in which the ownership and usage of the tomb were regulated. In this same group of texts the terminology used is still legal. The evidence from the tomb inscriptions is of particular value for our discussion of the Bactrian tallies, because it demonstrates that the two prepositions, 'm and 'l, could not only co-exist within one and the same dialect of Aramaic, but even could be used within one and the same text, in a similar clause and without an apparent difference in meaning. In all the clauses liability to a high placed person or deity for a certain amount of money is concerned. A good case in point is tomb inscription H31 from Hegra, in which the two prepositions eo-occur in two similar clauses (ll. 7, 11). This use of 'l in some of the tomb inscriptions is clearly an archaizing feature of the Nabataean language. As far as this specific meaning is concerned, the preposition 'm outnumbers the preposition 'l in these inscriptions.

11 5

C£ Yardeni A, 15 and Yardeni B, 19. The text is also quoted in DNWSI, 869, s.v. 'm mng. 2 "with, near," but the reading and interpretation of the text are superceded. 116

C£ also Healey 2009, 131, "but there is no authorization for you against me inside the said courtyard." Healey comments: "'my, literally 'with me', effectively 'in connection with me', 'over me,' 'over my rights.' Effectively 'no claim on me.'" (Healey 2009, 133).

3.2. "Second tally'' In most of the tallies only the flat interior surface was used for writing. There are, however, three tallies that also have writing on the convex exterior side:

117

The courtyard is not sold and remains with the seller of the house. See also the interpretation in Cotton/ Yardeni 1997, 35.

D1

118 Healey 2009, 131, "and you will have no right against me inside my courtyard, and no entry or exit over me (= my property)."

D2

For more instances in tomb inscriptions from Hegra, see Healey 1993, nos. 1:7; 5:9; 9:8; 11:4; 12:9-10; 16:7-8; 19:7; 28:5-6; 30:6; 38:7-8. See also Healey 1993, no. 31.

bwr bnw 3 blp Jpdn three white (and) resplendent [items not specified] in exchange for gifts (so ADAB, 237)

119

120

tnyn second 121

121

ADAB, 233: "a second (tally).

YouR TALLY Is FuLL!

WF.M. HENKELMAN and M.L. FoLMER

188 D 17 drywhwJ mlk' (of) Darius the king

The information in 017, part of the dating formula, seems to have been added in another hand, and perhaps also with another pen. The f on the interior surface looks different from the f on the exterior. 122 It looks therefore as if the royal name was added by a second person, perhaps an accountant or archivist. 123 The information tnyn (O 1) and bwr bnw 3 blp Jbdn (02) are both of importance to our discussion and will be examined here and in the following section (§3.3). The word tnyn (01) reflects the ordinal numeral "second" (sg.m.). Ordinal numerals are rarely attested in Aramaic texts from the Achaemenid period. The ordinal numeral tnyn is found in an Aramaic legal document from Elephantine. The word order in this text is unusual (ordinal numeral preceding the head noun) and there is disagreement in gender between the ordinal numeral (m.) and the noun fnh (f.): 124 TAD B3.1:7 (C 10) m(tnyn Jnh

a second year has arrived

In Biblical Aramaic, there is evidence for the expected word order with ordinal numeral tnyn (1. head noun; 2. subordinated noun [ordinal numeral]): Dan. 7:5

n?Ji?w ,,T.,~i¥'71 :i1'7 :-~;~1 :-~~;~1:1 'lr!t$ :-~ro 11~1 and look, another beast, a second, like a bear; and it was raised up on one side

In 01, the ordinal numeral tnyn is not accompanied by a head noun; its referent remains implicit and must therefore have been evident to the users of the tally. An explanation of the term as a self-designation therefore seems attractive; it is probably for this same reason that Naveh and Shaked understand tnyn to refer to a second tally (ADAB, 234). The occurrence of organizational information is in itself unsurprising in archival contexts and known from the Persepolis Fortification archive. Journal (register) text PF 1940, for example, concludes with the phrase tuppi hi 2-memanna bel15-na, "this is the second tablet of year 15" (1.29) and the small label PF 1920, undoubtedly once affixed to a larger tablet, merely states halmi 2-mena, "(this is) the second sealed document/authorization." In the case of split rallies, such information is unlikely to relate to archival procedure (second rally in a deposited

file), nor is there any reason to assume that "second" refers to the second half of the tally. Rather, the expression "second (rally)" is meant to alert to the fact that the debtor had more than one valid agreement or running account at the same time. The names of the recipient or debtor, ttk (Taitaka?), and that of the supplier or creditor, gwzh (Gauza), occur in other tally inscriptions (ttk in 09, 015; Gauza in 01-06), but not in the same combination. It is less likely that the mention of a second (tally) implies reference to any of these other existing tallies with ttk. Rather, Gauza had already issued a tally to ttk, which is no longer extant (or at least not in the Khalili collection), and 01 is the second one for the same year. One other recipient, :XSa6raka (bftrk), occurs more than once: with Gauza in 04 and without mention of the allocator in 018. 125 In this case the additional information "second (tally)" is not given. The use of "second (tally)" does not agree well with the idea that the tallies were part of a ration system (cf. §2.8 above). In fact, the recurrence of the name ttk as recipient or creditor in more than one rally dated to the same year already strongly pleads against this idea. 126 If the tallies served as a kind of vouchers of accounts from which a worker or soldier could draw amounts of flour or grain, they would have been destroyed (and replaced) after the account was depleted. The occurrence of "second (tally)" makes this problem more poignant, as it underlines the simultaneous validity of more than one account or agreement. It is hard to see why any one individual would have multiple tallies for flour allocations at the same time, unless he were a middle man - but that would make him part of the administration and render the use of split tallies redundant. In the dealings between the recipient/creditor and the administration the tag "second (tally)" clearly served as an operational aid: when presenting his half of the tally, ttk had to indicate (if it was not written on his half) that it was the second of his rallies, upon which Gauza would look for the tally inscribed accordingly. The tag thus neutralized the potential for confusion and conflict that could have resulted from matching two non-fitting halves.

3.3. Indications for silver in D2? There is only one rally that seems to break the silence as to the commodities or items involved in the agreements (cf. §3.1 above). Both the reading and the interpretation of the text on the convex side of 02, are ambiguous, however. Naveh and Shaked read and translate it as follows (ADAB, 237): pwr bnw 3 pip Jpdn three white (and) resplendent [items not specified] in exchange for gifts

122

Nothing can be concluded from the r, which also occurs on both sides.

The remark "written in the opposite direction" is probably a mistake (ADAB, 256) . 124 Alternatively, the text has been explained as "(when) the same time next year arrives" (tnyn as a noun, "repetition," as first element in a construct noun phrase; see DNWSI, 1223, s.v. tnyn 2). 123

125

A very similar name, pJtrkn, occurs in C4:41 as the patronym of wbfw'brdt, Vaxsuabradata, who is involved in the supplying of goods. For an analysis of the text see Tuplin forthc. 126

The same is probably true for XSa6raka, but as D 18 is undated we cannot be sure in this case.

WF.M. HENKELMAN and M.L. FoLMER

The orientation of the writing on the curved surface on 02 differs from the orientation on the other two tallies which have writing on their curved surface (0 1, 0 17). 02 was turned around its vertical axis, and the writing on the convex (reverse) side is aligned to the left edge (i.e. to the same end as the text on the obverse; cf. ADAB, 237 and §2.3 above). Some textual comments are in place here: f;wr The word f;wr is a well-known Aramaic adjective with the meaning "white" (see below).

bnw Naveh and Shaked interpreted bnw as a transcription of *bdnuva-, the nominative form of a loanword *bdnuvant-, "resplendent." They also suggested that it may be a term to refer to "high quality flour" (ADAB, 238). Whereas *bdnu, "beam, lustre, light" (Tavernier 2007, 549, 579) is well known from western Old Iranian dialects as a formative parr of personal names, it would hardly be an apt qualification of flour. Although the word can be used metaphorically (as in *XSaov/ chirographum means "handwriting" in this passage, rather than "written/ signed bond." In other words: both Gabael and Tobit wrote on (or signed) the same document, which was then split and divided between them. Both halves have handwriting on them, and both count as a chirograph in the developed sense of written bond. 213 To make things more convoluted, the mediaeval Hebrew translation for Tob. 5:3, as represented in Neubauer, has nnnj'j!'\, "bag," where the earlier Greek and Latin translations have XEtp6ypaq>ov/chirographum: '1'j'j ?:Jp •nnnj'j!'\1 •? 1m mnnj'j!'\, "his bag he gave me and my bag he received from my hands" (Neubauer 1878, 25; Weeks/Gathercole/Stuckenbruck 2004). 214 1he mediaeval Aramaic translation uses 1:11m, "load," in the same verse: ;,•? n':J;"J' 'J:l1m1 •? :J;-J' ;-J'J:l1m, "his load he gave me and my load I gave him'' (Neubauer 1978, 8). We assume that these differences result from confusion with the bags containing silver mentioned in Tob. 9:5 (cf. below) .215 Tobiah is subsequently told to find a travel companion, which he finds in the person of Azariah, "Yah has helped." It is in the end this person, the angel Raphael in disguise, who is 213

See also Skemp 2003, 160£, who comments on the Vg version of rhis passage as a summary rather than a free invention by Jerome. See also the texts of two other VL mss. in Weeks/Gathercole/Stuckenbruck 2004, 156, wirh different wording. Of these, L2 (cod Complutensis, 9'h/10'h cent.), described (ibid 24) as "the most expansive version of all rhe extant Latin texts" and probably a contamination of various recensions, gives ego quidem cirografom in testimonium depositi dedi illi et ille alium invicem ad suscepte depositionis fidem tradidit mihi, which sounds like a later attempt to explain the passage. 214 Constantinopel 1516. Fitzmyer also quotes the :1'!l1:!>17J, "bag," in one of the mediaeval Hebrew manuscript traditions of Tobit (for the text, see Weeks/Gathercole/Sruckenbruck 2004). The word reflects JlapafJ'ltmov/marsuppium , "bag" (Fitzmyer 2003, 186f.). In other mediaeval Hebrew manuscripts of Tobit, orher items were exchanged (nll:l~, "seal," 1' :ln:l, "handwriting," and :ln:l7J, "letter" (see Weeks/Gathercole/ Stuckenbruck 2004, 157).

YouR TALLY

FuLL!

22 7

instructed by Tobiah to take servants and camels to Gabael, give him the chirograph and take the money from him (Tob. 9:2). At this point finally some help from Qumran arrives: 4Q197 5:10 (= Tob. 9:2) [w]t'th by gb[']l whb lh ktb ws[b ksp']

[and] you come to the house of Gab[ae]l and give him (the) document and ta[ke the silver]2 16

Greekn and the two Latin translations are in concord with this Aramaic fragment . The wellknown word ::!lJ:p, cognate of the verb :Jn:J, "to write," can have several meanings, all of which go back to simple "(hand)writing," and "written document." The Aramaic parts of Daniel provide a good illustration of these basic meanings. In Dan. 5, :JJJ:p refers to the (hand)writing on the wall (5:7, 8, 15, 16, 17, 24, 25) and in Dan. 6, to a written document (6:9, 10, 11). 217 In developed sense, the noun can denote any written document: a legal record, a letter, a decree, etc. It is clearly not a specialized term. 218 As Greek XEtp6ypaq>ov can mean both "handwriting" and "hand-written document" (hence "written bond," etc.), it is not difficult to see why, if the original recension of To bit was in Aramaic, a translator would have picked this word to render :JJ):p. 2 19 It is at the same time not unlikely that the readers of the Qumranic Aramaic Tobit (or an older Aramaic version) would have had a different kind of document or device in mind from those reading, in later centuries, about the XEtp6ypaq>ov or chirographum. Be that as it may, we seem to be dealing with a symbolon, some kind of document split in two parts and inscribed with the names of the contractant partners and perhaps a date and the amount of silver. There is no reference to notched tally marks and no a priori reason to think of split tally sticks. A last glimpse is given in Tob. 9:5, in which Raphael arrives at Gabael's home. The Vulgate merely states that he gives him back his written bond (reddidit ei chirographum suum) and receives all the money in return. The other versions introduce three new elements: 1) checking, 2) the bags containing the money, 3) the use of seals:

215

The word nnMK deserves some special attention, rhough. It is Biblical Hebrew and interestingly enough only found in Gen. 42-44, the story of Joseph and his brothers. In it, nf]JJI,ltt, "bag" (HALOT, 67: "Sack") occurs 15 times and is clearly a keyword, togerher wirh 'lO:l, "silver"- both in the context of the brothers who are one by one accused by Joseph of having stolen his silver and silver goblet (ll':ll), which he uses for divination. In Gen. 42:27 one of them (Levi according to the tradition in Targum Pseudojonathan) opens his sack (viz!) to take fodder for his ass and finds Joseph's silver mnn7JK '!l:l, lit. "in the mouth of his bag." Apparently nnMK was a smaller bag or pouch put in the larger jliz!. Later, the other brothers find Joseph's silver in their bags as well, until the story reaches Benjamin, in whose bag the silver goblet is found (Gen. 44:12 1'7J'J:l nnn7JK:l, "in rhe bag of Benjamin"). All the Jewish Palestinian Aramaic translations of this text render nnn7JK with the Aramaic word 1li1U, "load." Apparently they understood the word as a kind of pack (c£ Cook 2008, 104). Jewish tradition is fascinated with the places where Joseph supposedly hid his silver and other treasures. Apart from Egypt and Mesopotamia (the banks of the Euphrates), Persia and Media are mentioned (see Ginzberg 1998, vol. ii, 125). It is well possible that this tradition reverberates in rhe mediaeval Hebrew and Aramaic translations of Tobit, also a story in which the adventures of a son travelling to a faraway country are told.

IS

Greek1

He then brought the bags with sealings intact (£v m1~ aQ :ll):;l~, "according to the writing in the book of Moses." See also HALOT

1728, s.v. ::11;9; D]PA, 271, s.v. ::11;9, "writing, document, decree, scripture"; DSA, 415, s.v. :ln:JN, "writing, book, letter, handwriting"; D]BA, 608, "written document, amulet, script, handwriting, orthography"; MD, 203, s.v. kddba 1, "book, writing." 218

The same applies to the noun spr in Aramaic texts from the Achaemenid period. See Folmer 1995, 629631 and 687-689. 219 The same holds true for the Syriac translation which has 'ef!drd, "document, deed" (Tob. 5:3). See n.211.

228

YouR TALLY rs FuLL!

WF.M. HENKELMAN and M.L. FoLMER Greeklll

And Gabael got up and brought the bags; he presented the sealings (e8et~e -cue; cr /i/ vowel shift), and 2) an original stem kiti-. The subsequent question is whether in Achaemenid Elamite there is one stem with one basic meaning, one stem with more than one (contextual) meaning, or two homonymic stems (i.e. homonymy caused by earlier kutuli- becoming kiti-). In the Fortification archive kiti- occurs in a number of texts that record the allocation and consumption of wine at the royal table. These texts always use the fixed expression HAL£SSANA

rather assuming - not very plausibly - that the phrase refers to animals that were slaughter~d ~~r consumption at the livestock station where they were kept, taking batmanna to denote an official m charge of domestic animals' in contrast to the ordinary herdsman (batera; ibid., 43, 50 n. 17), a distinction for which there is no support. Henkelman/Jones/Stolper (2006, 12 fu. 7) more recently suggested the possibility that the calculation represents a ten percent loss allowance - halbaka (ihalpi.k.01) means either 'was/were slaughtered' or 'was/were dead'. 16o The interpretation of the harika halbaka phrase discussed in the previous footnote is complicated by the appearance of multiple words of similar or the same shape that are certainly or. pos~ibly non-Ela~ite loa_nwords. Tavernier distinguishes between two different Old Iranian words featunng m Achaememd Elamtte texts. The first, ha-ri-ik-ka4 (*arika-), occurs in DBe meaning 'evil, unfaithful' . See Tavernier 2007a, 38 [1.4.9.1]; Hinz/Koch 1987, I, 628 s.vv. [ha-ri-i]k-kdn (with uncertain emendation, however) , ha-ri-ik-qa, ha-[ri-ik]-qa-if; edition ofDBe: Grillot-Susini/Herrenschmidt/Malbr;m-Labat 1993., Hinz ~1973, 125) tr~n~­ lates 'treulos, verworfen, gotdos', with reference to Old Indic alika-, falsch, unwahr ; cf. Hmz 1975, 36 ( m1t ungesicherter Etymologie'). According to Hinz and Koch ha-[ri-ik]-ka4-if in DBe 10.26 is a 'Verbalbildung' (also Hallock 1969, 691b). This suggests that the Elamite scribe(s) would have used *arika- as a verbal base and turned it into a conjugation I finite verbal form by adding the third-person suffix -f. While this is not impossible, of course, perhaps the possibility of a nominal sentence with *arika- plus (generalized) nominal suffix -fused as predicative adjective should also be considered. Tavernier (2007a, 506 [5.3.4.24], and previously Hinz/Koch 1987, I, 628-629 s.vv. ha-ri-ik-qa, ha-ri-kas) attributes the meaning 'pr~visions' to the second Iranian word. It seems more likely, however, that it represents the same word as that dtscussed above. C( GLS'mi'-ik-dim ha-ri-karif in NN 0559:1-2, presumably 'bad/rotten fruit' , PF 1965:10 where ha-ri-kaf ualifies seed (NUMUNMES) and NN 0870:4-5 where it qualifies a certain amount of wine. har;ikka also q c h b k" . DISLUMES . -zr-ra occurs in DPfe:23-24, quoted in Henkelman 2008, 369 m. 860: u- ea-nu m-nz-en ap-pa ha-ri-ik-ka el-ma-man-ra, which in his translation reads 'lest that come to pass, what evil man plots'. 4 161 Both end in -J in Achaemenid Elamite. 16 2

See Tavernier 2007b, 27 8-289.

