468 4.

Agriculture: the rural landscape, regional trends and diachronic change

land-holding systems already in the seventh century as well as with other parts
of the country in which there was little evidence for change and expansion at all. 5. Silver, silver money and money-based exchange
Nevertheless our case studies throw much light on the socio-economic back- 5.1. Introduction
ground of the initial phase of the long-term development that led to an unprece-
dented density of settlement and intensity of land use over the course of a mil- The criteria employed in most attempts at defining "money" in the context of the
lennium, into the Sassanian I early Islamic period. The texts necessarily present discussion of pre-modem economies2560 are owed in large measure to K. Po-
lanyi. A commodity can be designated as money, in this tradition of thinking, if
a far more fragmented and contradictory image of this process than the (simpli-
it serves as a generally accepted means of exchange especially by providing a
fied) model of this process based on the archaeological sources would suggest:
standard for attributing a quantifiable value to goods, if it is used as a means of
this development was certainly neither linear nor continuous, it was frequently
payment itself, and if it is used for 'hoarding,' i.e., for the storage of wealth? 561
interrupted, even reversed, for various reasons. Nevertheless, we can identify in
A particular Polanyian refinement to this definition, which has proved influen-
the sources the fundamental elements - apart from demographic growth as such,
tial in Ancient Near Eastern studies, involves his distinction between "all-
which is visible indirectly in the written documentation - which contributed to
purpose money" uniting all money criteria, and the "special-purpose money"
this development. They are not in evidence everywhere to the same degree, but
typical, according to Polanyi, of early societies including those of Ancient
they are sufficiently common to be clearly recognisable: large-scale royal in-
Mesopotamia, where individual commodities used as money are thought not to
vestments in the infrastructure resulting in a restructuring and expansion of the
have exercised all money functions at the same time? 562
cultivated area, and widespread intensification of production through a focus on
The most recent general studies on silver and money in first millennium
labour-intensive date gardening and horticulture, both on private and institu-
Babylonia were written by A. Bongenaar (1999) and P. Vargyas (2000, 2001).
tional land, to the extent that the urban elites, an important category of landown-
In addition to focussing on price development, Vargyas pays particular attention
ers, have only little or no interest in arable farming. Given the higher productiv-
to the different silver qualities and to the role of coinage in the Ancient Near
ity of date gardens in comparison with grain fields, this process necessarily re-
sulted in an increase in agricultural productivity per capita - and thus, once the East. He places emphasis on the fact that even coined silver, when it makes its
undoubted appearance in the Hellenistic period,2563 continued to be weighed,
process was sufficiently generalised, in an increase in the agricultural surplus
rather than counted, in the age-old Mesopotamian tradition. 2564 For Vargyas, this
produced by the society and thus in general economic growth. The sixth-century
evidence, especially from Sippar, shows this general shift towards the estab- fact is irrelevant for the definition of money; he sees no fundamental difference
lishment of dates as a 'leading crop,' on par with barley, as an on-going process between the use of silver in the Hellenistic period and in earlier centuries of the
first millennium BC, and concludes that "we have to draw the conclusion that it
and thus constitutes a model for the development of a new kind of agrarian re-
gime. One can even speak of an 'agricultural revolution,' as there is no evidence
for any earlier period of Babylonian history having seen such a development on
such a scale. Given the ecological and technological constraints of pre-modem °
256
From an Assyriological point of view, see esp. Renger 1995 and Powell 1996. See
agriculture in southern Iraq, this gradual shift to a regime with two principal also Stol2004: 900ff., especially 908f., for a sober critique of the material basis (regard-
ing the Old Babylonian period) on which Renger's conclusions rest.
crops of equal importance was decisive for maintaining, and increasing, the high 2561
These criteria are generally presented in a slightly abbreviated form, as "medium of
population density and level of urbanisation which characterise Babylonia in the
exchange, store of value and unit of account," in current reference works and text books
late first millennium BC and in the early first millennium AD. A new phase of
for general economics: e.g., Bannock et al. 1998: 280, Lipsey and Chrystal 1999: 428.