272

ELUSIVE SILVER IN THE AcHAEMENID HEARTLAND

M. TAMERUS

tibba kitka, which is translated as 'poured before the king' . 163 As such it parallels texts of the same kind that involve non-liquid commodities consumed at the royal table, which instead use the phrase HALESSANA tibba makka, indeed meaning 'consumed before the king'. 164 Hallock translated the above kitka as 'expended', 165 but took 'to pour' to be the literal meaning of kiti- that underlies the contextual meaning 'to expend'. It should be pointed out, however, that Hallock's 'to pour' for the royal wine texts is based only on his interpretation of the form kitinti (a 2.sg. conjugation Ill form) in DBe 61.76 and 67.89. There kitinti is used in the ex. DISNUMUNMES -m. a-nu k.t-tt-zn-tt, . . . w h"1ch h e trans1ate d 'mayest t h ou not pour out t h y presswn seed' . 166 This somewhat graphic translation is not quite necessitated, however. As Henkelman has noted, there are several Fortification texts in which non-liquid commodities are the object of forms of kiti- - including PF 1972 and other JaumarraJ texts - or where the context seems to require another meaning. 167 He further points out that in those texts kiti- perhaps continues Middle and Neo-Elamite kutuli-, 'to protect, to preserve', so that 'one may either assume two homonymic verbal bases with different meanings and used in different contexts[ ... ] or a verbal base kiti-, "to keep, to protect," that somehow came to be used for expenditures ofliquids.' 168 The former is the most likely - in other words, one may assume that in Middle Elamite both kutuli- and kiti- existed. The form kitinti in DBe is presumably based on kiti- (< kutuli-), the phrase NUMUN -ni anu kitinti meaning 'you will not keep (perhaps > 'have') your offspring', thus basically conforming to what is expressed by the parallel Old Persian and Babylonian versions. 169 Here, the same meaning is taken to apply in the phrase inni kitif in the JaumarraJ

163

Henkelman 2010, 676 fn. 24.

164

Henkelman 2010, 676 fn. 23.

165

Functioning as a parallel to makka which he translated 'dispensed'. See Hallock 1969, 25.

texts, i.e. 'he/they did not preserve/keep', which perhaps even came to denote 'he/they did not have'. To summarize the discussion thus far, the pertinent phrases from the Jaumarras text presented above in §6.1 (PF 1972) can be tentatively interpreted as follows. Following the statement that the storehouse officials withdrew 233 1. of barley, the text adds that they then did not 'keep' it (meni inni kitis') to use in an exchange procedure (appa sut huddamanba), which suggests that they did something else with the barley. Subsequently, 38 JaumarraJ were 'impressed' (harak). It remains unclear for now what the term Jaumarraf itself denotes. However, it is clear that one Jaumarraf is equal to the value of one-fifth of a silver shekel, so that 38 Jaumarras equals 7.6 shekels. One may surmise that '38 Jaumarras either means that 38 objects referred to as Jaumarraf were each impressed with something, or that one object called Jaumarraf was impressed with 38 marks, perhaps, of an unknown kind. In the latter case one might perhaps have expected a postposition with JaumarraJ to indicate that something was done 'on' it. This is in fact the case in NN 2358:18 (see below), though it is not certain whether -ma there functions as such. One may further note that since the 233 1. of barley withdrawn in PF 1972 would technically equal 38.8 Jaumarras (in accordance with the 5:1 rate), it was apparently necessary to round it down to a whole number. On the other hand some texts do include fractions of Jaumarraf. In the next section additional texts in which Jaumarraf feature are presented. The question why JaumarraJ may have been 'impressed' is discussed thereafter.

6.3 Additional saumarraJ texts PF 1980 (short account- wine, barley and fruit withdrawals)

166 Hallock 1969, 25, 714a. The Old Persian version (Schmitt 2009, 82) reads utdtaiy taumd md biyd, 'and may you have no offspring/family', the Babylonian version NUMUN-ka lu-u ia-a-nu (Von Voigtlander 1978, 44), 'may there be no offspring (lit. 'seed') of yours'.

Obv. 5

167 Henkelman 2010, 676 fn. 24. Hinz/Koch (1987, I, 490 s.v. ki-ti-is} attempted to solve the issue of non-liquid objects occurring with kiti- by taking it to mean 'verbrauchte, verausgabte' in these cases, i.e. similar to Hallock's 'expended'. 168

Henkelman 2010, 676-677 fn. 24.

169

Hinz/Koch (1987, I, 499 s.vv. h.ki-ti and ki-ti-in-ti) further complicate the matter by suggesting that the verb kiti- is related to the rare word kiti- which they take to be a continuation ofNeo-Elamite ku-tuMEs. The former is thought to mean 'GroGvieh' - one may add here that Achaemenid Elamite kiti and orthographic Neo-Elamite kutu can hardly be anything other than Iranian *gaiOa- or a form derived from it- and to have come to denote 'Habe' by transfer of meaning. On this basis they take kiti- in kitinti to mean 'haben, behalten', regardless of the meaning 'verbrauchen, ausschenken' attributed to occurrences of the 'gleichlautenden kiti-' in Fortification texts. See Hinz/Koch 1987, I, 489-491, s.v. ki-ti-in-ti, and cf. 490 (s.vv. ki-ti-iS, ki-ti-if-dd), 494 (s.vv. ki-ut-ki, ki-ut-qa). This is problematic, however, not only because Iranian loanwords in Elamite are not commonly verbalized, but also because it is not certain that kiti actually did come to mean 'Habe'. Interestingly, a transferred meaning 'livestock' > 'possessions' may be implied by the use of the logogram GUD in Fortification texts, for which Hinz and Koch point out a parallel in Avestan *gaefJa, namely both 'Herde' and 'Hab und Gut' (Hinz/Koch 1987, I, 497 s.v. GUD.lg; Henkelman 2010, 738 fn. 266; cf. *gaiOd-, 'cattle, herd', noted in Tavernier 2007a, 186 [4.2.623]; note thatTavernier translates 'cattle, household', but Schmitt [2014, 178], who translates 'Vieh[besitz], Herde', points out that the etymological con-

273

22 7 QA GISGESTINMES

6

maz0 -zf-ifGISGESTINMEs hu-be-ik-mar 6 HALku-ri-

7

ma ma:zo-zf-if me-ni 30 GISJd-u-mar-rdf ha-rak0 a-ak

8

HALma-tur-za-na a-ak HALu-ma-ak-ka 4 a-ak HALzir0 -ru-si-

9

ka 4 PAP 3-be-ud-da in-ni ki-ti-if ap-pa su-ut hu-ud-

10

da-man-ba me-ni un-ra 83 GISJd-u-mar-rdJ pfr-nu-ib-

11

ba GISJd-u-mar-rdf ha-rak0

They withdrew 227 I. of wine. 6-7 From that wine, Kurima withdrew 60 I. 7 Then, 30 Jaumarrafwere impressed. 7- 10 And Maturzana and Umakka and Zirrusika, a total of3 (persons), did not keep (it) to make an exchange(-request). lO-ll Subsequently (by? for?) each 83 Jaumarraf (and) half a Jaumarraf were impressed. 5- 6

nection with Vedic gehd-, 'Haus', is controversial). This does not, however, mean that the same development occurred for kiti.

ELUSIVE SILVER IN THE ACHAEMENID HEARTLAND

M. TAMERUS

274 Obv. 14

Jaumarraf were impressed, and (from?) the balance of each 153(?) 172 Jaumarraf were withdrawn.

92 GIS

That was received 'on' a parchment document (or, perhaps, 'a parchment document was received for it'?- see the use of ukku below), Missabadda took (it).'

31-32

15

ir-du-ma-um SE.BARMES maZQ-zi-if 'me-ni' SE.BARMES hu-

16

be HALzf-iz-za hi-Je ASJi-Ju-da-nu-if r mu!' -tuk HAL' hu' -pir-

17

ri maZQ-Zi-if pir-tu-ka4 me-ni HAL'zi-iz '-za

18

ri df in-ni ki-ti-if ap-pa su-ut hu-'ud -da-man-ra

19

me-ni hi pa-ri-ma-ak 1 si-ri-maZQ-si 60-na 1 ME

20

60 GISfd-u-mar-rdf me-ni in-ni hu-ut-tuk fu-tur

PF 1987 (general commodity account- fruit)

HAL' hu' -pir-

Amounts of fruit are mentioned under three headings: ukkap daka (lit. 'placed on them', presumably meaning 'deposited to their account'), duka ('received') and mazzika ('withdrawn'). The following text pertains to the latter column.

They withdrew 2,760 I. ofbarley. 15-17 Then, Zizza, dwelling 170 at Sisudanus, withdrew that barley, ... 17- 18 ThenZizzadid notkeepafto make an exchange(-request). 19-20 Subsequently this was issued (lit. 'went'): 1 sirimazsi 'of 60' (and) 160 Jaumarraf. 20 No balance was subsequently made. 14- 15

Obv. 24

SEMES SALMES_na ha-sa-na nu-ut-tam 5-ma su-ut hu-ut- 'tas"'

26

16 2 BAR GI5mi-ik-tam5 za-ak-me me-ni ANSEMES HALba-' ka 4'-

27

pdr-na hi kur-ma-ak A5be-ul20-um-me-ma ANJTJMES ha-du-

28

kdn-nu-ia 12 na-an pir-ka4 fu-tur 11 2 QA maZQ-Zi-man-

29

pi hu-be-na in-' ni' ki-ti-if ap-pa su-ut hu-ud-da-man-

30

ba me-ni un-ra 55 GISJd-u-mar-rdf 1 SUMES GISJd-

31

u-mar-rdf ha-rak0 da-tam 5 ap-pu-ka4 -na-ma

They withdrew 832 I. of fruit. 24-25 They then exchanged it for an adult female donkey of lowest quality. 26 500 I. of fruit (was) the price. 26-28 Then the donkey was entrusted to Bakaparna, in the twentieth year, in the first month, on the twelfth day. 28 -29 They were (also) to withdraw 17 1 the balance, 332 I. 29 -30 They kept nothing of that to make an exchange(request). 30-31 Subsequently (by? for?) each 55 Jaumarraf (and) one-third of a Jaumarraf were impressed. 3 1 (This was) in accordance with former regulations.

a-ak ap-pa maZQ-zi-ka4 hu-'be' HALza-su-ma a-ak HALpdr-mi-ia-na I-IALpi-ut-tuk-ka 4 PAP 3-be-ud-da df 'in' ni ki-ti-if ap-pa su-ut hu-ud-da-man-ba me-ni un-ra mt'!-kak-ka4

30

50 GISJd-u-mar-rdf ha-rak0 -ka4 a-ak fu-tur un-ra-na 1 ME 50+ '3?' GIS-, 'kh u- be A5 tup-pt' KUS~MES u k -k u-na sa-u-mar-ras,. maZQ-Zt-t

31 32

1

5

3

3

3

GISkaz0 -fa HALba-ba-ak- r ka4' 27 GISJd-umar-rdf ha-rak0 -ka4

4

2

2

GISkaZQ-la HALmi-ud-du-zi-ia 18 GISJd-umar-rdf ha-rak0 -ka4

HALu-pi-iz-za dJ in-ni ki-ti-if me-ni 45 GISJd-u-mar-rdf ha-rak0

G 1Skaz -la 0

2

kazla fruit, Upizza did not keep af. Then 45 Jaumarraf were h.' ed.

3

kazla fruit, Babakka: 27 Jaumarrafwere h.'ed.

4

kazla fruit, Midduziya: 18 Jaumarrafwere h.'ed. ap-pa ASkdn-ti-ma maz0 -zi-ka4 hu-be uk-ku '1' rdp-tam 5 a-ak 1 hi-du PAP 2

31

'UDU.NfTA'MES HALmi-if-fi-ia-ma-na a-ak HALdf-ba-ia-u-da a-ak HALda-ut-ti-ia-

32

, na' pAP 3-be-ud-da su-ut hu-ud-da-if A5ul-hiMES HALESSANA-na-ma r kapo' -pa-ka4

33

[me-ni UDU]. 'NfTA'MES hu-be HALra-iJ-da-ma hi-Je HALba-te-ra hi kur-ma-u-ut

10

uk-ku du-ka4 HALmi-if-Jd-ba-ud-da mar-ri-if

As suggested in Hallock 1969, 584; Tavernier 2007b, 267 with fn. 21. On the interpretation of conjugation forms with auxiliary -ma- see Stolper 2004a, 80 with references.

6

Obv. 9

And what was withdrawn, (of) that Zasuma and Parmiyana (and) Pittukka, a total of 3 (persons), did not keep aJ to make an exchange(-request). 29 -31 Then (by? for?) each ... 50

171

2

PF 2078 (short account - barley withdrawal)

27-29

170

maZQ-zi-ka4

30-32 What had been withdrawn at the storehouse, for that (lit. 'upon that') Missiyamana and ASbayauda and Dattiyana, a total of 3 (persons) made an exchange for 1 ram and 1 ewe, a total of 2 head of small livestock. 32 They were requested (lit. 'summoned, gathered') for(?) the House of the king. 33 Then we entrusted those sheep to Rasdama, the herdsman.

PF 1986 (general commodity account- fruit)

29

du-ka4

Obv. 30

24

28

uk-ka4-ap da-ka4

Obv. 1

27 '2 BAR' 2 QA GI 5mi-ik-tam5 maZQ-Zi-if'me'-ni 'AN',

25

Obv. 27

275

172

32 1 BAR SE. BAR MES maZQ-zi-if me-ni in-ni

11

ki-ti-if ap-pa su-ut hu-ud-

12

da-man-ba me-ni un-ra 1 ME 60+

13

1 GISJd-u-mar-rdf 2 QA (read presumably: 2/3) GISJd-u-

14

mar-rdf ha-rak0!

See Aperghis 1997, 287. The meaning of the phrase futur ... mazzik is unclear.

______________________J

ELUSIVE SILVER IN THE AcHAEMENID HEARTLAND M. TAMERUS 173

9-10 Th ey Wit . hdrew 970 I. of barley. I0-12 then the did (-request). 12-14 Subsequently (by? for?) each 161 tmpressed.

saZmarr:;:~}eep -(thlt). to mak~ an exc~ange two irds of a saumarras were .

7 10 2 and 6-7 They withdrew 1,604.5 I. of wine. · Then they made an exchange (for) 1 p. horse one m. 174 female donkey, a total of 2 (animals). 10 1,210 I. of wine (was) the price. II -! They were (also) to withdraw the balance, 394.5 I. of wine. !2-!3 They did not then keep (it) to make an exchange(-request). 13 (By? for?) each 198 faumarraf were impressed.

NN Oh146 (general commodity account- fruit; cited after collated edition by W Henk I ) · e man See t e comment to pp 1987, above.

Rev. 21

( .. .) 20 1 BAR 9

QA SE.BARMES maZo-

22

zi-ka4 hu-be-iK-mar HALmi-ut-ra-da

(. · .) md-ni df in-ni ki-ti-if ap-pa su-

23

hi-Jeba-ra-bar-rdSlO 1 BAR9

ut hu-ud-da-man-ra me-ni 12 GlSJd-u-mar-' ras,., ha - rak 0- ka4

24

zi-if me-ni in[-ni] 'ki-ti' -if ap-pa 'su'-

25

' ut' hu-ut-man-ra '50?' (+)3 cGIS'Jd-u-mar-rdf ' rak0'

26

HAL'kar-ma'-na hi-se ba-'ra-bar'-rdf 10 SE.BARMES maZo-

27

zi-if 'me-ni in-ni ki' -ti-if 'ap' -pa 'su'-

( ... ) me-ni df in-

28

ut hu-ut-man-ra '50' GISJd-u-mar-rdf ha-

14

ni ki-ti-if ap-pa su-ut hu-ud-da-man-ra

29

rak0

15

me-nr· 15 GIS·sa-u-mar-ras ' ·· ha-rak0-ka4

Obv. 9 10

9impressed. -10 Then h e d"d 1 not keep af to make an exchange(-request). IO Subsequently 12 faumarraf were

Obv. 13

175 22 24 21 -22 619 I. of barley were withdrawn. - From that, Mitrada, a storekeeper,25 withdrew 319 I. of barley. 24·25 He did not then keep (it) to make an exchange(-request) . 53(?) faumarraf 27 28 were impressed. 26-27 Karmana, a storekeeper, withdrew 300 I. of barley. - He did not then keep (it) to make an exchange(-request) . 28 -29 50 Jaumarraf were impressed.

13-14Th en he d"id not keep af to make an exchange. 15 Subsequently 15 faumarraf were impressed. Rev. 30

(. · .) me-ni ap-pa maZo-zi-

31

'su, -ut hu -ud-da -man-ra if-ti in-ni ki-ti-if-da an-pa r

32

me-ni 2 ME 362h GlSJd-u-mar-rdf ap-'pu' -ma ha-' rako, -ka4

NN 1015 (short account- fruit withdrawal; cited after collated edition by W Henkelman)

30-31 Th en wh at he had withdrawn he did not keep to make 32 an exchange(-request). Subsequently 236 2/3 faumarrafwere impressed by/for(?) them.

Obv. 8

( ... ) 7 2 BAR 1/z GISmi-' ik' [-tam 5]

Lo.E. 9

maZo-zi-if in-ni ki-ti[-if ap-pa]

10

NNHenkelman) 0704 (short account - wine and barley withdrawals·, cited after collated edition by W Obv. 6

( ... ) 1 ME 60

8

4 QA 20-ir-ma-ki GIS'GESTIN'MES maZo-Zt-ts · ., me-m· 1 ANSE · KUR· 'RAMES AS?' prr-ra' ' nu·, -tam 5-maf 1 ANSE •

9

MES SAL -na ma- ' a, -kur-tt-ta. . ' na PAP' 2-e-da su-

7

10

ut 'hu'-ut-taf 1 ME 21 GIS'GES.TIN'MES za-ak -me

11

'Iu'-tur '30'(+)9 4 QA 20-ir-ma-' ki' GISGESTJNMES maZo-

12

zi-man-ba me-' ni' in-ni ki-ti-'if su-ut hu-ut-

13

man-ba un-ra 1 ME 98 GISfd-u-mar-rdf ha-rako

'QA' SE.BARMES 'maZo'-

su-ut hu-ud-da-man-ba me-' ni GIS?'

Rev. 11 12

un-ra 39 6-ir-ma-ki GISJd-

R.E. 13

2-um-'me-na'

u-mar-rdf 'ha-rak0'

10 an exchange(-request). 8-9 They withdrew 235 I. of fruit. 9· They did not keep (it) to make 3 - Then (by? for?) each 39 and 1/6 faumarraf were impressed. (1 This is) the second (tablet). 10 12

173

Read presumably pir-ra-tam -mi-ia- or a variant spelling (*jratamya-, 'of the best quality' : Tavernier 5

2007a, 405 [4.4.2.6]) . 174 See Tavernier 2007a, 413 [4.4.6.4] on makurtiya (*[h]uvagrdya-), translated '(animal raised) in the own house'. l75 According to Tavernier 2007a (417 [4.4.7.17]) this word represents *barabara-, 'he who carries a burden, load', but given its occurrence in a similar context (PF 1977) it is more likely that this is a variant spelling of ambarabarraf, 'storekeeper'.

ELUSIVE SILVER IN THE ACHAEMENID HEARTLAND

M. TAMERUS

NN 2206 (journal; cited after collated edition by W Henkelman) me-ni ap-pa A5kdn-ti-ma maZo-'zt -ka4 'in' -ni ki-ti-if ap-pa su-ut hu-ud-'da'-

Rev. 32 33

man-ba me-ni un-ra 20 GlSJd-u-mar-rdJ ha-rak0

of wine, each (withdrew) 311 1. of wine. 7 They did not then keep (it) to 8 10 make an exchange(-request). 7-8 Then they made an exchange(-request). • 1 male m. donkey 10 (and) 1 female donkey of lowest quality, a total of 2 (animals) they made an exchange for. 3601. of wine (was) the price. 11 x They withdrew 5731. of wine. tl-!2 They did not then keep (it) to make an exchange(-request). 13- 14 Subsequently 285 (and) half a JaumarraJ were impressed.

5-6 They withdrew 9331.

32 33 -

Then what was withdrawn at the storehouse they did not keep to make an exchange (-request). 33 Subsequently (by? for?) each 20 JaumarraJ were impressed.

NN 2208 (short account, barley and fruit withdrawals - the first, although intelligible, is broken and not presented here; cited after collated edition by W Henkelman) ( ... ) 1 BAR 9 QA GI 5mi-ik-tam5 maZo-zi-

Rev. 19 U.E. 20 21 22

if hu-be in-ni ki-ti-if su-'ut' hu-

22

'ni 56' [GISJ'Jd' -u-mar-rdJ ha-rako

Rev. 26

19- 20

They withd rew 19 1. of £rutt. · 20-21 Th at t h ey d.d 1 not keep to ma k e an exchange(-request). 21 22 - Then (by? for?) each 2 JaumarraJ (and) half a JaumarraJ were impressed.

'12' G!Sir-

r

21

They did not then keep (it) to make an exchange(-request).

Subsequently 56 JaumarraJ were impressed.

21 - 22

0

Obv. 9

me' -ni r in' [-ni] r ki' -ti-iJ'su-ut hu-ud' -da-man-' ba' me-

21

They withdrew 3361. of barley.

20

ud-da-man-ba me-ni un-ra 2 GlSJd-u-mar-' rdJ' pir-nu-ip-Ju GI 5Jd-u-mar-rdJ ha-rak

NN 2266 (short account- fruit withdrawal; cited after collated edition by W Henkelman)

( ... ) 11 '6 Qj\ SE.BARMES maZo-zi-if

Rev. 20

( .. . ) 12 2 BAR 7 'QA' [GISJ

27

'mi-ik' [-tam 5] maZo-Zi-iJ me-ni 12 2 BAR 'T [QAJ

28

[SE].'BAR' MES su-ut 'hu'-ut-taJ GIS'mi'[-ik-tam 5]

29

r

za' -ak-me SE.BARMES rHAL'za-ir-du-' if' [hi kur-ma-ka4]

They withdrew 387 1. of fruit. barley. 28 · 29 The fruit (was) the price.

27-28

27 29

They then made an exchange(-request) for 387 1. of The barley was entrusted to Zardus.

10

. .k-tam maZo-Zt-ts , "' du-ma- r urn' G!S mt-t 5

NN 2355 (short account - barley and wine withdrawals; cited after collated edition by

11

me-ni 'in' [-ni] 'ki' -ti-iJ 'ap' -pa

W Henkelman)

12

su-ut 'hu' [-udj -da-man-ba me-ni

13

un-ra '60' GlSJd-u-mar-rdJ ha-

14

rak0

Obv. 5

6

9 10 -

They withdrew 360 1. of fruit. ll - 12 They did not then keep (it) to make an exchange(-request). 12- 14 Subsequently (by? for?) each 60 JaumarraJ were impressed. 5

NN 2289 (short account- wine, barley and fruit withdrawals; cited after collated edition by W Henkelman)

Obv. 5 6

7

me-ni in-ni ki-ti-iJ ap-pa su-' ut hu-ud'da-man-ra me-ni 1 ME 45 GI 5Jd-u- 'mar'-rdJ

8

ha-rak0 ( ... )

They withdrew 8701. of barley. 6-7 They did not then keep (it) to make an exchange(-request).

7- 8

Subsequently 145 JaumarraJwere impressed.

Lo.E. 11

( ... ) '90'[(+)3 3 QA]

[G1sGESTINME5 maZo-zt]-' if un-ra' 31 1 QA rGISGESTINMEs'

29 Gisr ir' -du-' ma-um' SE.BARMES maZo-Zi-if

Rev. 12

( ... ) 1 ME 82 GI 5GESTINMES maZo-ziif me-ni 1 SAL fi-ba-um-me pir-ra-tam 5-mi-ia su-ut hu-ut-taJ A5u!-hiME5 HALESSANA-na-ma me-ni

7

[me-ni in-ni ki-ti-is1 'su' -ut hu-ud-da-man-ba me-ni

13

8

[su-ut hu-ut-tas1 'I' ANSEMES GURUS ma-a-kur-ti-ia-' nu' -

14

SALfi-ba-'um'-me hu-be HALam-pir-za GI 5GESTINME5

9

[if 1 ANSEMESJ 'SAL'MES! -na nu-tam 5-'maJ' PAP 2-e-da 'su'[ut] 'hu-ut-taf' [30](+)6 mar-ri-if G15 GESTJNMES za-ak-me

15 17

ku-ti-ra ir du-nu-ik HALESSANA-ik-mar iJ1 Je-ra-ka4 1 ME 80 GI 5GESTINMES za-ak-me su-ut hu ut-tuk-ka4 Ju-tur 2 mar-ri-iJ GI 5GESTINMES 'maz0'-

10

GI 5GESTINMES maZQ-zi-man-'ba me-ni'

16

11

'x-uf 50(+)'T [3] QA

12

r

in-ni' ki-ti-iJ su-ut hu-ud-da-man-' ba'

18

zi-ma-'ak' hu-be in-ni ki-ti-iJ ap-pa su-

13

[me-ni] un-ra 2 ME 85 pir-nu-ba GIS'Ja"-u-mar-rdJ

19

ut hu-ud-da-man-ra me-ni 10 GlSJd-u-mar[-rds1

14

[ha-rak0] ( ... )

U.E. 20

'ha'-rak0

279

280

M. TAMERUS

ELUSIVE SILVER IN THE AcHAEMENID HEARTLAND

11 12 - 1hey withdrew 1,820 I. of wine. 12"13 They then made an exchange(-request) for 1 firstclass f~male slave ~or the House of the king. 13-15 1hat slave was subsequently given to Ampirza, 6 the_wme master. s- It was ordered by the King. 16-17 1,800 I. ofwine (was) the price (for wh1ch) an exchange was made. 17-18 1he balance, 20 I. of wine, was (also) to be withdrawn. 18-19 That they did not keep to make an exchange(-request). 19-20 Then 10 JaumarraJ were impressed.

1

NN 2358 (journal; cited after collated edition by W Henkelman) Rev. 16

( ... ) [in-ni]

17

ki-ti- 'is"' [ap-pa su-ut] hu-ud-da-man-ba ap-pa AS kdn-ti-ma ma.zo-'zi' [ -ka me-ni un]-

18

ra 22 [G 15Jd-u]-' mar' -rdJ 1 SUMES GISJd-u-mar-rdJ-ma ka

4

4

-za-ip

1 17

61hey did not keep (it) to make an exchange(-request) (for) what had been withdrawn at 17 18 the storehouse. - Then (by? for?) each 22 faumarraJ (and) 113 of a JaumarraJ(?) were struck.

6.4 Another attempt at interpreting the saumarraJ texts The texts presented above add a number of things to what is already known from pp 1972 that_ ~ay clar~f)r, bur also complicate the interpretation of the JaumarraJ phenomenon. One addltlon that Is most obvious is that whereas texts like PP 1972 identifY the location at which the initia~ transaction (i.e. t~e _withdrawal) is to be situated - in that case identifYing it as a halmarras- others make explicit that commodities were indeed withdrawn from a storehouse. ~ example is the journal NN 2206:32, which refers to appa kanti-ma mazzika, 'what was Withdrawn at the storehouse (kanti)'. 176 It remains unclear, however, where precisely the subsequent exchanges were undertaken, or where the transactions involving JaumarraJ took place. A number ~f texts sh~w that in some cases the full amount of a withdrawn commodity was used to acqUire somethmg, bur in other cases only a part of it was used. The short account NN 0704 records the withdrawal of 1,604.5 I. of wine, of which 1,210 I. were exchanged for a _horse and a donkey. _The remainder, referred to by the term futur, 177 was apparently also ~Ithdrawn. The fo11owmg statement, that inni kitif appa sut huddamanba, 'they did not keep It to ~ake an exchange(-request)', pertains to this remainder. Since it was not exchanged for anyrhmg, the fact that subsequently faumarraJ are involved must be the direct result of this ~11. 6-13). ~e same text also records a barley withdrawal for the same purposes, but none of It was used m an exchange. Whereas in another short account, PP 1980, withdrawals of wine (1~. 5-11) and b~rley (11. 14-20) were not followed by an exchange, a part of the fruit that was withdrawn was 1ll fact used to acquire a donkey. NN 2289 shows the same picture: nothing 176 On kanti see Henkelman 208a, 398 fn. 915. 177 In administrative contexts Jutur seems to mean 'balance' · See' e·g., Hallock 1969 , 15 , 60 , 759 . 0 ccur. . rences m M1ddle and Neo-Elamite inscriptions and in DB have been assigned meanings such as •· t" · ' ' (I ) (I , e 1us ICe, !ustness, correct y .'~raper y). Se~ Hinz/Ko;h 1987, II, 1187-1188 s.vv. According to Hallock (1969, 15) presumably the umfy1~g concept 1s t~a-t of correctness"'. Cf., however, the discussion in Fisher!Stolper 2015, 17-18, commentmg on Hallocks Interpretation and on the alleged adverbial uses of the word.

281

was acquired for the withdrawn wine and barley, but the fruit was traded off. Interestingly, barley was acquired for the fruit. The amount of barley that was acquired was identical to the amount of fruit, and these were of equal value, as can be expected. 178 Difficult to understand is the passage NN 2289:5£, where it seems that first the wine that was withdrawn was not used for sut, yet subsequently it was used, and for part of the wine two donkeys were acquired. 179 Matters are perhaps complicated by the occurrence of a variant to the (meni) inni kitif phrase. Several texts instead read aJ inni kitif. 180 The word aJ causes some lexicographical difficulty. An Elamite word aJ is attested from Middle Elamite onwards and is commonly translated 'livestock, herd'. However, according to Hinz and Koch it also occurs with a different meaning in Achaemenid Elamite, which they suggest is 'Teil(?)' .181 It is probably safe to assume that when the storehouse officials' task was to dispose of commodity surpluses, it was a regular procedure to attempt to exchange them for animals. Given the regularity of this phenomenon one should at least consider the possibility that in the pertinent faumarraf texts af refers to (a set of) animals. Another possibility that one could consider is that Hinz and Koch interpreted correctly and that it refers to the officials' 'share' or 'possession' in the sense of the goods they had taken into their possession by (formally) withdrawing them from the storehouse (which is recognized as such by having been recorded explicitly). However, here it should be noted that their examples for the alleged aJ all come from either JaumarraJ or similar texts. So the interpretation of this word actually hinges on the question whether it is a separate word at all. On the one hand it seems to make sense to take aJ to refer to the commodity that was withdrawn from a storehouse. If this is indeed correct, af may be implicit in texts that only read (meni) inni kitif. NN 0146:30-31 would seem to indicate that af indeed refers to the withdrawn goods. In an earlier part of the same text (recording the accounting of the nineteenth year) two separate instances of fruit withdrawals are each followed by the statement that af inni kitif appa sut huddamanba, 'they did not keep af to make an exchange(-request)'. At the end of the text, in the part that records the accounting of the twentieth year, it is stated that appa mazzifti inni kitifda appa sut huddamanra, 'what they had withdrawn they did not keep to make 178 It is difficult to explain, though, why according to NN 2289 at one and the same place barley was withdrawn (11. 20-22), presumably to dispose of it, while at the same time an even greater amount of barley was acquired with withdrawn fruit. It seems that both transactions pertain to the same year. 179 Note, however, that the first occurrence of meni inni kitif in I. 7 has been fully restored. Perhaps it is not unreasonable to expect mazzif at the beginning of I. 7, following un- 'ra' 31 1 QA 'G 15 GESTINMES' in I. 6. And so there are other complications, such as the fact that in PF 1979 withdrawals of barley and fruit are used to undertake an exchange (at least as implied by the use of the term sut) , yet both the grain and the fruit were allocated (kurmaka) to one of the officials who initially withdrew them. 180 See PF 1980:18, PF 1986:28-29, PF 1987:2 and NN 0146:9, 13-14. 181 See the two separate entries dJ in Hinz/Koch 1987, I, 84. They connect their second af with dJ-ki, which occurs a couple of times in DBe and seems to mean 'anything' (see, e.g., Hallock 1969, 670b, taking it to represent Old Persian *CiJCiy-) . See the references in Hinz/Koch 1987, I, 88 s.v. aJ-ki. They translate 'erwas', though whether its literal meaning is indeed 'ein Teil' depends on the interpretation of as. See also Hinz 1970, 438, taking 'df fur Teil, und entsprechend df-ki fur "erwas" im Worrsinne von "einen Teil'" (for which the analysis would then presumably be la5+ki.01, lit. 'part+one'). However, that hereby one can avoid 'dJ stets mit "Vieh" uberserzen zu mussen' is hardly an argument.

M. TAMERUS 182

an exchange(-request)'. More explicit (despite the text's syntax at this point) is NN 2358, reading [inni] kitif [appa sut] huddamanba appa kanti-ma mazzi[ka], 'they did not keep to make an exchange(-request) what had been withdrawn at the storehouse (kanti)'. Similarly, in PF 1980 (1. 24ff.), after a part of the fruit that had been withdrawn was used to pay for a donkey that was acquired, the remaining amount was also withdrawn. At that point the text continues to record that hube-na inni kitif appa sut huddamanba, lit. 'of that they did not keep to make an exchange(-request) '. 183 Given the otherwise standardized nature of the phrasing it would be logical for appa mazzifti and hube-na in these texts to refer to the same thing as af in other texts. In that case the phrase af inni kitif would refer to the fact that what had been withdrawn was not subsequently used to exchange it for something else, in the sense that something other than the 'normal' procedure had followed instead. Support might additionally be found in related texts that, however, mention neither sut nor !aumarras, though there are also potentially further complications. For instance, PF 1974:3-9 records that a storekeeper called Mauparna withdrew an amount of barley, but then died (hafpika) , followed by the statement that a!-e inni kitif, 'he did not keep his/its aJ.I B4 In the subsequent part of this text another withdrawal is referred to, this time by two other storekeepers. Both the reference to Mauparna's withdrawal and that to the second one are followed by the statement that ta!!up-ikmar inni pirru !arra!da. It is not entirely clear how this is to be translated, as it is not immediately dear to whom ta!!up refers. It can generally mean 'people' or 'troops', and in case of the former the phrase could mean 'from the people they did not collect anything', following the translation that Hallock suggested. 185 The word pirru, placed between the negation inni and the verbal form is difficult to interpret, but its use in other contexts perhaps suggests that it might have to do with conversing one thing into another, 186 and one might surmise that in the present context it concerns exchanging the one thing for another. Given that the use of this statement following the second withdrawal by two storekeepers is followed by the statement that !utur sut inni hudda!, '(with) the balance (which would in this case seem to be the full amount) no exchange was undertaken', the statement ta!!up-ikmar inni pirru !arrafda perhaps does not refer to the sut procedure but to a different context of exchanging or selling off the foodstuffs to others. Another possibility is that tafsup refers to the (group of) storehouse officials, though then the syntax of the texts might be a complicating factor. As !arra- (or !ara-) more literally means 'to cut off', it might in fact mean that nothing of the commodity that these officials had withdrawn was disposed of (nothing was 'cut off' from

182

The same text has huddamanba with a plural referent earlier, so either plural -ba would have been expected here, or the group of officials is construed singularly. 183 NN 2208:20 and NN 2355:18 both read hube inni kitH. 184 See also PF 1973, 1975-1977. 185 Hallock 1969, 580.

Hall~ck (1969, 745a) translated pirru as 'together', though noted that its literal meaning is uncertain. Stolper kindly suggests to me that its use in pirru mufa- in Treasury texts may refer to the calculation (mufa-) of the conversion of one thing into another (cf. the equivalency of one sheep and three shekels of silver, and one marrif of wine and one shekel of silver). 186

ELuSIVE SILVER IN THE AcHAEMENID HEARTLAND

it). 18 7 In administrative contexts (both Achaemenid and earlier) sarra- also has the meaning 'to apportion, to disburse' (i.e. in the sense of 'cutting off' a part from something in one's possession and disbursing it for someone or something else, rather than in the sense of taking something away). 188 In any case either interpretation would seem to make sense in the light of the statements that follow these phrases in PF 1974. The statement in connection with Mauparna that af-e inni kitif would make sense if it means that because he died the goods that he had withdrawn were lost to the storehouse (and thus to the institution). 189 To conclude, despite one or two complicating factors the most plausible interpretation would seem to be that af refers to the goods that were withdrawn and were subject to either being used in an exchange or to be disposed of in another way. What still requires explanation is what is described by the statement x !aumarraf harak. This explanation must focus both on the actual meaning of the words and on the contexts in which and possibly the reason(s) why the expression occurs. It has already been noted above that the etymology of the word saumarraf remains unclear. It seems most likely that the word written spr(?) in the Aramaic Bowman texts that has previously been connected with it has a different (Semitic) background. Therefore there is little point in, for instance, deriving the nature and shape of saumarras- assuming that the word refers to objects- from the objects on which the self-descriptive word spr(?) is written. It has also been pointed out that in the context of the Fortification archive hara- is commonly used to denote the impressing of seals, 190 and that there is no support whatsoever for the interpretation that takes the phrase saumarras harak to mean that saumarraf were 'brought in'. The point, however, is precisely that it is perfectly possible that something was (ultimately) brought in- since as Aperghis rightly pointed out !aumarraf feature instead of another commodity - 191 but that this is simply not what is recorded in the texts. Despite the fact that hara- has nothing to do with 'bringing in', it does seem to make sense to assume that since the officials involved formally withdrew commodities from institution-owned storehouses, they owed the institution something (of a value that was at least deemed equal) in return for them. It is probably the case that the normal procedure was to try and dispose of those commodities through exchange, and that whatever was thus acquired essentially became part of the institution's assets in the sense that it was formally recorded as 'intake' and, as some texts show, assigned to individuals under the administration's purview. Since it is clear that 'one saumarras represents the value of one-fifth of a shekel, that multi187 Note, however, PF 1976:6ff., where it is first stated that taffup-ikmar inni pirru farraf, yet subsequently that hupirri, 'he' (referring to the storekeeper who was previously mentioned in the text), exchanged (sut huttas} a part of the barley he had withdrawn for a female donkey, and similarly PF 1977:5ff. 188 The meaning of (forms based on) fara- is discussed in Henkelman 2008, 128 fn. 285 ; cf. the relevant entries in Hinz/Koch 1987, II, 1134, 1138; and see Stolper 1984, 104. 189 Perhaps this explains why PF 1978:14 reads df-e in1-ni SA-ri (the text reads mi-ni though Hallock [1969, 582 note e) points out that mi is written over an erasure and perhaps an error for in), lit. 'his af was not there' . Here, too, an individual who had made a withdrawal is said to have died (ha/pika). Note, however, that if -ri in sA-ri is taken to represent the suffix -r this would seem to imply that here aJ is animate. 19° And earlier as well, see, e.g., Hinz/Koch 1987, I, 625 s.v. ha-ra-qa. 19 1 Aperghis 1997, 284-285 .