fundamental agrarian change probably came only in the Islamic period, with the An alternative definition, which owes less to classical economic theory, considers money
2559
introduction of new crops (such as sugar cane or indigo). But even this Mus- as "a means of payment in economic, political and social contracts," see von Reden
lim 'Green Revolution' can be seen as a continuation of a basic paradigm which 2007: 4+19. This definition places less emphasis on the use of money to facilitate the
took shape in our period: the new crops were 'industrial crops' which were exchange of goods and contributes to dissociating the issue of money and monetisation
grown largely for the benefit of expanding urban markets. We will see in the from the discussion of the existence, or absence, of 'markets' in pre-modem (money-
following chapter that also during our period, ubanisation and an increasing using) economies. For our purposes in the present context, these are two separate prob-
monetisation of economic exchange were closely linked to defining features of lems.
2562
the agrarian regime, such as cash-crop production and commercialisation of E.g., in the Old Babylonian period, according to Polanyi, barley was the means of
agriculture. payment, silver the standard, and barley, silver, wool and other commodities were used
as a means of exchange. See Bongenaar 1999: 160.
2563
We WI'II re fier t o Vargyas ' mterpretat10n
. · of the Achaememd · kaspu sa ginni as coi-
nage below.
2564
2559
See below section 6.2.3. See Haf"'10rd20 05a: 369+31 for the term "weight-money."
.
470 5. Silver, silver money and money-based exchange 5.1. Introduction 471

was a money economy that ha d deve1oped m . h century Baby1oma

. s1xt . ,2565 - a
are vety similar to those ofPolanyi and his followers? 571
money economy based on, and operating within, "a smoothly operating market (1) money has the power to meet 'social obligations' (another way of refer-
2566
system." Bongenaar's study, after surveying the discussion, in particular the ring to the use of money as a means of payment and of exchange) and is in gen-
various definitions for "money" in the tradition of Polanyi, and after giving a eral valued and stored for future use for this reason, rather than for its "use-
brief overview of the use of silver in Neo-Babylonian institutional archives, value" (as a raw material for prestige items, for instance, in the case of silver);
concludes that "silver served as money, although not exclusively"- staples were (2) it tends to be quantified;
used in the same way. The basic criteria for "money" are met, but silver was not (3) it provides a measure of value;
"all-purpose money," it was not universally acceptable. "Silver could never (4) it should have general acceptability (i.e., Seaford is unwilling to accept
develop into an everyday money because its value was too high in comparison the Polanyian "special-purpose money" as 'real' money);
. other commod'I ties,
wtth . , s1'1 ver was " a h'tgh-range money. ,2567
(5) it should (may?) have exclusive acceptability (by virtue of being the only
Within Assyriology, there is a near-consensus that it is legitimate to speak of substance to unite all the criteria listed above);
silver as "money" for most periods of Mesopotamian history. M.A. Powell ar- (6) money may be fiduciary, in that it is attributed by convention with a
gued that in northern Mesopotamia, silver, in conjunction with other, cheap value over and above its intrinsic value; the essential point is that there has to be
metals (lead, copper [bronze], tin), was very close to forming a "pre-coinage a "general belief in the future acceptability" of this conventional value; 2572
money system," whereas in Babylonia, silver, "the ultimate standard of liquid- (7) the state may be involved in guaranteeing that money be attributed the
ity," was complemented by barley, and in the first millennium by dates, as a characteristics listed above.