ELUSIVE SILVER IN THE AcHAEMENID HEARTLAND

M. TAMERUS ple faumarraf 'units' feature when the value is greater than 0.2 shekel, and that given the use of hara- it is likely that here too something is impressed, the most plausible solution is that faumarraf either refers to something that could be impressed into something else or refers to the impressed object itself. In this respect Giovinazzo was right to point at the relevance of the common use of hara- to refer to the application of seals, though her conclusion that faumarraf 'serait une valeur d' equivalence des differents types de rations attribuees selon la categorie a laquelle appartiennent les salaries, selon leur sexe, leur age ou leurs responsabilites', and that it 's' applique egalement au type de recipient destine aux differentes denrees' is incorrect. 192

As for contexts and reasons, it is clear that normally faumarraf occur when either a part of the withdrawn commodities remained unused, or none of it was used in any exchange at all. But while the faumarraf texts seem to be concerned mostly with surplus disposal, on closer scrutiny things are a little more complicated. 193 An important question is whether an initial intention (or order) to seek to exchange is an absolute necessity for faumarraf to come into play at all. For instance, according to PF 1980:5-10 storehouse officials first withdrew an amount of wine. This is one of the texts in which the reference to a withdrawal is not directly followed by a reference to what the officials did with it, but where a part of the initially withdrawn amount is subsequently withdrawn by another individual. Here, a certain Kurima withdrew 60 1. of wine from the 227 1. that the storehouse officials had taken. There is no reference to what Kurima did or did not do with it, and all that is recorded is that 30 faumarraf were 'impressed', conforming to the 5: 1 faumarraf: marrif rate. It may not necessarily be the case that the fact that this individual took an amount means that he would have gone on to attempt to exchange it. Instead, here 30 faumarraf may simply have been 'impressed' (at a location that goes unmentioned) in return for Kurima's withdrawal. If this is indeed the case, this passage might in fact record a sale of wine. NN 2355 suggests that not all transactions recorded in faumarraf texts served to dispose of storehouse surpluses. According to this text 1,820 1. of wine had been withdrawn and subsequently used to acquire a female slave. In this case it is explicitly stated that this had been ordered by the king (HALESSANA-ik-mar if-fe-ra-ka4). Possibly surplus disposal through exchanges (and perhaps selling to individuals) did not constitute the phenomenon reflected in the faumarraf texts itself, but was a (substantial) part of it. While the expression sut hutta- is commonly translated 'to exchange', it should be pointed out that the term sut derives from 192

Giovinazzo 1993, 127. A couple of newly read texts which M.W. Stolper kindly brought to my attention show that Jaumarraf occur in contexts not attested in the texts presented in this paper. Fort. 1394-101 is a livestock (GUDMES) account recording for a number of consecutive years several categories of withdrawals of livestock, with separate entries for each individual involved. The column that gives the total number of withdrawn animals is followed by one in which an amount/number of Jaumarraf is recorded. It remains unclear for now why faumarraf appear here. One could surmise, for instance, that this is a rather abbreviated way of recording that the institution had to be reimbursed for the animals withdrawn (for different purposes) from its stocks, but that remains guesswork. In any case it is to be noted that the Fortification archive contains many animal accounts yet none apart from this one is known to mention Jaumarraf. Fort. 2182-102 is a short account that records a number of small livestock (UDU.NITAME5). In connection with them the word Jaumarraf occurs, but because the tablet is broken at this point the context remains unclear. 193

suda-, which as already noted earlier means 'to seek, request'. Therefore one should consider the possibility that a more literal translation of sut hutta-, namely 'to make a request' > 'to seek to

(acquire through) exchange', more accurately reflects the phenomenon it refers to. This would cover the occurrences of surplus disposal, selling off to individuals, and acquisitions directly ordered by (or in name of) the king. There may have been times when for one reason or another no exchange was possible, or the preferred animals or commodities happened not to be available, so that the withdrawn goods were otherwise disposed of. It is at this point, but apparently also in cases like Kurima's individual withdrawal, where faumarraf comes in. In other words, if this analysis is correct, the texts discussed here refer both to faumarras having been obtained in lieu of commodities or animals and to saumarras obtained through direct sales. The texts presented in §6.3 provide some additional information on the act described by the phrase Jaumarraf harak. For instance, in NN 0146:32 it is stated that 236 2/3 faumarraf appu-ma harakka. The translation of this phrase naturally depends both on the syntax ~f the tex~ and on one's interpretation of the texts in general. One could take appu-ma to mean by them (reflecting the instrumental use of the postposition -ma), so that the individuals involved in the transaction are the referent of appu, Alternatively the postposition indicates purpose, so that appu-ma would mean 'for them' . 194 In that case appu-ma would essentially have the same function unra ('each') has in several other texts that read (meni) unra x saumarras harak(ka), though 195 it must be noted that this would mean that unra is not the grammatical subject of harak(ka). One text uses a form of another stem in the place where the other texts use hara-, and this might be suggestive as to what might be referred to. Thus, NN 2358:18 uses kazap (a 3.pl. conjugation II form of the stem kaz[z]a-), the full statement reading meni unra 22 saumarras 1-SU (= 113) saumarras-ma kazap. In the interpretation that follows (and in the text as presented above in §6.3) this is taken to mean 'then by?/for? each 22 faumarraf and one-third of a saumarraf were struck' . 196 It is, however, only fair to point out that the syntax of this text is a complicating factor, since it makes sense to assume that the postposition -ma is to be construed with kazap. And since -ma is unusual here and one would rather expect -na to follow the second instance of saumarraf if the postposition was part of the fractional expression, and since it would not be fair to assume all too easily that it might be a scribal error, the postposition requires an explanation. One option would perhaps be to take -ma, construed with kaza-, to refer to the striking made on the object in question, but it admittedly remains difficult syntactically. Another option would be to change the interpretation of the entire expression and take unra to be the subject of kazap here (perhaps with instrumental -ma, so that it would literally read something like 'and each were struck with .. .') and of harak(ka) in the other texts. However, See Hallock 1969, 668 s.v. apma, 'for/to them', taking ap-pu-ma to be a variant of ap-ma. The use of the word unra in this phrase- which when it occurs here is consistently written without any postposition - is grammatically peculiar. If it refers to the officials and served to record tha: so~ethi~g was done for them one would perhaps expect unra-na (compare PF 1986:29-30, where unra IS wntten m the Jaumarraf harak phrase, but when subsequently the text seems to record that a number of Jaumarraf were withdrawn/taken away from their balance, it reads Ju-tur un-ra-na). 196 It is not clear why most texts use harak, with inanimate referent in the interpretation given here, but this text uses a form in -p for which one would normally expect a plural animate referent.

194

195

286

ELusiVE SILVER IN THE AcHAEMENID HEARTLAND

M. TAMERUS

in several texts unra is not used, and one sees the one clause with the expression inni kitif appa sut huddamanba followed directly by meni x faumarraf harak, as in NN 2355:6-8, for instance. Although given the syntax of the text the expression with kazap is perhaps not simply a variant with precisely the same meaning to the one that uses harak(ka) (see the discussion below), it is assumed here that it refers ultimately to the same phenomenon. Elsewhere conjugation forms of kaza- occur in administrative contexts as well as in royal inscriptions. Although they occur in a rather wide variety of contexts, requiring different specific translations, it seems that 'to hammer, beat, smite, forge' ultimately covers all contexts. In several cases forms built on this stem denote destruction (doubtlessly deriving from 'beating/striking down' something or someone), particularly but not exclusively in royal inscriptions. 197 In administrative contexts- including the Persepolitan archives but also the Neo-Elamite Acropole texts from Susa- the stem kaza- is known in particular from compound terms denoting various kinds of smiths, such as AN.BAR kazzip, 'blacksmiths', KU.GI kazzip, 'goldsmiths' and indeed KU.BABBAR kazzip, 'silversmiths'. 198 There are additional relevant occurrences in a number of Acropole texts. These refer to the forging of particular items, usually in connection with metal. For instance, S 49:10-12 records that 1 li-gi AN.BARMES 7 MA.NA Y2 GISsir-ri-me zik0 -kur-tium ANpir-mu-na a-ah kaz:o-za-ak, '1 peg, 199 (with) iron, 7 Y2 minas (being) its weight(?), 200 for 197

For instance, a difficult passage in PF 1857, a letter about the handling of grain at various villages. It was sent by a certain Ukkama and his associates to Parnakka, the Persepolis administration's director. The sender notes that AShu-ma-nu-if ASkan-sa-an hi-fe hu-be- 'ma' SE.BARMES ASHAL.AMES AStur-ra-tan-nu-if ka4-za-if, lit. '(at) a village, Kansan (being) its name, there, barley of irrigated land, turratannuf "struck it down" (i.e. "destroyed it")'. Hallock took turratannuf, which is attested only here, to be a geographical name, though translated with a measure of doubt '(at?) the Turratannus' (1969, 511 [text], 765 [glossary: 'GN']). Hinz/Koch (1987, I, 372 s.v. h.tur.ra-tan-nu-is} suggested- though equally with a question mark- that the word represents Iranian *fJratanya-, 'Unwetter, Hagelsturm, Gewittersturm', a meaningTavernier (2007a, 510 [5.3.4.56]) recently reiterated, translating 'thunderstorm'. In Achaemenid royal inscriptions forms built on kaza- occur meaning 'striking/smiting' in the sense of destroying, defeating, suppressing. Thus, in DSee it is used in the phrase sa-ap [DI 5ib]- 'ba '-ak-ra DIS 'if'-tuk-ra in-ni ir ka4 -za-ma-ak, 'so that by the strong the weak shall not be smitten' (c£ Hinz/Koch 1987, I, 457 s.v. qa-za-ma-ak). XPhe contains the phrase hu-be A 5da-a-ya-ma DISu ka -za, 'I smote those lands' (c£ Hinz/Koch 1987, I, s.v. qa-za). In this as well as the previous case the parallel Akkadian version uses a form of the verb daku, 'to kill, defeat', namely Id idukku and aduk, respectively (though DSee uses a conjugation II form with passive meaning whereas the Akkadian text has an active form). The Old Persian versions ofDSe and XPh use jantiy and ajanam, respectively (Schmitt 2009, 126, 167), active present indicative and imperfect forms of the verb *)an-, 'to strike, kill, defeat' (Schmitt 2014, 179). 198 As Hallock (1969, 712b) translates: 'lit. "smiter(s)"'. The use of kaza- in these compound terms certainly derives from the fact that smiths commonly worked their material by hammering it. 199 See Stolper 1984, 53 for a Middle Elamite attestation of li-giMES in TTM 24:2, and reference to the equivalency of ligi or like with Akkadian sikkatu, 'peg, bolt' (sikkatu A in CAD S, 247-251) in the Achaemenid royal inscriptions DPi and XPi. See also the discussion in Basello 2012, especially 24-30 on Middle and Neo-Elamite evidence. According to Basello (2012, 30) it is also possible that like refers to 'tile knobs in metal (i.e. knob-plates) [... ] used as a decoration for royal and/or public/institutional buildings.' 200 Hinz/Koch (1987, II, 1091 s.v. sir-ri-me) translate 'Gewicht' and quote Eilers (1090 s.v. sir-me) connecting sir- to the stem sira-, which he translates 'aufhangen' (as do Hinz and Koch). Note, however, the Middle Elamite references cited in the entries si-ra-df and sir-a-df in Hinz/Koch 1987, II, 1088, which they translate 'er wog aus'. See especially the discussion in Stolper 1984, 12-13; and compare Powell's discussion (19871990, 510) of Sumerian gin and Babylonian fiqlu, 'shekel', for an example in another language of one verb

· )' Another example is . h d the ziggurat of Pirmu, therewith20I was fo_rged (l.e. a~mer~ ' smtttehn . k 100 80 zik . h d 1 MA NAAN BARMES kur-mdn DISbar-rt-man-na u-ma- a4 + o h S 114, w lC rea s · · . · · d 180 zikti · · ah kaz za ka '1 mina of iron, by allocation of Barnmanna lt was acqmre · tt-tp ao- - 4> were therewith forged (i.e. hammered, smitten)'. If the phrase Jaumarraf harak indeed means that something referred to by •the term. h h . 1· ·tl mentioned) then saumarras h k , d k zan might Jaumarraf was impressed (with somet ing t at ts not exp lCl y rn · NN 2358 can be taken to refer to the same phenomenon. ara an. ~ r . kazar m f, h · · f a neganve Image mto two different aspects of it: the former re ers to t e tmpressmg 0 h d the latter refers to t e act cover . . d f) an object (regardless of what material th~t obJeCt ts ma e. o ' an ed here from two different of hammering/punching into a metal obJe~t. Thus, what~~ convey . h h he ob'ect in ers ectives ma be the marking of an obJect. The question, then, lS .w et er t J. was silver. Since Jaumarraf is not explicitly silver, other (indirect) evidence would be required to estabh~h thts.. ~ a~IS does not mean t texts resented above the term Jaumarraf is preceded by the etermmanve it necessarily an object made of wood. This determinative is generallk u~dl to mark;~~y rial objects. 202 To take one example, in PT 1963-3:4, 8, 21, 22, 23 it mar t e ogograp IC written word f,or 'SI'lver' , GISKU'MES ·

~ue~ion

in~eed

.referr~d ~mfth; tt~:~si~s:le:.gt~!

w:

As for the question whether the object called Jaumarraf was ~n~eed ~ilver t~~ p~~a~:r~F 1980:14-20 is potentially revealing. Here it is recorded that 92 trttba V~~ ~76 v .) b e uen~ were withdrawn from a storehouse, and that a certain Zizza at the place. Stsu anu~· su s l q ne ly withdrew the whole amount. Since nothing of it was used to acqmre '~omet mJ, el::t;ad ex ects to find on analogy with the other texts that 460 Jaumarraf were tmpre.s~e ·. . ' . . . " f 60" (and) 160 Jaumarraf were issued for lt (ht partmak p d h , the text recor S t at one Strtmazst O · d h' . l" that • • 11 19-20) 203 As Aperghis has pomte out, t 1s tmp tes . . . 1 smmazst 6 0-na 160 saumarras, . . al f h k l . . . als 300 Jaumarraf 204 Since five Jaumarraf represented the v ue o one s e eh' one smmazst equ · al f · shekels i e to be t e it is tempting to take one sirimazsi 'of 60' to represent the v ue o stx~ d . PF .1980·14-20 uivalent of a mina 205 It would seem that the amount of barley recor e m l ·. h' eq · Wh · clear is why - at east m t ts was such that it was worth more than a mina. at remams un

. b d h f, rmer) noting that 'the denoting both 'hanging' and 'weighing' (with t.he latter ~e~n~ng ase on ~ f:o; a baiance'. Akkadian probably means "weight," i.e., somethmg suspen e Y some mean . . . -1 . e eneral oblique resumpnve pronoun 201 On the interpretation of a-ah m given contexts as a ~on doca%v. , gd in particular the discussion in see Hallock 1973, 148 fn. 4; Hinz/Koch 1987, I, s.vv4. a- an a-h a,han saseeJ·n S 114 a-ah is a resumptive 1 · · for S 9 assumes t at ere Stolper 2004a, 76-77. Th e trans anon given . h b' . . (the li-gi) is forged, hence it takes ronoun referring back to the material with whiCh t e o yect m qu~snon ~he text to refer to a peg made of iron (rather than refernng to an Iron peg). Henkelman 20 11 b, 617. h p r . al. d . b use at least in texts from t e ersepo Jtan The etymology of sirimazsi is unclear. It IS norn: Jze -"';azskt) eCca 'd ri·ng the context of weight measures · MA's· h h al n?fl?', (as e g m maz-zt- a 4 • ons1 e archives the sign as t e v ue '.'~ GIS : ''. bl ·n 'weight', inS 49, quoted above. it seems pertinent to compare it to strrt m str-rz-me, presuma Y meani g

202 203

204

Aperghis 1997, 285.

205

Aperghis 1997, 285.

288

particular context- any amount smaller than a mina apparently had to be expressed in multiples of the faumarraf of 0.2 shekel value.

It is particularly the combination of this sirimazsi, the 0.2 shekel value of faumarraf and the explicit references to 'impressing' and (once) 'striking' Jaumarraf that makes the idea plausible that faumarraf refers to something made of precious metal. Yet it need not have been strictly silver coinage, as Aperghis suggested with reference to the archer coins from the time of Darius and the use of a coin to seal Fortification tablets. 206 In the discussion on coinage above (§3) reference was made to Babylonian ginnu silver, which according to Jursa refers to either a distinctive mark on or a particular form of silver that provided instant information about the silver's purity. It would not be wise to suggest any kind of direct relation between the two phenomena. But there would be merit in considering the possibility that 'impressing' faumarraf refers to a process that entailed weighing and measuring out silver in fixed units that were subsequently marked to convey certain information. On the basis of the texts it cannot be told where exactly this weighing and marking - i.e. the acquiring of silver for goods disposed of- would have taken place. As for the reason for marking, one option is that silver was marked to guarantee its purity, possibly on the basis of weight. It is, however, difficult to imagine that silver units acquired locally in the procedure implied in the faumarraf texts were of a consistent purity, unless there was a market that was closed to such an extent that no silver of insufficient purity circulated. It is equally unclear for now how these marked units of silver would relate to the fractional amounts of weighed silver with which workers were remunerated according to texts from the Treasury archive. It is thus suggested here that the pertinent expression in the faumarraf texts refers to the 'marking' of silver objects. The primary purpose of the phrase was not, however, to record the act of marking, but to record that the required amounts of silver qualifYing as faumarraf had been acquired- or could at least be booked as such. If the above interpretation is correct, the faumarraf texts might reflect another example of silver circulating in the Persepolis economy. Thus, there may indeed have been a market where commodities could essentially be sold for money - not minted, but money nonetheless. The texts do not, however, provide information on what was subsequently done with the silver, i.e. whether local storehouses kept stocks of silver or the silver was ultimately collected centrally. Another question that cannot be answered easily is who were the immediate sources of the silver thus acquired. Although one does not expect institution-owned commodities to have been sold for institutional silver, one possibility is that individuals who held silver because they were remunerated with it (e.g., kurtaf workers or soldiers) used it to buy directly from storehouse officials, so that the silver truly circulated within the confines of the Persepolis economy.

As for the 'impressing' of faumarraf it is interesting to point (with caution) at the Achaemenid-period presence of small silver bent bars in the Chaman-i Hazuri hoard from the Kabul region and bent bar and cup-shaped 'coinage' from the northern and northwestern part of the Indian subcontinent. 207 On the basis of the Kabul hoard context the mid-fourth century 206 20 7

Aperghis 1997, 285. See the discussion in Cribb 2003, 15ff.

ELuSIVE SILVER IN THE AcHAEMENID HEARTLAND

M . TAMERUS

BCE has been taken to be the time when such bent bars could be hoarded ~or the ~rst ti_me, but according to Cribb it 'would not be impossible to push the dat~ of thelr first Issue mt~ ' 20s Notably the bent bars from the subcontment as well as the cup t h e 1ate fifth century BC · ' d ·bl d h ed ob. ects that are thought to have been based directly on the former an possl y ate tury as well 209 are punch-marked. Interestingly, as Cribb notes, what both s aph 1 )fifth tot e ate cen ' . h 'h d of them have in common as opposed to later punch-marked coins lS that t ey av~ a en~m. . h. h · 1 des fractional coins'.2iO Cribb, who states that the Ind1an obJects mauon system w 1c me u f, h · ' 211 'were issued according to a system which suggests a centralised issue sy~tem or eac senes ' consistently calls these and the Kabul objects coinage. This means that~~ one ~oull~ ~c~p~t~e .bili that the faumarraf were something similar and that these o JeCts s ou m ee e ~~::~fiedr:s coinage- albeit one clearly different from the Achaemenid c~inage.gen~r~y folun~ in the western part of the empire - the statement that no coinage was m use m ~ e eart ~n. at the time of Darius would have to be emended somewhat. For the moment, owever, lt lS safest not to go this far. Accordin to Cribb ' [t]he Chaman-i Hazuri hoard suggests that Achaemenid mo~etary ractice was br~u ht into the empire's eastern provinces in Afghanistan as another part of lts lo~ ~ administratio~ and, more specifically, he notes that ' [t] he evidence of the C~aman Ha~u~~ hoard suggests that the prototype of India's first coinage, i.e; the bent ba.r comag~, w~ ~tse f an ada tation of a local Afghan imitation of Greek coinage. 212. Co~cludmg on t e ongm o the In!an coina e tradition he suggests that it 'originated as an idea lmp~rted from t~e Greek . A h g · d Afighanistan' 213 Yet it need not be necessary to mvolve the mfluence world v1a c aemen1 · b fi h h and · · the Greek world With regard to the o jeers rom t e nort ern o f monetary pracuce m ·

208

Cribb 2003, 17-18. . , h k d bent The silver objects from Bhir Mound, Taxila, 'similar in technique :m~ ':"eight to t~e ~unc -~a:n eAchae[J the Kabul re ion (Callieri 2006, 12), are according to Calhen (zbzd.) clear evld~nce o . . , Magee et al (2005 717) argue, however, that whde the appearance bars rom . h ~ id presence m t e 1ronuer regwn · · ' .d M · · _ punch-marked silver at Taxilations indicate knowledge of Aramatc and that aurhy a~ pfl a f th Achaemenids' '[s]uch evidence is of e , b ' hasized as testimony to t e m uence o h 1 scu pture - as een emp . . . . . . because it rovides no insight into indigenous in to serious the alleged (material) link questionable. responses to tmpe~ s~ategt~ o~~nl :o~d and th~ punch-marked objects from Taxila, indeed despite the

209

:ret~ese

object~

relev~~ce

~ongside t~e t7~t~~:hti~:~t~~;t:a~~:e:u::~ed :;z:::a~:~~fd

e~amt~mg ;n~t'e~o:~:;l1rl~~relevant t~ke

~:::~: ~e~:7: n::v;l:c~ ~0 ind~cate that Taxila was necessarily part of the Achaemenid administrative network at this time (Magee et al. 2005, 717 fn. 68). . c h h of these and related objects. 2 1° Cribb 2003, 17. See there, PP· 16 - 17 wr P otograp s 2 11

212

Cribb 2003, 17.

Cribb 2003, 18. d 'h · l"ke Indian curCribb 2003, 18. Bivar (1988, 195) went so far as to suggest with regar to t es; 1·7go\ I . ts in the 'd · c. h me economic system that produced a currency o SI ver ar-mgo ' rencies that they enve aom t e sa . d h . B C ' which 'evidently continued in use · I d · the e1ghth an sevent centunes . · , f ~ph~e Io Assr~an~:~::o an~r;:t~r in that of the Achaemenids.' See Bivar 1971 (especially the discu:sion o~ m~ 1e0;~;0~) ~: 'bar-in~ors' found in a 'Median'-period hoard of sil~er objects at Nus-i Jan. On ent an ~~aight silver bars found in another Afghan hoard (Mir Zakah) see B1var 1982.

213

ELUSIVE SrLVER rN THE AcHAEMENID HEARTLAND

M. TAMERUS northwestern parts of the Indian subcontinent Achaemenid Mghanistan may well be a direct source, but perhaps local practice there did not imitate Greek coinage but was based ultimately on the practice of using weighed silver for money in the imperial heartland, extending towards the east as Achaemenid administrative practices did, too. 214

7·

CoNCLUSION

It remains difficult to assess the economy of the heartland in terms of notions such as 'monetization' and 'silverization'. This does not, however, mean that there is nothing to be said on the basis of the texts discussed in the foregoing sections. In addition to administrative texts that refer to transports of silver and 'treasures', simply attesting to the presence of silver in and its movement throughout the heartland region, as well as the valid assumption that wool production and trade may have increased the presence and circulation of silver, there are some indications that there were markets where silver was in use. In the first place the central treasury disbursed proper amounts of silver on at least a couple of occasions. If silver disbursements or gifts such as those received by the group of individuals in PT 04 resulted in those people having more silver than they needed to provide for subsistence, they could hoard it or use it to purchase from storehouses surpluses or otherwise acquire goods. The same applies to the craftsmen, herdsmen and other workers who according to Treasury texts were (partly) remunerated in silver. The fact that they were means that there must have been places where they could spend their income. It is very plausible that such payments were not limited to the period for which Treasury texts are currently attested. If the Treasury texts do witness an increase in silver payments, perhaps facilitated by the profits from wool production -which is admittedly something that cannot actually be checked - there may in this respect have been an increasing silverization of the heartland economy. The second part of this article treated the texts involving faumarraf in detail. The idea, posited by Aperghis, that according to these texts silver was acquired seems the best answer to the question what went on in the procedures referred to in these texts. There was much more to these texts that required analysis and explanation compared to what had hitherto been written about them, and for this reason detailed philological discussions accompanied their interpretation. It is argued here that these texts refer to some kind of marking of objects called faumarraf, and in the light of this reference was cautiously made to the phenomenon of punch-marked silver bent bars found in Achaemenid-period contexts towards the east. Although it would go too far at this point to simply state that there must have been direct connections, given the fact that the evidence from the heartland does remain scarce and complicated, it may be clear that careful and detailed scrutiny ofPersepolitan texts like those discussed here has the potential to provide important insights into economic aspects of the Achaemenid period.

214

Though it is good

to

take Magee et al.'s (2005) critical views into consideration.

BIBLIOGRAPHY

- 1988: The Indus lands. In: J. Boardman et al. ALRAM, M. 2012: The coinage of the Persian (eds.), The Cambridge Ancient History IV: Empire. In: WW Metcalf (eds.), The Persia, Greece and the Western Mediterranean Oxford Handbook of Greek and Roman c. 525 to 479 B. C. Cambridge, pp. 194-210. Coinage. New York, pp. 61-87. BoWMAN, R.A. 1970: Aramaic Ritual Texts from ArERGHIS, G.G. 1997: Surplus, exchange and Persepolis. Chicago. price in the Persepolis Fortification tablets. In: J. AndreauiP. BriantiR. Descat (eds.), BRANDENSTEIN, w I MAYRHOFER, M. 1964: Economie antique. Prix et formation des prix Handbuch des Altpersischen. Wiesbaden. dans les economies antiques. Saint-BertrandBRIANT, P. 2002: From Cyrus to Alexander: A de-Comminges, pp. 277-290. History ofthe Persian Empire. Winona Lake. - 1999: Storehouses and systems at Persepolis: - 2010: Suse et l'Elam dans I' empire achemenide. Evidence from the Persepolis Fortification In:]. Perrot (ed.), Le Palais de Darius aSuse. tablets.]ESH042, pp. 152-193. Une Residence royale sur la Route de Persepolis ARFAEE, A. 2008a: Persepolis Fortification Tablets. aBabylone. Paris, pp. 22-48. Fort. and Teh. texts. Tehran. BRosrus, M. 1996: Women in Ancient Persia (559- 2008b: PT lOa, collated and completed. 331 BC). Oxford. ARTA 2008.001. CAHILL, N. 1985: The Treasury at Persepolis: - 2008c: The Geographical Background of the Gift-giving at the city of the Persians. A]A Persepolis Tablets (Diss. University of 89, pp. 373-389. Chicago). CALLIERI, P. 2006: India m. Relations: BASELLO, G.P. 2012: Doorknobs, nails or Achaemenid period. Encylopaedia Iranica pegs? The function(s) of the Elamite and 13, pp. 10-13. Achaemenid inscribed knobs. In: G.P. CAMERON, G.G. 1948: Persepolis Treasury Tablets BaselloiA.V. Rossi (eds.), DARlOSH Studies (OIP 65). Chicago. I!. Persepolis and its Settlements: Territorial - 1958: Persepolis Treasury tablets old and new. System and Ideology in the Achaemenid State. Naples, pp. 1-66 and pis. 1-:XX. ]NES 1713, pp. 161-176. BEAULIEU, P.-A. 2013: Arameans, Chaldeans, and - 1965: New tablets from the Persepolis Treasury. Arabs in cuneiform sources from the Late ]NES2413, pp. 167-192. Babylonian period. In: A. BerlejungiM.P. CRIBB, J. 2003: The origin of the Indian coinage Streck (eds.), Aramaeans, Chaldaeans, and tradition. South Asian Studies 1911, pp. 1-19. Arabs in Babylonia and Palestine in the First FISHER, M.T. I STOLPER, M.W 2015: Achaemenid Millennium B.C. (LAOS 3). Wiesbaden, Elamite administrative tablets, 3: Fragments pp. 31-55. from Old Kandahar, Mghanistan. ARTA BrANCH!, G. 1986: Le signe SA3 en Elamite. In: 2015.001. L. de Meyer/H. GascheiF. Vallat (eds.), GARIBOLDI, A. 2014: Elements of 'globalization' Fragmenta historiae elamicae: Melanges in ancient Iranian numismatics. In: M.J. o.lferts aMJ Steve. Paris, pp. 255-261. Geller (eds.), Melammu. The Ancient World BrvAR, A.D.H. 1971: A hoard of ingot currency in an Age of Globalization. Berlin. of the Median period from Nush-i Jan, near GARRISON, M.B. I RooT, M.C. 2001: Seals on the Malayir. Iran 9, pp. 97-111. Persepolis Fortification Tablets. Vol 1: Images - 1982: Bent bars and straight bars: An appendix of Heroic Encounter. 2 Vols (OIP 117). to the Mir Zakah hoard. Studia Iranica 11, Chicago. pp. 49-60.

ELuSIVE SILVER IN THE AcHAEMENID HEARTLAND

M. TAMERUS GERSHEVITCH, I. 1969: Iranian nouns and names in Elamite garb. Transactions of the Philological Society 1969, pp. 165-200. GIOVINAZZO, G. 1993: Les Saumarra.S clans les textes de Persepolis.A/ON53, pp. 121-127.

-

2011 a: Of tapyroi and tablets, states and tribes: The historical geography of pastoralism in the Achaemenid heartland in Greek and Elamite sources. BICS 54/2, pp. 1-16. 2011b: Cyrus the Persian and Darius the Elamite: A case of mistaken identity. In: R. Rollinger/B. Truschnegg/R. Bichler (eds.), Herodot und das persische Weltreich. Akten des 3. internationalen Kolloquiums zum Thema "Vorderasien im Spannungsfeld klassischer und altorientalischer Oberlieferungen. Innsbruck, 24.-28. November 2008. Wiesbaden, pp. 577-634.

GRILLOT, F. 1974: Review of R.T. Hallock, Persepolis Fortification Tablets. RA 68, pp. 179-183. GRILLOT-SUSINI, F. I HERRENSCHMIDT, c. I MALBRAN LABAT, F. 1993: La version elamite de la trilingue de Behistun. Une nouvelle lecture. fA 254, pp. 19-59. HALLOCK, R. T. 1960: A new look at the Persepolis Treasury tablets. JNES 19/2, pp. 90-100. - 1969: Persepolis Fortification Tablets (OIP 92). Chicago. - 1973: On the Middle Elamite verb. JNES 32/1-2, pp. 148-151.

2012: The Achaemenid heartland: An archaeological-historical perspective. In: D.T. Potts (ed.), A Companion to the Archaeology of the Ancient Near East. Oxford, pp. 931-962.

1978: Selected Fortification texts. CDAFI 8, pp. 109-136, 1985: The evidence of the Persepolis tablets. In: I. Gershevitch (ed.), The Cambridge History of Iran II. Cambridge etc., pp. 588-609.

-

2013a: Administrative realities: The Persepolis archives and the archaeology of the Achaemenid heartland. In: D.T. Potts (ed.), The Oxford Handbook of Ancient Iran. Oxford/New York, 528-546.

HAWKINS, S. 2014: Greek and the languages of Asia Minor to the Classical period. In: E.J. Bakker (ed.), A Companion to the Ancient Greek Language. Chichester, pp. 213-227.

-

2013b: Persepolis tablets. In: R.S. Bagnall et al. (eds.), The Encyclopedia of Ancient History. Maiden, pp. 5179-5181.

-

forthcoming: Local administration: Persia. In: B. Jacobs I R. Rollinger (eds.), A Companion to the Achaemenid Persian Empire. Chichester.

HENKELMAN, WF.M. 2005: Animal sacrifice and 'external' exchange in the Persepolis Fortification tablets. In: H.D. Baker/M. Jursa (eds.), Approaching the Babylonian Economy. Proceedings of the START Project Symposium Held in Vienna, 1-3 July 2004 (AOAT 330). Munster, pp. 136-165. -

-

2008: The Other Gods Who Are. Studies in Elamite-Iranian Acculturation Based on the Persepolis Fortification Texts (Achaemenid History 14). Leiden. 2010: 'Consumed before the king'. The table of Darius, that of lrdabama and Irta.Stuna, and that of his satrap, Karkis. In: B. Jacobs I R. Rollinger (eds.), Der Achiimenidenhof Akten des 2. internationalen Kolloquiums zum Thema "Vorderasien im Spannungsfeld klassischer und altorientalischer Oberlieferungen. Landgut Caste/en bei Base!, 23.25. Mai 2007. Wiesbaden, pp. 667-775.

I

-

]ONES, C.E. I STOLPER, M.W 2006: Achaemenid Elamite administrative tablets, 2: The Qa~r-i Abu Na~r tablet. ARTA 2006.003.

I KLEBER, K. 2007: Babylonian workers in the Persian heartland: Palace building at Matannan during the reign of Cambyses. In: C. Tuplin (ed.), Persian Responses. Political and Cultural Interaction with(in) the Achaemenid Empire. Swansea, pp. 163-176.

HINZ, W 1973: Neue Wege im Altpersischen. Wiesbaden. -

1975: Altiranisches Sprachgut der Nebenuberlieferungen. Wiesbaden.

-

I KocH, H. 1987: Elamisches Worterbuch. 2 Vols. Berlin.

HoFTIJZER, J. I JoNGELING, K. 1995: Dictionary ofthe North-west Semitic Inscriptions. 2 Vols. Leiden etc. ]oNES, C.E. I SEUNGHEE, Y. 2011: From the Persepolis Fortification Project, 3. The first administrative document discovered at Persepolis: PT 1971-1. ARTA 2011.003. JuRSA, M. 2003: Observations on the problem of the Median "empire" on the basis of Babylonian sources. In: G.B. Lanfranchi et al. (eds.), Continuity of Empire: Assyria, Media, Persia. Padova, pp. 169-179. 2008: The remuneration of institutional labourers in an urban context in Babylonia in the first millennium BC. In: P. Briant I WF.M. Henkelman I M.W Stolper (eds.) , L'archive des fortifications de Persepolis: Etat des questions et perspectives de recherches: Actes du colloque organise au College de France par la "Chaire d'histoire et civilisations du monde achemenide et de !'empire d'Alexandre" et le "Reseau international d'etudes et de recherchs achemenides" (GDR 2538 CNRS), 3-4 novembre 2006 (Persika 12). Paris, pp. 387-427. 2010: Aspects of the Economic History of Babylonia in the First Millennium BC. Economic Geography, Economic Mentalities, Agriculture, the Use of Money and the Problem of Economic Growth (AOAT 377), with contributions by J. Hack!, B. Jankovic, K. Kleber, R.B. Payne, C. Waerzeggers and M. Weszeli. Munster. KAuFMAN, S.A. 1974: The Akkadian Influences on Aramaic (Assyriological Studies 19). Chicago. KAWASE, T. 1980: Sheep and goats in the Persepolis royal economy. Acta Sumerologica 2, pp. 37-51. KocH, H. 1993: Achiimeniden-Studien. Wiesbaden. KuHRT, A. 2007: The Persian Empire: A Corpus of Sources from the Achaemenid Period. London. 2008: The Persepolis archives: Concluding observations. In: P. Briant/WF.M. Henkelman/M.W Stolper (eds.), L'archive

293

des fortifications de Persepolis: Etat des questions et perspectives de recherches: Actes du colloque organise au College de France par la "Chaire d'histoire et civilisations du monde achemenide et de !'empire d'Alexandre" et le "Reseau international d'etudes et de recherchs achemenides" (GDR 2538 CNRS), 3-4 novembre 2006 (Persika 12) . Paris, pp. 563-570. LEVAVI, Y. 2014: ginnu-silver from the time of Nebuchadnezzar {and Nabu-a\}ge-iddin's term as temple administrator). NABU 2014/102. MAGEE, P. et al. 2005: The Achaemenid Empire in south Asia and recent excavations in Akra in northwest Pakistan. AJA 109, pp. 711-7 41. MooREY, P.R.S. 1994: Ancient Mesopotamian Materials and Industries: The Archaeological Evidence. Oxford. NrMCHUK, C.L. 2010: Empire encapsulated: The Persepolis Apadana foundation deposits. In: J. Curtis IS.]. Simpson (eds.), The World of Achaemenid Persia: History, Art and Society in Iran and the Ancient Near East. London, pp. 221-229. PoWELL, M.A. 1987-1990: MaBe und Gewichte. RLA 7, pp. 457-517. PoTTS, D.T. 2008: The Persepolis Fortification texts and the royal road: Another look at the Fahliyan area. In: P. Briant/WF.M. Henkelman/M.W. Stolper (eds.), L'archive des fortifications de Persepolis: Etat des questions et perspectives de recherches: Actes du colloque organise au College de France par la "Chaire d'histoire et civilisations du monde achemenide et de !'empire d'Alexandre" et le "Reseau international d'etudes et de recherchs achemenides" (GDR 2538 CNRS), 3-4 novembre 2006 {Persika 12). Paris, pp. 275-301. RENGER, J.M. 2003: Iraq. Ancient period. In: J. Mokyr (ed.), The Oxford Encyclopedia of Economic History. Oxford, pp. 147-151. 2011: The role and the place of money and credit in the economy of ancient Mesopotamia. In: H. Ganssmann (ed.), New Approaches to Monetary Economics and

M. TAMERUS

294

lheory: Interdisciplinary Perspectives. New York, pp. 15-36.