form of cheap money? 568 These general conclusions were also followed by These definitions are then applied by Seaford to material from the "early
2573
scholars from neighbouring fields, especially Ancient History, studying the de- Near East" taken from a (quite thorough) survey of the pertinent secondary
velopment of coinage or the history of money before coinage?569 D. Schaps, in literature in his discussion of whether "money" was used in this area. 2574 Pre-
his study of what he calls the "monetization of Ancient Greece," accepts on the dictably, his answer to the question is mostly negative. First of all, and one can
basis of the usual criteria that silver functioned as money in Mesopotamia, but of course agree on this point, coinage was absent from Mesopotamia in the pe-
he emphasises that it always remained a commodity and never came to represent riod Seaford discusses. Also he sees little evidence for standardised and/or
only money. This limited its effect and its spread throughout society: it was not marked pieces of silver or other metal. The absence of coinage and the need to
"synonymous with wealth itself."2570 The debate was taken up and partly turned weigh silver for every transaction is seen as a major impediment to its wide-
in a new direction by R. Seaford (2004), who, in the context of evaluating the spread use. In the same vein, while Seaford is willing to accept that silver served
impact of the use of money on early Greek society and thought, discusses not as a standard of value, he disagrees with (for instance) van de Mieroop's claim
only the usual definitions of money, but also gives a detailed assessment of prior that (in OB Ur) the use of physical silver in payments was not infrequent. 2575 He
attempts to argue for (or against) the existence of money in different periods of quotes Renger for the assumption that even when silver was used to indicate the
Near Eastern history. As it will be useful to engage with some of his arguments, value of an item to be bought, the actual payments were made with barley or
we will repeat the characteristics Seaford attributes to money even though they other money media and that silver was not a general means of exchange: in other

2565
Vargyas 2000: 520.
2566 2571
Vargyas 2001: 284. See note 2561 above. See Seaford 2004: 16ff. We have already briefly referred to his arguments in Jursa
2567 25
Bongenaar 1999: 174. As an example ofthe impracticability ofthe use of silver for 2005b: 176 . See now also Seaford 2009 for a convenient summary of his views.
2572
everyday transactions he states that "[a]s long as one shekel of silver still constituted a Seaford 2004: 7. This is a cornerstone of his overall argument.
month's pay, small products at the local market had to be bought with other, 'cheaper' 2573
His cut-off point is the fall of the Assyrian empire in 612 BC, which roughly coin-
commodities." Bongenaar's conclusions are cited with approval in the final pages of cided with the invention of coinage in western Anatolia (Seaford 2004: 125ff. on his
Milano's essay on the methodology of the investigation of the circulation of goods in view of who invented coinage and when; see also Schaps 2004: 93ff.).
ancient Mesopotamia (Milano 2003: 128). He insists rightly on the need to study as a 2574
Seaford 2004: 318ff. We need not go into Seaford's treatment of the Old Assyrian
whole the economic system(s) which determine(s) the use of money (media), rather than evidence (mostly as presented in Veenhofs Aspects of Old Assyrian Trade and its Ter-
the technical aspects of their use in isolation.
2568
minology [1972]). It contains a number of misunderstandings which might have been
Powelll996; summary in Powelll999: 14f. avoided by referring, i.a., to Dercksen's The Old Assyrian Copper Trade in Anatolia
2569
See, e.g., Bulgarelli 2001, where much material from three millennia of Mesopota- (1996).
mian history is assembled to support the argument for accepting (weighed) silver as pre- 2575
Van de M'1eroop' s actua1 words are even more cautious
. than our paraphrase: "silver
.
coinage money. must have played a role as a means of payment, and it cannot have been used solely as a
2570
Schaps 2004: 42ff.; the quote is from p. 51. measure to calculate equivalences ... " The quote is from van de Mieroop 1992: 248.
 5. Silver, silver money and money-based exchange 5.1. Introduction 473
472

words, silver is denied one of the essential characteristics of 'money. '

2576
Redis- because of a substantial margin of error in the small weights, 2580 corresponding
tribution and barter left little need for money-based exchange, which also had no amounts of silver would not have changed hands regularly; only quantities of
role to play in the essentially undocumented sector of the economy constituted one shekel or more could have been weighed with sufficient precision. He adds
by subsistence farming and mral exchange.