-

RooT, M.C. 2008: The legible image: How did seals and sealing matter in Persepolis? In: P. Briant I F.~. Henkelman I M.W Stolper (~ds.), Larchzve des fortifications de Persepolis: Etat des questions et perspectives de recherches: Actes du colloque organise au College de France par la "Chaire d'histoire et civilisations d~ monde achemenide et de /'empire d'Aiexandre" et le "Reseau international d'etudes et de recherchs achemenides" (GDR 2538 CNRS), 3-4 novembre 2006 (Persika 12). Paris, PP· 87-148. SANCISI-WEERDENBURG , H ·WA V · ·M . 1980·. 1auna en Persai. Grieken en Perzen in een Ander Perspectief (Diss. Rijksuniversiteit Leiden). Groningen.

-

:V.

1998: Baji. In: M. BrosiusiA. Kuhrt (eds.) Studies ~n Persian History: Essays in Memory of Davtd M Lewis (Achaemenid History 11). Leiden, pp. 23-34. ScHEIL, V. 1911: Memoires de delegation en Perse. IX: Textes elamites-anzanites. Troisieme serie. Paris. ScHMIDT, E.F. 1957: Persepolis. Il· Contents ofthe Treasury and Other Discoveries (OIP 69). Chicago. ScHMITT, R. 1967: Medisches und persisches Sprachgut bei Herodot. ZDMG 117, PP· 119-145 . 2009: Die altpersischen Inschriften Achaimeniden. Wiesbaden.

der

2014: Worterbuch der altpersischen Konigsinschriften. Wiesbaden. SPEK, R.J. VAN DER 2011 : The 'silverization' of the economy of the Achaemenid and Seleukid empires and early modern China. In: z.H. ArchibaldiJ.K. Davies/Y. Gabrielsen (eds.), lhe Economies of Hellenistic Societies, lhird to First Centuries BC. Oxford, PP· 402-420. STOLPER, M.W 1984: Texts from Tall-i Malyan. I: Elamite Administrative Texts (1972-1974). Philadelphia. -

1992: The Mura.Su texts from Susa. RA 86, PP· 69-77.

-

2000: Ganzabara. Encyclopaedia Iranica 10, pp. 286-289. 2004a: Elamite. In: R. Woodard (ed.), lhe Cambridge Encyclopedia of the World's Ancient Languages. Cambridge, PP· 60-94. 2004b: Hidali. Encyclopaedia Iranica 12, pp. 308-309.

TAVERNIER, J. 2007a: Iranica in the Achaemenid Period (ea. 550-330 B. C.): A Lexicon of Old Iranian Proper Names and Loanwords, Attested in Non-Iranian Texts (Orientalia Lovaniensia Analecta 158). Leuven. -

2007b: On some Elamite signs and sounds. ZDMG 157, pp. 265-291.

TuPLIN, C. 2008: Taxation and death: Certainties in the Persepolis Fortification archive? In: P. Brianti~F.~. HenkelmaniM.W Stolper (~ds.), Larchwe des fortifications de Persepolis: Etat des questions et perspectives de recherches: Actes du colloque organise au College de France par la "Chaire d'histoire et civilisations d~ monde achemenide et de /'empire d'Aiexandre" et le "Reseau international detudes et de recherchs achemenides" (GDR 2538 CNRS), 3-4 novembre 2006 (Persika 12). Paris, pp. 317-386. -

SILVER IN SEARCH OF HIS FATHER: A COMPARATIVE FOLKLORISTIC APPROACH TO AN EPISODE OF THE SONG OF SILVER

2014: The changing pattern of Achaemenid Persian royal coinage. In: P. Bernholz I R. V~ubel. (eds.), Explaining Monetary and Fmanczal Innovation. A Historical Analysis. New York etc., pp. 127-168.

VARGYAS, P. 1999: Kaspu ginnu and the monetary reform ofDarius I. ZA 8912, PP· 247_268. WAERZEGGERS, C. 2010: Babylonian in Susa. The travels of Babylonian businessmen to Susa reconsidered. In: B. Jacobs 1 R. Rollinger (eds.), Der Achlimenidenhof Akten des 2. Internationalen Kolloquiums zum lhema "Vorderasien im Spannungsfeld klassischer und altorientalischer Oberlieferungen'; Landgut Castelen bei Easel, 23.-25. Mai 2007. Wiesbaden, pp. 777-813. YAGHMAEE, E. 2010: Excavations in Dashtestan (Borazjan, Iran) (Abstract). In: J. CurtisiS.J. Simpson (eds.), lhe World of Achaemenid Persia: History, Art and Society in Iran and the Ancient Near East. London, p. 317.

Willemijn ].I. Waal

INTRODUCTION

Excavations in the Hittite capital ljattusa, often written sumerographically as URUKLJ.BABBAR, the 'Silver City', have yielded disappointingly few economic records and information about the economic role and value of silver in the Hittite Empire is limited. In this paper, I would therefore like to address a more ethereal aspect of this precious metal, focusing on the enigmatic character of the boy Silver, the main protagonist of the 'Song of Silver'. Considering Ben's broad interests and expertise, I hope he can also appreciate a mythological story about personified silver.

THE KuMARBI OR KINGSHIP OF HEAVEN CYcLE

The Song of Silver is commonly thought to belong to the so-called 'Kumarbi' or 'Kingship of Heaven' cycle. Central theme of this cycle is considered to be the battle between Tdsub and Kumarbi for kingship of heaven. It is generally assumed that the Song of Kumarbi (or the 'Song ofEmergence' 1) is the first song of the cycle, followed by the Song ofLAMMA, the Song 2 of Silver, the Song of ljedammu and the Song of Ullikummi. Over the years, proposals to include other compositions, such as Ea and the Beast, Wasitta, Eltera and the Song of the Sea have been made. 3 The idea that all these stories belong to one coherent cycle in which the stories are directly related to each other and can be placed in a consecutive order is, however, far from self-evident and has been questioned. 4 In recent publications, this matter has been (re)-addressed. Alfonso Archi has shown that the cycle consists of stories composed at different times and by

Thus Coni 2007. Van Dongen 2012 prefers the translation 'The Song of Going Forth'. See Houwink ten Cate 1992, 109-20 and also Siegelova 1971, 82-84, Hoffner 1998, 40-42, Rutherford 2 2001, 604-605, Schwemer 2001, 45 +375 1, Van Dongen 2012, 30-31 and Archi 2009, 211. For a recent evaluation of the compositions belonging to the cycle, see Van Dongen 2012, 29 with refer3

1

ences. See already West 1997, 104 who suggests that each story may have been an independent and self-sufficient 4 entity.

SILVER IN SEARCH OF HIS FATHER

WJ.I. WAAL

different bards and suggests that an 'archipelago' may be a more suitable image than a cycle.s ~na Maria Polvani has argued that, rather than forcing all these compositions into one cycle, 1t would be better to distinguish different cycles. 6 Despite these objections, the stories are still generally co~sidered to belong to the Kumarbi-cycle. Most recently Erik van Dongen, though ac~o"':'ledgmg the a~ove arguments, considers it more useful to retain the cycle as previously mamtamed, at least nll further evidence comes to light.? One could, however, also argue that until more evidence comes to light, it is best to leave all options open and to appreciate the members of the supposed cycle as individual stories. As ~olvani has rightly observed, the inclusion of some of the songs in the Kumarbi-cycle has sometimes led to strained interpretations without any consideration for the specific characteristics of each of the stories. This certainly holds true for the Song of Silver.

above stories ofUllikummi and Ijedammu, Silver is seen as an instrument ofKumarbi, created to dethrone Tdsub and to take over kingship of heaven. This interpretation is, however, for a large part based on and restorations and the story is probably to be understood quite differently. 8 In this paper, I would like to re-evaluate an essential scene of the Song of Silver from a comparative folkloristic point of view, which sheds a new light on the character of Silver.

The boy Silver twitted with illegitimacy sets out to search for his father In a reasonably well-preserved passage at the beginning of the Song of Silver, the adolescent Silver is confronted with the fact that he does not have a father. This scene runs as follows: Fragment I9 [ ] ... [Silver [hit] the orphan boy with a st[ick ]. The orphan boy spoke an evil word t[o] Silver: "My Silver, why [do you hit us?] Why do you strike us? You [are] an orphan boy like us". [When Silver heard these words] he began to weep. Weeping, Silver went into his house. Silver began to speak w[ords (.. )] to his mother: "The boys, whom I struck down in front of the gate, they are insulti[ng] 10 me. I hit a boy with a stick and he spoke an [evil] word bac[k] to me. The words which the orphan boy spo[ke] to [m]e, hear [the] m, o my mother! 'Why [do] you [hi]t [us? Why do you strike] u[s? You are an orphan like] u[s]'.

THE SoNG OF SILVER

In its first edition by Harry Hoffner (Hoffner 1988), the Song of Silver was identified as a of th_e Kumarbi Cycle, which showed strong affiliations and similarities with the Song of Ulhkumm1 and the Song of Ijedammu. The restorations proposed in Hoffner's edition of the Song of Silver, which are virtually entirely followed by Haas 2006 and the online edition of Rieken et al. 2009-, are very much based on the assumption that the Song of Silver basically has the same outline as the two aforementioned songs. mem~er

In the Song ofUllikummi- the better preserved one of the two- the god Kumarbi raises a supplanter in order to dethrone the storm god Te.Ssub. He does this by pairing with a giant rock, which gives birth to a stone monster. Kumarbi hides the stone child, who is deaf and blind, from the other gods by placing it on the right shoulder of the Atlas-like figure Ubelluri. The s~one creature grows very fast and when it reaches the sky, he is discovered by the sun god. Hornfied, th~ sun god warns the storm god Te.Ssub, who despairs when he sees the stone giant. The goddess Sauska tries to seduce the monster, but being deaf and blind he is immune to her charms: After a failed attempt to overcome Ullikummi in a fight, Te.Ssub goes to the god Ea for advtce. Ea eventually saves the day by sawing off Ullikummi's feet from Ubelluri. The less well-preserved song of Ijedammu seems to contain similar elements: here, Kumarbi has created a monster with the daughter of the sea, Ijedammu, who is enticed by Sauska.

297

Fragment II 11 [The mother?] too[k] the stick [aw]ay.? 12 [ ](s)he turn[ed] around? [began] to answer: [ "My Silver13, do not hi[t]/kil[l] me. [ the ci]ty, I will tell you about it. [ ] of [the c]ity ofUrkd, he (is) its father [ ] he dwells in Urkd[ ] ... for/of all lands the court cas[es ]. Your brother (is) the storm god. He is [kin]g in heave[n] and he is k[ing] in the land. [Your 14 sister is !Star, she is quee[n] ofNiniveh. [ not anybo]dy/somebo]dy you fear/respect. One 15 go[d? ] he ... (e)s the land and the enemies [and?] the wild anim[als? ] he ... (e)s the top

A new edition of the Song of Silver is currently being prepared by the author. KUB 17.4, 1'-15'; KBo 22.82, 1'-11'; KBo 26.146, 1'-5', see Hoffner 1988, 149-151 (fragment 3) and

8

9

Rieken et al. 2009- , hethiter.ner/: CTH 364.4. 1 For the translation of rhe hapax form te-ri-if-k[dn-zi], see Hoffner 1988, 151 and Kloekhorst 2008, 874

°

s. v. terifkela-.

KUB 36.18 obv. ii 1'-22', KBo 22.80 r.col. 1'-10', KUB 33.115 obv. ii 1'-12', see Hoffner 1988, 152-155 (fragment 2 ii) and Rieken et al. 2009-, hethiter.net/: CTH 364.2 12 Hoffner 1988, Haas 2006 and Rieken et al. 2009- assume that it is the mother who takes away the stick,

11

In the Song of Silver, Kumarbi has sired a child with a mortal woman. Kumarbi apparently leaves the mother, since the boy, named Silver, grows up without his father. When he finds ~ut t~at his fat~er is the god Kumarbi, Silver sets out to search for him. The rest of the story ts. qmt~ uncertal~vand badly da~aged, but it is generally assumed that Silver takes away the kingship from Tessub and terronzes the gods, ruling over them like a tyrant. In analogy to the 5

Archi 2009, 211. Note that H aas 2006, 130 also has reservations about the idea of a closed cycle.

6

Polvani 2008, especially 620-624.

7

Van Dongen 2012, 29.

which indeed seems most plausible. !3 The form is a dative-locative, but a vocative makes more sense here (also Rieken et al. 2009-). Alterna82 tively, one could interpret the form as a dat.-loc. in final position, see Hoffner 1988, 154• . l4 KUB 36.18 and KBo 22.80 read na-ab-ti 'you fear' , KUB 33.115 obv. ii 4' reads na-a-bi which can be understood as a 3'd pres. sg. or 2nd imp. sg. Hoffner 1988, 153 completes a negation here ('you must [not] fear [an]y [deity]', see also Haas 2006, 149 and Rieken et al. 2009-, which is probable, bur not certain. 15

KUB 36.18 obv. ii 16' reads: the land of the enemies [and'] the wild[ animals.

WJ.I. WAAL do[wn? ] the bottom u[p? 16 [(When) he] heard the [w]ords of his mother, he went fTrom] the city [ ] he reached the city Urkd. He [ J he did nor find in his house. He we[nt]? to roam the lands [ ] he [w]anders. U[p?] in the mountains ...

The boy Silver is called a wannumiya'S DUMU, a term which is discussed by Hoffner, who convincingly argues that this term refers to children or women without a protective male figure like a father or husband, i.e. (half) orphans or widows. 17 1he above scene can be summarized thus: The boy Silver is hitting orphan boys. When he is confronted with the fact that he is an orphan just like them, he starts to cry and goes back home. After a short broken part his mother, apparently after having been threatened by Silver, reveals the identity of his father. The name of the father is not preserved, but considering his association with the city ofUrkes, it seems very likely that this was Kumarbi. Silver then sets out to go and look for his father.

As already mentioned above, Silver is generally regarded as one of Kumarbi's monstrous creations against Tdsub and his character is considered negative. The above scene is seen as indicative for his later violent behavior. In the words ofVolkert Haas (2006, 149): "Der heranwachsende Knabe Silber steigert sich vom Raufbold zu einer Gefahr fur die Weltordnung: Zuerst verprligelt er seine Altersgenossen, dann bedroht er seine eigene Mutter; spater greift er sogar die Cotter an und bedroht die Gestirne Sonne und Mund." It has not been recognized, however, that this scene in fact represents a common folk tale moti£

A COMPARATIVE FOLKLORISTIC APPROACH

The field of comparative folkloristics is not very well known in the field of cuneiform studies. Notable exceptions are recent studies such as those of Amar Annus (2009) , Michael Haul (2000), Wouter Henkelman (2006, 2010a & b) and Herman Vanstiphout (2001), which show how fruitful this approach can be when studying ancient Near Eastern literature. 18 The motif of a hero who grows up without a father or both his parents is very common and widespread. Many variants exist, but typically, the child is of high birth (his real father is a king or a god) but he is raised by humble people, or sometimes animals. Interestingly, the above scene in which Silver is confronted with the fact that he does not have a father, which gives the impetus to go and look for him, can be identified as a very specific folktale motif, namely H 1381.2.2.1.1: Boy twitted with illegitimacy seeks unknown fother. 19 In his motif-index, Thompson lists various examples of folk tales in which this motif can be found, to which several more may be added. Below, an overview is given of the most important ones:

16 17

KUB 33.115 obv. ii 8' adds an extra line here, which seems to be a repetition of obv. ii 4'. See Hoffner 1988, 151.

SILVER IN SEARCH OF HIS FATHER

299

Other stories with folktale motifH1381.2.2.1.1

The tale of Truth being blinded by Falsehood - This tale from ancient Egypt was written down in the 19th dynasty, and thus more or less contemporarily with the Song of Silver. 20 The story is about Truth, who is wrongfully accused of theft by his brother False and therefore blinded. He is found by a lady who makes him her doorkeeper. She is attracted to his beauty and give birth to his child. The boy, who is very tall and resembles the child of a god, does very well in school: He practiced all the arts of war, and he surpassed his older companions who were at school with him. Then his companions said to him: "Whose son are you? You don't have a father! " And they reviled him and mocked him: "Hey, you don't have a father! " Then the youth said to his mother: "What is the name of my father? I want to tell it to my companions, for they quarrel with me. 'Where is your father?' So they say; and they mock me." 21

His mother then tells him that his father is the blind doorkeeper. The boy is very angry with his mother, brings his father into the house and treats him respectfully. When he hears his life story, he sets out on a journey to avenge his father.

The story ofSohrab and Rustem - In the story of Sohrab and Rustem from the Shahmaneh (the Persian Book of Kings) by Abolqasem Ferdowsi, Rustem conceives a child with a princess on one of his travels, leaving right afterwards. When he departs, he gives the princess an onyx to place on their son's arm, if a son should be born, which indeed happens. The boy is much stronger than his companions: By the time he [Sohrab] was ten, no one dared to compete with him and he said to his mother: "Tell me truly now, why is it I'm so much taller than other boys of my age? Whose child am I, and what should I answer when people ask about my father? If you keep all this hidden from me, I will not let you live a moment longer." 22

His mother reveals the identity of his father, and Sohrab goes looking for Rustem, swearing he will place his father on the throne of Iran. Eventually, the father and son meet in battle and Rustem, who refuses to reveal his identity, inflicts a fatal wound on Sohrab. Only when he hears the laments of the dying Sohrab about his failure to find his father and he discovers the onyx under his armor, Rustem realizes that he has unwittingly killed his own son.