2577
Evidence from first millennium that few people could afford to buy something for such an amount of silver, and
Babylonia (which strictly speaking is mostly beyond Seaford's chronological therefore Powell "cannot imagine silver circulating from hand to hand on a daily
focus) is cited to the effect that "despite this relatively widespread use of silver, basis throughout the whole populace - including mral areas - in the Old Baby-
which is as good evidence for money as we have seen so far," "the value of lonian period or in any later period. That, however, is not the same thing as say-
silver was too high for it to develop into an everyday money," which meant that ing that its role in the economy was insignificant" (Powell 1999: 16f.). We will
barter was used for "the most usual exchanges."
2578
In his summary, he evalu- return to this matter below.
ates his findings in light of the above criteria for defining "money." In particu- Every attempt to establish the range of the "silver economy" also implies
lar, he does not see a general preponderance of 'money value' over 'use-value' necessarily a discussion of the use of other money media, in particular that of
for any of the proposed money-stuffs, including silver (criterion 1). He assumes barley, dates and to a lesser extent wool, and in even more general terms, of the
. . o f the economy as a whole. 2581 It can be seen even from a superfi-
monetlsahon
that while silver was used as a measure of value, this was not universally the
case (3), and emphasizes that the acceptability of silver as a means of payment, cial survey of any corpus ofNeo-Babylonian texts that we are dealing with what
while difficult to assess, was determined by particular contexts, in certain could be called a binary economy, an economy characterised by an interdepend-
spheres of the economy, and was not general (4). Exclusive acceptability (5), if ence of economic spheres relying on exchange in kind and on exchange in cash
it occurred at all, would have been limited to certain very specific cases. Fiduci- (silver), respectively. However, the details of this interdependence, the internal
ary money (6) was entirely absent, all commodities performing certain money dynamics of the system and its diachronic development have never been dis-
functions remained commodities and retained a use-value; institutional (state) cussed systematically and comprehensively. A step in this direction shall be
guarantees for fiduciary money, or even for guaranteed weight or quality of attempted here.
We begin with the most important money medium for the purposes of the
silver, were unknown (7).
It is not intended to offer here a detailed critique of Seaford's findings; we present chapter, silver. The discussion focuses on the forms and qualities in
are mostly interested in the question he asks. His criteria for money are narrower which it circulated, and on the role of the state and more generally the institu-
than those of several other scholars, but this has the advantage of bringing the tions in determining these forms and qualities. In the following sections on the
problem into clearer focus. We shall mostly engage with criteria 1, (3), 4, 5 and actual use of silver and other money media, it is obviously neither possible nor
7. This means that the discussion will focus mostly on silver as a means of pay- desirable to attempt a complete overview of the evidence: the material is vast.
ment, on the contexts in which it is used in this way, and, as far as possible, on Because of this even the most careful studies published so far offer impression-
the frequency and cumulative importance of such transactions. The problem one istic observations on which it is hard to base warranted generalisations. Here we
faces when evaluating the criteria for "money" is primarily one of gradual dis- shall try to avoid the pitfalls of such an approach by focussing closely on se-
tinctions based on quantification rather than of ('binary') qualitative judge- lected bodies of particularly promising evidence, or on particular types of eco-
ments: qualitative judgements can only follow from quantification. In other nomic transactions.
words, we will try to arrive at a clearer understanding of the range, and therefore Records from the eighth century are studied first with the hope that they will
also of the character, of the silver-based sector of the economy in the first mil- put into sharper relief the later data on which the rest of the discussion will be
lennium. One key issue follows directly from Seaford's summary of Bon- based. The sixth-century temple archives give the most nuanced view of the
genaar's and Dandamaev's work - is silver really essentially "high-range
money," causing most everyday transactions to rely either on other money me- als to obtain goods for their daily needs through barter, and a large-scale exchange sys-
dia (barley, wool, dates) or on simple barter? In general terms the problem is tem for institutions to which the silver equivalences quoted in the Ur III merchant ac-
posed very clearly by Powe11. 2579 He assumes t hat 10r ~
teehn"tea1 reasons, 1.e.,
. counts refer. Silver was not suitable for small-scale exchange in the local 'market.'
2580
He states that statistically the error would have been on average about three percent
2576
Seaford 2004: 324f. of the total mass of the (small) weights. For the problem of the accuracy of Mesopota-
2577 mian weighing see now Bry 2005: 190. Note that this inaccuracy of weighing technology
Here Seaford explicitly follows Renger (1995 etc.).
2578 Seaford 2004: 333. He uses mostly Bongenaar 1999 (whence the [implicit] refer-
was as likely to work to on individual's benefit as to this detriment.
2581
ence to silver as "high-range money") and Dandamaev's essay in the Cambridge Ancient It should be emphasised again that payments in kind are not (necessarily) 'barter.'
For definitions of 'monetisation' and references to treatments of the problem in Ancient
History, 2III/2, of 1991.
2579 Powell 1999: 16f. The matter is taken up by Widell2005. He assumes that in the Ur
History see most recently von Reden 2007: 3r+ 17 (citing, i.a., Schaps and Seaford) and
3lff.
III period, there were two exchange systems: one for local exchange, allowing individu-
474 5. Silver, silver money and money-based exchange 5.2. Silver qualities 475

interplay of 'cash and kind' in a particular sector of the economy. Our discus- of silver and staples and the degree to which their range of use was complemen-
sion of this material starts with complete surveys of all transactions attested in tary or overlapping will be discussed in the following sections; here we will deal
the Eanna and Ebabbar archives for certain years. Subsequently it turns to some with the descriptions of the physical characteristics of silver in its use as money.
types of institutional business which are of particular relevance in this context. The various qualifications attributed to silver in texts from the seventh cen-
For the non-institutional sector of the economy, chapter 3 contains a similar, if tury onwards have been treated extensively in the secondary literature, so a brief
more general survey of types of economic activity and their relative importance: survey emphasising contentious interpretations and a few new aspects will suf-
the results are briefly summarised here. The problem of silver as 'high-range fice.2586
money' and the issue of 'the market' and market exchange are addressed in l~ght For convenience, the results of these discussions are summarized in the fol-
of some new evidence; otherwise, money as a means of exchange needs httle lowing table.
further discussion.Z 582 The use of money for payments is studied for three con-
texts of particular importance: labour, taxation, and agriculture and agricultural pe$ii "white, " ~ non-oxidised?; no indication of a standard fineness
rents. The information on payments in these areas is far richer than the docu- qalii "refined, ~ examined," no indication of a standard fineness (higher than
mentation for everyday monetary exchange, especially for buying and selling on that of"white" silver)
the city markets. 2583 Several sections will deal with the changes in prices and nuhhutu "hacked," "scrap (silver)," not "alloyed"
wages over time and with the implications of the partly dramatic changes in sibirtu "ingot (of stated/known fineness)," sometimes also "cut up for use in
price (and wage) levels that can be observed in the course of the sixth century. pieces (and sealed in a bag)"
tuhhii an 'undesirable' form of silver, perhaps oxidised, 'ugly' silver, silver slag,
5.2. Silver qualities or silver whose quality has only been examined by a touchstone and not
Silver, barley, dates and wool were used as money media in Babylonia in the by controlled refining/alloying; not simply "refined"
first millennium BC. Gold was only very rarely used as a means of purchase (in (silver) sa probably "marked/stamped silver, " originally perhaps intended for tax
2584 1 (. ginni payments or the like, from the reign ofNabonidus onwards; not identical
institutional contexts) and never as a standard of value. Base meta s tm,
bronze) likewise play no role in this context, at least in the written record, which with the Croeseids and Darius' sigloi
2585 (silver) erbu silver designated as "income of the 'treasury',"~ silver whose fineness
does not entirely exclude their use as low-range money. The extent of the use
sa aranni and perhaps weight is guaranteed by the royal treasury; the way this guar-
2582 For the simple reason that important transactions, in particular slave, animal and antee was conveyed is a matter of conjecture