The legend of Oedipus - The motif also occurs in the legend of Oedipus as found in The Library of (Pseudo-)Apollodorus. His father, king Laius, exposes Oedipus at birth because of an ominous prophecy. The child is found and raised by king Polybus of Corinth and his wife. He grows up to be stronger than all his comrades, who start calling him a bastard out of envy:

18

W ith respect ro earlier publications, the case-study of the Sargon-Legend by Lewis 1980 is worth mentioning. See on this topic also Unal 1986.

The story is preserved in one copy, P. Chester Beatty 11, P. Brit. Mus. 10682. For a translation, see Lichtheim 2006, 211-214.

19

21

See Lichrheim 2006, 212.

22

For translation, see Davis 2006, 190.

According ro the classification ofThompson 1955-1958; for the present motif see 498-499; also Uther 2004 type tale *369, 231.

20

i'

SILVER IN SEARCH OF HIS FATHER

W.J.I. WAAL

300

When the boy grew up and excelled his fellows in strength, they spitefully twitted him with being supposititious. He inquired of Periboea, but could learn nothing; so he went to Delphi and inquired about his true parents. 2 3

In contrast to the other stories, Oedipus grows up without both his parents and his adoptive mother does not tell him anything about his origins, and he therefore decides to visit the oracle of Delphi looking for answers.

The Story ofAli and Zahir from Damascus - The tales from Arabian Nights include a story about a rich but childless man named Zahir from Damascus, who marries princess Farha but is captured by genies right after their marriage. In his absence, Farha gives birth to a son, Ali. When he is grown up, he one day beats a slave of his mother, who then insults him, saying that though he himself may be only a slave, he at least knows who his parents are, as opposed to Ali, who is an illegitimate child: 27 Eines Tag schlug ich einen ihrer Sklaven, der mir von meiner Kindheit an verhaBt war, und sagte ihm: "Verruchter Schwarzer, wie oft habe ich dir schon gesagt, ich wollte dich nicht mehr vor meinen Augen haben, warum weichst du mir nicht aus, wenn du mich kommen siehst?". Da antwortete er mir: "Ich bin allerdings nur ein Sklave, doch jeder kennt meine Eltern und weiB daB Sie auch Sklaven waren, wie ich; aber du vaterloser Bube, weiBt du, daB dein Vater ein unbekannter, auf dem Meer aufgefundener, hergelaufener Mensch war! Gott verdamme dich, weil du mich so miBhandelst, und lasse dich gleich mir unter fremden Leuten umher-

The tale ofla Hija del Carbonero- In this Spanish tale, the daughter of a coalman marries the king and bears him two children. Because of deceitful evil talk of her jealous sisters, the king orders the boys to be killed. In order to save them from this fate, their mother puts them in a crystal urn on the river where they are found and raised by a village woman and her husband. One day they argue at school with a fellow-pupil and they get into a fight. Then the boy who quarreled with them tells them that they do not have a mother, but that she [the adoptive mother] found them in the river. When the boys hear this, they go home crying and ask their adoptive mother if she is indeed not their real mother. She confirms this, whereupon the children decide to go out for adventures. Eventually they get reunited with their parents (and the evil sisters are punished). 24 Similarly to the story of Oedipus discussed above, the- in this case two -heroes who grow up without both their parents.

wandern, die kein Mitleid mit dir haben!"

After speaking these words, the slave runs away and Ali goes after him with his sword, but he cannot catch him. Ali contemplates the slave's words and confronts his mother, demanding explanations and threatens to kill her if she does not tell him: "Bei dem erhabenen, barmherzigen Gott, ich muB wissen, wer mein Vater war, oder mein Schwert soli deinem lasterhaften Leben ein Ende machen!" Meine Mutter erwiderte weinend: "Mein Sohn, bei dem, der die Berge geschaffen und ihre GroBe und Schwere kennt, dein Vater war einer der Besten und Edelsten seines Volkes!" - "Nicht so vide Worte!" rief ich ganz auBer mir: "nenne mir meinen Vater, oder ich bringe dich und mich urn, denn ich habe heure etwas gehort, daB meine jugendliche Locken grau farbt." Als meine Mutter dies horte, sagte sie mir: "Stecke dein Schwert ein und setze dich ruhig zu mir, ich will dir alles

The legend ofLaka - In the Hawaiian legend of Laka, which is set on the island of Maui, the father ofLaka is killed on a trip for a birth gift for his son before the child is born. The boy is raised by his grandmother and when other boys tease Laka because he is fatherless, he decides to seek his father's bones. 25 The story ofprince ]ivaka as the king ofphysicians - In the intricate Tibetan tale about prince Jivaka, the woman Amrapali, who works as a courtesan, gives birth to the child of king Bimbisari, who has left her before the child is born. The child grows up without his father: Now when nine months had gone by, Amrapali gave birth to a fine, good-looking boy. While he was a growing lad, he was playing one day with the sons of the Licchavis ofVaisali and they addressed him with scoffing words, saying: "Son of a female slave, born from among many hundreds of thousands, who is your father?" He went in tears to his mother. She asked him why he wept, and he told her the whole story. 26

gestehen." 28

His mother tells him the whole story and Ali decides to go and look for his father, against his mother' wishes.

The story of Muladeva and the Brahman's Daughter - In this story, which forms part of the Sanskrit poem Kathiisaritsdgara ('Ocean of the Streams of Story')) composed by the poet Somadeva, the Brahman's daughter, who is left by her husband Muladeva, gives birth to a son. And when she was in Pa!aliputra she gave birth in due time to a son. And that boy grew up and learned all the accomplishments. And when he was twelve years old, that boy, in a childish freak, happened to strike with a creeper a fisherman's son of the same age. When the fisherman's son was beaten, he flew in a passion, and said: "You beat me, though nobody knows who your father is; for your mother roamed about in foreign lands, and you were born to her by

After telling her son that his father is king Bimbisari she sends him off to go and meet his father.

23

301

Appollodorus, The Library 3.5.7. Translation: Frazer 1921. See also Constans 1974, 25-26.

24

See Espinaso 187, 314. See Beckwith 1970, 263. The Laka Legend is widespread in the South Pacific. The motif of a fatherless boy taunted by other boys is also found in the story from Tuamotus about the boy Rata, see Beckwith 1970, 265. Similarly, in a story from Marquesas the orphan Ata is teased by other children about his dead parents, see Beckwith 1970, 269. 25

26

For translation, see Ralston 1906, 89-90.

See Weil 1865 (1978), volume 4, 211 and Chauvin 1901, 72. Note that this story is contained only in the translation ofWeil and relies on a manuscript form Gotha, see Marzolph/van Leeuwen 2004, 102.

27

28

Translation Weil 1865 [1978), 211.

302

SILVER IN SEARCH OF HIS FATHER

WJ .I. WAAL

303

so:ne hu~band or other. When this was said to the boy, he was put to shame. So he went and said to his mother: Mother: who and where is my father? Tell me!"29

rogates his (adoptive) mother about his origins (b). When the mother reveals his true identity -sometimes after being threatened (c), the hero decides to go and search for his father (d).

r:is mo.ther tells him her life story and the boy goes out to search for his father and in the en d b nngs hrm back as a captive.

Despite the fact that these stories share the same motif, it should be clear that that we are dealing with independent stories, and that the circumstances and the way in which the motif is used within each of the stories differ to a considerable extent. The variants eloquently demonstrate the intricate mechanics of the spread of folktale motifs, which are continuously adjusted, reshaped and attuned in order to fit the contexts of a particular story, yet maintaining their structure. The legend of Oedipus, for example, is the only one in which the mother does not tell the hero about his origins, for this would interfere with the rest of the story, wherein

~e overview given here is far from exhaustive and more tales could be added, such as the

Armeman s~ory of the son of king Chosrou, who grows up without his father and becomes a danger to hts playm~tes and then sets out to seek his father. 3° Further, Hassan el-Shamy lists folktales from Palestma and Kuwait containing motifH13812.2.1.1.31

Oedipus has to go and consult the oracle of Delphi. . In addition, elements of the above-discussed motif may be found separately in other stones. The element of an ille.gitimate hero being teased at school out of envy, for example, is also _ran of the story of the Bzrth ofAbu Zayd in the Egyptian Sirdt Bani Hila! romance. In the vers~on collected by Abnoudy, the black boy Abii Zayd, who does not know that he is adopted, t~vo~~s the jealousy of his schoolmates because he excels in everything and is the teacher's favonte. . They then start to tease him with the fact that he is a bastard whereupon Abii Zayd reacts funously · 1u d e r h" fand h beats the slanderer until he bleeds ·33 This story d oes, h owever, not rnc a search ror ts at er. Fi~ally, _one ~ay compare the element of a young (adopted) hero who stands out when c~mpetmg. wtth hrs comrades, thus displaying his noble birth, to the folktale motif of the boyking (Motw des Knabenkonig). A well-known example hereof is the Legend of Cy ld b H d · h" h h rus as to e~o otus, m w tc t e young Cyrus, supposedly the son of a herdsman, shows traits of kingshtp when he has been chosen to play the role of king in a play with other children. 34

!

Let us now return to the Song of Silver. As mentioned above, attention has been drawn to the violent behavior of Silver: he is first found hitting an orphan boy, who then points out to him that he does not have a father. Similar scenes are found in the tale Miiladeva and that of Ali, in which the protagonists are hitting a fisherman's son and a slave respectively. The arguments of their victims are the same: the protagonist has no right or reason to hit them, since he does not (even) know who his father is. In the Spanish fairy tale the two heroes are fighting with one of their classmates who then insults them by saying that they are illegitimate children. In the other stories, the use of violence is not explicitly mentioned, but the protagonists are usually mocked and harassed by their schoolmates, presumably out of envy because they excel in physical strength and surpass their companions. After being confronted with the fact that he does not have a father, the boy Silver again shows violent behavior, this time towards their mother, threatening to kill her if she does not tell the truth. The intimidation of the mother is also present in the tale of Ali and that of Sohrab. The use of violence cari thus be seen as typical for certain variants of motif H 13812.2.1.1.

SILVER REHABILITATED

e ~ the above stori.es demonstrate, motif H 13812.2.1.1 may consist of the following elbments. the hero, who ts a (half) orphan, is as a boy confronted with his illegitimacy usually Y other boys of his age (a). The hero thereupon goes home- occasionally in tears- a~d inter-

; ;rranslation of C.H Tawne~ as edited by Penzer 1928 [reprint 1968], vol. 9, 82-83. Penzer 1928[1968], 2 Bloo:~;;;1;~~t, t1h9e5~ccusauon of bastardy is a well-developed motif in Sanskrit literature, referring to 30

See Golzio 1996, 25 s.v. Knabenkonig.

31

See Shamy (el) 1995, 1972.

N~te that ~he motif of jealousy invoked by classmates because the teacher favours the hero is also found later m the Tibetan story of prince Jivaka, see Ralston 1906 95-98.

32

33

.

Rather than being a mere instrument in the evil plans of Kumarbi, the boy Silver is likely to be the main protagonist and hero of the story, just like the fatherless boys in the stories discussed above. There are more indications that support this assumption: the proemium of the 35 Song of Silver- as far as it is preserved- seems to describe Silver in positive terms. Further, in

,

See Sch1ppers 2003, 358. I am indebted to Johan WeststeiJ·n for this referen ~ See Herodorus 1.114-115. The story of Candagutta is found in the Mahrkrhe Mahavamsa chronicle from Ceylon , see Golzio 1996 , 24 s.v. 1\.na r? ben komg. .. a.vam~a-. I a, a commentary on .

M

In sum, the scene in which the boy Silver is hitting orphans boys, as well as his aggressive behavior towards his mother reflects a popular and widespread folkrale motif, and is not a demonstration of his violent nature, as hitherto thought. It may incidentally be noted that this particular scene also shows Silver's sensitive side, since he leaves the playground crying to go to his mother - another typical folk tale motif, which is also present in the Spanish and the Tibetan tales. Emotional and aggressive heroes are of course not uncommon in ancient Near Eastern literature, nor in later literatures, one only needs to think of the at the same time rash and brutal behavior of heroes such as Gilgame5, Achilles or the protagonists in the Shahmaneh or Arabian Nights.

See HFAC 12, 1'-13' (CTH 364.1). In addition we may mention here the possible, but uncertain attes62 tation of Silver as king in KUB 27.38 rev. vi 19', see de Marrino 1993, 129• , Houwink ten Cate 1992,

35

SILVER IN SEARCH OF HIS FATHER

WJ.I. WAAL

line 7 it is explicitly stated that the song is about Silver ('I will [s]ing of him, Silver, the fi[ne? .. .') and Kumarbi does not appear to be mentioned at all. By contrast, in line 4 of the Song of Ullikummi Kumarbi is presented as the central character ('It is Kumarbi, Father of All Gods, of whom I sing'). 36 Last but not least, the fact that Silver goes out looking for Kumarbi and not the other way around, pleads against the idea that Silver merely served Kumarbi's ends. In fact, the role of Kumarbi in the story is likely to have been much more modest than is generally assumed; a number of times the apparent involvement ofKumarbi is solely based on restorations. Too little is preserved of the Song of Silver to entirely exclude the possibility that the rest of the story followed the same outline as Ullikummi and Ijedammu, but in light of the new interpretation presented here this seems unlikely. The above observations confirm the afore-mentioned doubts and objections with respect to the existence of a closed and coherent Kumarbi-cycle. The stories attributed to the Kumarbi cycle may share a general theme, the struggle between the gods for kingship of heaven, but this was a very well-known and universal theme. Within Hittite Anatolia, for example, this theme is also found in the stories of Iluyanka, which are generally considered to be of Hattic origin and to be unrelated to the Kumarbi songs. 37 Here as well, the storm god is stripped off of his powers and has to regain kingship with the aid of other gods. This is not to say of course, that the stories ofljedammu and Ullikummi would have been unknown to the audience listening to the Song of Silver and vice versa. In addition, more than one version of these stories may have existed, undoubtedly together with many more stories about divine battles, which were transmitted orally and as a consequence have not come down to us. The written records only reveal glimpses of a much larger oral tradition, which makes it all the more hazardous to force these little shreds of evidence together to form a single logical narrative. Though similarities between some of the 'Kumarbi Songs' should of course not be ignored, there is at the moment not enough evidence to support the idea of a coherent uniform cycle and these compositions should first and foremost be studied as stories their own right. 38

Ill, Salvini 1977, 88-90 and Wilhelm 2003, 395. The passage is generally translated as 'Silver, Ruler and Heroic King' vel sim. If this interpretation is correct and this line indeed refers to our mythical Silver, the reign of Silver would be judged positively here. Interestingly, in the next line Ijedammu appears to be mentioned, of whom it is said that he was raised or installed as king by Kumarbi. The same is, however, not said of Silver, which could tentatively be seen as an indication that he became king without Kumarbi's help. If this passage indeed refers to the mythical kings l)edammu and Silver, then it shows that the traditions about these two gods were well known and could be linked together. The passage is, however, too uncertain to allow any conclusions. 36 CTH 345.1, for the most up-to-date text edirion see Rieken et al. 2009-, hethiter.net/: CTH 345.!.1. 37 CTH 321, for an edition of these tales, see Beckman 1982 and Rieken et al 2009-, hethiter.net/: CTH 321. 38 See also West 1997, 104, Polvani 2008, 620-624 and Archi 2009, 211. An additional argument that the songs presently ascribed to the Kumarbi cycle are not necessarily related to each other is the different role the god Ea plays in each of the stories. Rather than a development in character (Hoffner 1998, 41-42), this is likely to reflect different traditions.

BIBLIOGRAPHY AARNE, A./THOMPSON, ST. 1961: 7he Tjpes ofthe Folktale. A Classification and Bibliography. Helsinki (second revised edition). ANNUS A. 2009: The folk-tales of Iraq and the literary traditions of Ancient Mesopotamia. ]ANER9.1, pp. 87-99. ARCH!, A. 2009: Orality, direct speech and the Kumarbi cycle. AoF 36, pp. 209-229. BECKMAN, G. 1982: The Anatolian myth of Illuyanka.]ANES 14, 11-25. BEcKWITH, M.T. 1970: Hawaiian Mythology. Honolulu. BLOOMFIELD, M. 1919: 7he Life and Stories of the ]aina Savior Pdrfvandtha. Baltimore. CHAUVIN, V 1901: Bibliographie des ouvrages arabes ou relatifi aux arabes publie dans !'Europe chretienne de 1810 a 1865. Tome V Les Milles et une nuits (deuxieme partie). Liege/ Leipzig. CoNSTANS, L. 1974: La Legende d'CEdipe. Geneve. CoRTI, C. 2007: The so-called "Theogony'' or "Kingship in Heaven". The name of the song. SMEA 49, pp. 109-121. Dongen, E. van 2012: The Hittite Song of Going Forth (CTH 344). A reconsideration of the narrative. WdO 42/1 pp. 23-84. DAVIS, D. 2006: Shahnameh. 7he Persian Book of Kings by Abdolqasem Ferdowsi. New York. EsPINOSA, A.M. 1987: Cuentos Populares de Castilla y Leon, Tomo I. Madrid. GoLZIO, K.H. 1996: 'Knabenkonig'. In: K. Ranke et al. (ed.), Enzyklopiidie des Miirchens. Handworterbuch zur historischen und vergleichenden Erziihlforschung, Band 8. Berlin/ New York, pp. 23-25. HAAs, V 2006: Die hethitische Literatur. Texte, Stilistik, Motive. Berlin/New York. HAUL, M. 2000: Das Etana-Epos. Ein Mythos von der Himmelfohrt des Konigs von Kif (Gottinger Arbeitshefte zur alrorientalischen Literatur 1). Gottingen. HENKELMAN, WF.M. 2006: The birth of Gilgame5 (Ael. NA XII.21). A case-study

in literary recepnv1ty. In: R. Rollinger and B. Truschnegg (eds.), Altertum und Mittelmeerraum: Die antike Welt diesseits und jenseits der Levante. Festschrift for Peter W Haider zum 60. Geburtstag (Oriens et Occidens 12). Stuttgart, pp. 807-856. HENKELMAN, WF.M. 2010a: Beware of dim cooks and cunning snakes: Gilgame5, Alexander, and the loss of immortality. In: R. Rollinger et al. (eds.), lnterkultura.litiit in der Alten Welt: Vorderasien, Hellas, Agypten und die vielfoltigen Ebenen des Kontakts (Philippika 34). Wiesbaden, pp. 323-59. HENKELMAN, WF.M. 2010b: Iter ad paradisum, Terug naar Gilgame5, en verder. Lampas 42.2, 2009, pp. 111-135. HoFFNER JR., H.A. 1988: The Song of Silver: A member of the Kumarbi cycle of "songs". In: E. Neu and C. Riister (eds.), Documentum Asiae Minoris antiquae. Festschrift for Heinrich Otten zum 75. Geburtstag. Wiesbaden,pp. 143-166. - 1998: Hittite Myths (SBL WAW 2). Atlanta. HoUWINK TEN CATE, P.H.J. (1992). The Hitrite storm God: His role and his rule according to Hittite cuneiform sources. In: D.J.W Meijer (ed.), Natural Phenomena: 7heir Meaning, Depiction and Description in the Ancient Near East, volume 152 of Verhandelingen der Koninklijke Nederlandse Akademie van Wetenschappen Afdeling Letterkunde. Amsterdam, pp. 83-148. KLoEKHORST, A. 2008: Etymological Dictionary of the Hittite Inherited Lexicon (Leiden Indo-European Etymological Dictionary Series 5). Leiden. LEWIS, B. 1980: 7he Sargon Legend. A study ofthe Akkadian Text and the Tale of the Hero who was exposed at Birth. Cambridge (MA). LICHTHEIM, M . 2006: Ancient Egyptian Literature. Volume 11: 7he New Kingdom. Berkeley. MARTINO, S. DE 1993: KUB XXVII 38: ein Beispiel kultureller und linguistischer Oberla-

306

WJ.I. WAAL gerung in einem Text aus dem Archiv von Bogazkoy. SMEA 31, pp. 121-134.

MARZoLPH, U. I LEEUWEN, R. VAN 2004: 1he Arabian Nights Encyclopedia (2 vols.). Santa Barbara/Denver/Oxford. PENZER, N.M. 1928 [1968]: 1he Ocean ofStory:

Being CH Tawney's Translation of Somadeva's Kathd sarit sdgara (or Ocean ofStreams of Story) (10 vols.). Delhi/Varanasi/Patna. PoLVANI, A.M. 2008: The God Eltara and the Theogony. SMEA 50, pp. 617-624. RALsroN, WR.S. 1906: Tibetan Tales Derived from Indian Sources. Translated from the Tibetan of the Kah-Gyur by E Anton von Schiefner. London. RANKE, K. eta!. (ed.) 1977: Enzyklopiidiedes Miirchens. Handworterbuch zur historischen und vergleichenden Erziihlforschung. Berlin/New York. RIEKEN, E. I LoRENZ, ]. I BAUER, A. I ESSER, F. 2009-: Mythen der Hethiter. http:!lwww. hethport.uni-wuerzburg.de/txhet_myrh/. RuTHERFORD, I_. 2001: The Song of the Sea (SA A. B. BA SIR): Thoughts on KUB 45.63. In: G. Wilhelm (ed.), Akten des IV lnternatio-

nalen Kongresses for Hethitologie: Wiirzburg, 4.-8. Oktober 1999 (StBoT 45). Wiesbaden, pp. 598-609. SALVINI, M. 1977: Sui testi mitologici in lingua hurrica. SMEA 18, pp. 73-91. ScHIPPERS, A. 2003: An Episode in the life of a hero in the Sirat Bani Hila!: Abu Zayd as a schoolboy. Oriente Moderno N.S. 22, pp. 347-359. SCHWEMER, D . 2001: Die Wettergottgestalten Me-

sopotamiens und Nordsyriens im Zeitalter der Keilschriftkulturen: Materia/en und Studien nach den schriftlichen Quellen. Wiesbaden.

of the Arab World. A Guide to Motif Classification. Volume I Bloomington/Indianapolis. SIEGELOVA, ]. 1971: Appu-Miirchen und Hedammu-~thus (StBoT 14). Wiesbaden~ THOMPSON, Sr. 1955-1958: Motif-index of Folk-Literature: a Classification of Narrative Elements in Folktales, Ballads, Myths, Fables, Medieval Romances, Exempla, Fabliaux, jestBooks, and Local Legends, Revised & enlarged edition (6 vols.). Bloomington. EL-SHAMY, H .M. 1995: Folk Traditions

DTHER, H.-J. 2004: 1he l}pes of International

Folktales: A Classification and Bibliography (3 vols., Folklore Fellows Communications 284-6). Helsinki. ONAL,. A. 1986: Das Motiv der Kinderaussetzung m den altanatolischen Literaturen. In: H. Kuhne et al. (eds.), Keilschriftliche Literatu-

ren. Ausgewiihlte Vortriige der XXXII Rencontre Assyriologique lnternationale. Berlin. VANSTIPHOUT, H.L.]. 2001: Shamshum aj-Jabbar: On the persistence of Mesopotamian literary motifs. In: WH. van Soldt et al. (eds.), Veenhof Anniversary Volume. Studies

Presented to Klaas R. Veenhofon the Occasion of His Sixty-Fifth Birthday (PIHANS 89). Leiden, pp. 515-527. WEIL, G. 1865 [1978]: Tausend und eine Nacht. Arabische Erziihlungen (4 vols.). Wiesbaden. WEsT, M. 1997: 1he East Face of Helicon: West

Asiatic elements in Greek Poetry and Myth. Oxford.

WrLHEL~, G. 2003: Konig Silber und Konig l:;ltdam. In: G .M. Beckman et aL (eds.), Hittite Studies in Honor of Harry A. Hoffner Jr. on the Occasion of His 65th Birthday. Winona Lake, pp. 393-395.

MYCENAE, RICH IN SILVER ]orrit M Kelde/

Whilst gold is the metal that is most commonly associated with Mycenae - mostly because of Homer's reference to that city, but also because of the fabulous death masks that were uncovered by Schliemann in the shaft graves of Grave circle A- relatively little is known about the role and appreciation of silver in the Mycenaean world. Silver artefacts have been found only sporadically at Mycenaean sites, suggesting that the metal was rare, yet there is good evidence that the Mycenaeans were already mining silver in the Laurion, south of Athens. This article will try to explain this apparent dichotomy, by exploring the role and importance of silver in the Mycenaean world.

INTRODUCTION

The Mycenaean world, and Mycenae in particular, is known primarily for the fabulous wealth of its elite. Homer called Mycenae 'Rich in Gold', and Schliemann's discovery of the Shaft Graves, which contained fabulous golden death masks and other precious objects, seemed to confirm the Poet's description of Agamemnon's citadel.' Apart from numerous gold objects, the Shaft Graves also contained a large amount of silver artefacts. Interestingly, the silver objects 'zeigen ausnahmslos gute Qualitat und saubere Arbeit? whereas the quality of the gold objects is more varied (Karo reports several poorly made cups with defects). Whilst over 30 silver objects were recovered from the Shaft Graves, silver is only sporadically found in later strata at Mycenae. Indeed, the precious metal is rare in the Mycenaean world as a whole, whereas gold has been found at numerous sites, and in far greater quantities. Indeed, silver (a-ku-ro, as part of the decoration of a set of wheels) is mentioned only once (!) in the extant Linear B texts (on Sa 287 from Pylos), although the ideogram WE (which might

• I thank the editors, K. Kleber and R. Pirngruber, for their invitation to contribute to this Festschrift to my PhD supervisor, Bert van der Spek. Much of the research for this paper has been done in Oxford and Amsterdam, and I thank my colleagues at those places for their help and inspiring discussions. In particular, I should like to thank Joanna Palermo and Lisa Bendall for their feedback and references. 1

The Shaft Graves, of course, turned out to be much older than Agamemnon's age (dating as they do to the beginnings of the Mycenaean period, to ea. 1650-1550 BC, whereas the stories of the Iliad in essence seem to hark back to the late Mycenaean period, shortly before 1200 BC), but their wealth remains a staggering testimony to the apparent power and prestige of the early Lords of Mycenae. The gold in the Shaft Graves is generally thought to have come from Egypt and the golden death masks similarly seem to hint at early contact between Mycenae and Egypt. However, many of the Egyptian objects in the Shaft Graves probably reached Mycenae via Minoan Crete (see already Karo 1930, 318; for a recent assessment Cline 2007, 193.) 2

Karo 1930, 225.

308

J.M.

MYcENAE, RicH IN SILVER

KELOER

indicate silver) is found on a handful of other tablets (including an interesting occurrence on Oa 734bis, where WE is written over what has been identified as an INGOT ideogram). 3 This scarcity of silver in the archaeological and Linear B record is remarkable, for the Mycenaeans did not have direct access to natural gold reserves, whereas silver could be extracted from the mines in the Laurion. Stos-Gale and Gale conclusively demonstrated that these silver reserves were already exploited from the (late) Middle Helladic period onwards, until at least Late Helladic Ill C (12-11 th centuries BC). 4 Despite its sporadic occurrence in the archaeological record, there can be no doubt that silver was highly valued in the Mycenaean world: indeed, some of the most iconic objects from the Shaft Graves, including the 'siege rhyton' and a rhyton in the shape of a hull's head (Fig. 1), were made of silver, whereas several later (14 1h-l3 1h century) procession frescoes from Knossos and Tiryns demonstrate the use of silver (and gold) vessels in religious ceremonies. 5 The sudden 'boom' of silverwork in the Shaft Graves, and the later scarcity of silver in the Mycenaean archaeological record thus requires an explanation.

SILVER AND THE RisE oF MYCENAE

Silver was used in the Aegean from at least the third millennium BC onwards. Early examples of silverwork include a superbly Fig. 1: Silver rhyton from Shaft made silver diadem from Amorgos (in the Grave IV at Mycenae (photo by Oren Cyclades), and a beautifully made two-hanRozen, CC License unported). dled sauceboat from Troy Ilg, weighing a staggering 600 grams. Chemical analysis suggests that, during the third and early second millennium BC, silver from various sources (in Anatolia, the Cyclades (Siphnos), and possibly already the Laurion) was used throughout the Aegean. Towards the end of the Middle Bronze Age, however, the Laurion seems to have become the main source of silver: chemical analysis

of a small number of objects from the Shaft Graves at Mycenae suggests that 80% of the silver found in those tombs originated from the Laurion.

6

The question rises as to how such a large quantity of silver from the Laurion arrived ~t Mycenae. It seems unlikely that Mycenae already exercised some sort of control over the Launon as early as Late Helladic I (the era of the Shaft Graves), and it seems equally implausible to attribute the wealth in the Shaft Graves exclusively to successful raids. Instead, it is more likely that most of the objects from the Shaft Graves were acquired in the process of gift exchange between the rulers of Mycenae and other potentates7 , most likely partners (or competitors) in a trading network centred on the Saronic Gulf, which connected the Argolid to Crete, the Cyclades, and the Near East. 8 At the centre of this network stood Kolonna on Aegina, which, by the time of the Shaft Graves, reached its zenith in terms of wealth and power. Testimony to Kolonna's wealth and importance as trading centre is the so-called Aegina Treasure (dated to ea. 1700-1500 BC). It seems likely that the rulers of Kolonna, Mycenae and other Aegean states, cemented their relations Fig. 2: The Saronic Gulf and nearby regions, by sending gifts, such as with LH I-III sites mentioned in the text (afterTartaron 2013). this type of precious metalwork. It is equally likely that Kolonna was instrumental in facilitating Mycenae's access to the Minoan world and that various objects from the Shaft Graves at Mycenae reached that citadel via Aeginite middle10 men,9 although direct contacts between the Argolid and Crete must also have existed. Although it remains unclear how and why Mycenae became involved in the Aeginite network, there is good evidence to suggest that, towards the end of LH 11 (i.e. the late 15th century BC), Kolonna was eclipsed by Mycenae. There is no evidence for significant building Sros-Gale/Gale 1982, 476. Voutsaki 1999, 110. s For a recent assessment of the Saronic trading network, see Tartaron 2013, esp. 236 ff.

6

3 C£ Ventris/Chadwick 1973, 351; Lujan 2011, 30. But see Olivier/Vandenabeele (1979, 152-153) who dismisses the identification of* 166 as an INGOT ideogram, whereas WE may more plausibly be identified as a designation for textile (wehanos). 4 Stos-Gale/Gale 1982. 5 On the procession frescoes, see Blakolmer 2007, 43; see also Weilhartner 2012, 223-4.

7

Tartaron2010, 172. The adoption of numerous Minoan elements in Mycenaean culture further testifies leading region in cultural, political and economic life in early Mycenaean times.

9

Jo

to

Crete's role as the

3IO

].M.

MYCENAE, RicH IN SILVER

KELDER

activity at Kolonna during LH liB, and although it is generally assumed the 'mansion' at Kolonna continued to function, it was evidently out of use in LH IliA (when a large kiln was constructed inside its remains). 11 The decline of Kolonna in LH II-IliA, which coincides with the rise of Mycenaean Troezen, may well be seen as a conscious Mycenaean attempt to eliminate a contender for regional hegemony and to establish more direct control over important trading routes. 12 Kolonna was not the only important centre in the area that suffered a setback in LH liB-IliA: numerous important sites in Attica similarly seem to have lost in importance, too. Thorikos, which had been graced with the so-called 'elliptical tholos' (or built chamber tomb IV) during LH I-liA and a tholos tomb in LH IIA, ceased to be of importance: there is no evidence for later substantial architecture (such as fortification walls, mansions or tholos tombs), although habitation at the site continued. 13 Other sites in Attica which had been of importance during MH-LHI, such as Kiapha Thiti, also ceased to be of significance between LHII and LHIIIA. Indeed, the abandonment or decline of hitherto 'princely' villages throughout LHII-IIIAl Attica is so widespread and marked, that it has been connected to Theseus's mythical 'synoikismos'. 14 Yet there is, in fact, no reason to suppose that this phenomenon was the result of Athenian intervention. Quite the contrary in fact, for there are precious few indications of substantial activity on the Athenian acropolis predating LH IIIB. 15

3II

Whilst the exact degree of control over this area may have changed per region and over time 17, there are reasons to believe that by LH IIIB (the 13th century BC), Mycenae's grip on the area was tightened. At Korphos: Kalamianos on the Saronic coast of the Korinthia, a large and monumental harbour town was constructed (pottery evidence suggests ?nly ~imited habitation on the site prior to LH IIIB) which is thought to have been Mycenae s mam port in the region. Daniel Pullen, one of the site's excavators, has described it as a 'short-lived maritime outpost, purposely founded as a component of state expansion in a climate of intense 18 peer-polity competition in LBA (fifteenth-thirteenth centuries BC) Greece.' Similarly, ~he sudden rise during LH IIIB early of the acropolis of Athens as the dominant centre of Atnca may be seen as conscious attempt to bring the region under palatial control: Hurwitt assumes (and I agree) that a Mycenaean palace was functioning on the acropolis in LH IIIB, but it was only towards the end of LH IIIB, in the last quarter of the 13th century BC, that the acropolis was surrounded by a cyclopean wall. The impulse for the fortification (and quite possibly for the establishment of Athens as the palatial centre for Attica as a whole), Hurwitt (op. cit. 80) suggests, may well have come from the Argolid, i.e. Mycenae. If we were to assume that all of the above is true (and the problem with all this is, of 19 course, that nothing can be conclusively proven or, for that matter, rejected ) and that Mycenae, from ea. LH 11-IIIAl onwards managed to control the entire Saronic Gulf, and thus had full control over the Laurion, we are faced with two questions:

It is tempting to see the decline of the early Mycenaean 'princely' villages in Attica as the result of Mycenaean intervention, aimed at gaining control over the Laurion mines and the sea routes to the Cyclades, Crete, and further east and south, to Anatolia and Egypt. After the decline of Attica in LH liB and the eclipse of Kolonna (where the mansion was abandoned) towards the end of the same period, Mycenae was left in effective control of the Saronic Gulf and, by extension, the mines ofLaurion. 16

1. What triggered the collapse of the early Mycenaean (MH late-LH IIA) system of peer polity interaction and gift exchange, and caused the rulers ofMycenae to expand their realm? 2. What happened to the silver that was extracted from the mines in the Laurion following Mycenae's annexation of Attica? In my view, the answers to these two questions are closely related, and have to be sought beyond the Aegean, in Egypt.

11 Gauss 2010, 476. But note that Kolonna remained a major exporter of ceramics to, e.g., Attica and northern Korinthia, until LH IIIA2: c£ Tartaron 2010, 175. 12

Cf. Konsolaki-Yiannopoulou 2010, 73.

13

Cf. Servais 1969, esp. 68 and n. 7. Servais (1971, 21-102) suggested that the LH IIA 'proper' tholos at Thorikos may have belonged to a new ruling house at Thorikos. 14

Lohman (2010, 44-45) prefers 'sympoliteia', noting that some of the settlements remain inhabited (although clearly of lesser status), whilst there is no evidence for early concentration of settlement at Athens.

15 Hurwitt 1999, 72; although Mountjoy (1995a, 16) notes that two sherds in the so-called Palace Style that have been found on the acropolis suggest that the settlement there must have been of some importance. 16

N. Papadimitriou (2010, 255) similarly suggests a Mycenaean take-over ofThorikos. I consider Mycenae to have been the dominant centre- the 'capital'- of the Argolid by LH IIA at the very latest (when the first tholos at that site was built). The funerary record also points to the early pre-eminence of Mycenae: S. Voursaki (2010, 97 ff.) observes that the rise of Mycenae in LH II is accompanied by the rise of several (secondary I dependent) sites, and that burials are notably wealthy in the eastern part of the Argolid (i.e. the region around Mycenae), although she argues that only during LH IIIA2 the process of concentrating wealth at Mycenae reached its final stage.

17 Tartaron (2010, 171: see now also Tartaron 2013) notes that the northern part of the Korinthia, as far as the archaeological record is concerned, shows very little Mycenaean impact until LH II~, wh.en the export of pottery at Kolonna declined and local pottery production started to show strong. h~ks Wlt~ :he Argolid. It should be stressed, however, that archaeology is a problematic ~o~l for estabhshmg polmcal boundaries and limits of'imperial control' (cf. Kelder 2013, 41-52), whereas Hmtte texts (cf. below) suggests

a very early expansion of Mycenae. 18

Pullen 2013, 245.

19

As I have argued elsewhere (Kelder 2013), archaeology alone is not a very reliable tool for identifying

state formation and expansion. The same goes for the Linear B evidence (although there are some clu~s, most notably the uniformity of script and administration throughout Greece, that suggests a degree of umty towards the very end ofLH IIIB (cf. Postgate, 2013,412 [and references therein]), which is too late in date anyway to be of much use in reconstructing the development of the early Mycenaean state.

312

}.M.

MYcENAE, RicH IN SILVER

l