real estate sales, are overwhelmingly conducted in silver. Table 53: some common des1gnatwns of silver
2583 Note that the same is true for papyrological evidence from Ptolemaic Egypt: von

Reden 2007: 79. This is owed to the structural similarities of the documentary evidence Explicit (i.e., quantified) references to the fineness of silver appear around
from the two periods and cultures (the main difference being the availability of a large 600 BC; only late in the reign ofNebuchadnezzar do they become frequent. 2587
amount of material on land sales from Babylonia which has no counterpart in Egypt). The most common type was silver of 87.5 %fineness, described as silver with
2584 Very rarely debts were discharged in gold (YOS 6, 242; 11 Nbn). What one does
one-eighth alloy, sa ina iSten siqli bitqu, literally "which (contains) one-eighth
find are texts like BIN 1, 149 (12 Npl, Uruk): Eanna acquires gold and silver for wool; (of a shekel) (of alloy) in (each) shekel." 2588 From the end of the reign of Nebu-
the total price is stated in silver, the gold having been converted into silver according to a chadnezzar onwards and at least until the end of the reign of Xerxes, when this
stated rate. It would be a fallacy to state that from the viewpoint of the recipient of the
wool this was a purchase of wool with silver and gold: there can be no doubt that gold
was ~rimarily a commodity and not a money, its 'use-value' (as a prestige item) vastly
of tin in this list (the reading is probably correct although the signs are heavily damaged:
out-weighs its value as a means of exchange.
2585 The sale of a house by Eanna for large amounts of copper, tin and purple-coloured
there is no plausible alternative) suggests that this metal was more common than one
would expect on the basis of its use-value (which is restricted to metal-working): hence
wool documented in YOS 19, 1 is exceptional (for the context see Dandamaev 2006; see
the assumption that it may have been mentioned here because of its use as a means of
also Joannes 2005b: 44f.). If base metals were used as low-range money only, they
exchange.
would have remained largely unrecorded in the texts. An explicit reference to bronze 2586
coinage is found only in an astronomical diary of the third century, when describing a Joannes 1994 and Vargyas 2001 are the most important studies. For silver in the
situation of silver scarcity caused by political intervention (van der Spek 2007: 418; the material record from Mesopotamia, the survey of the evidence given by Moorey 1994:
word used for "coin" is zfpu). Note that Nabonidus' tariff stela, which quotes the (fa- 232 is most useful.
2587
vourable) rates for the most important commodities allegedly prevailing during the An early unpublished reference (16 Nbk) to silver with 1/8 and 1112 alloy can be
king' s reign, probably mentions tin (one mina for one shekel of silver) after barley, dates, found in BM 26706 (Borsippa).
2588
sesame, cress and wool and before imported wine (Schaudig 2001: 531). The inclusion Joannes 1994: 141.
 5. Silver, silver money and money-based exchange 5.2. Silver qualities 477
476

. 2589 E l (sa ina iSten siqli bitqu). Vargyas also claims that "white silver cannot be pure
phrase falls (nearly) out of use, th1s was probably the standard fineness. ar y .1 ,2595 Th 1 . IS
. mcorrect,
. 2596
Sl ver. e atter pomt and the former not compelling: if
in the sixth century, especially in Uruk, silver of 83.33 % fineness (i.e., silver
with 116 alloy) was used; 2590 in Uruk at least its use was discontinued suddenly pe$U had implied a specific degree of fineness, it would have rendered the ex-
591 plicit statement of the fineness redundant. Since pe$U can demonstrably also
towards the end of 39 Nbk, certainly as a result of official intervention.Z Other
degrees of fineness occur rarely: silver of 91.66% (1/12 alloy) is attested late in apply to refined, pure silver,2597 it seems best to assume that the word simply
the seventh and early in the sixth century, 80 % fineness appears in the last dec- denotes the metal's colour without implying any specific degree of fineness.
ade of Nebuchadnezzar's reign. Silver of 95.83 and 90 % fineness is attested The adjective qalU is used nearly exclusively for silver from the beginning of
592
only in a few instances in the Achaemenid period.Z The fact that silver quali- the reign of Artaxerxes I onwards (although it also appears in earlier texts); at
ties are mentioned with growing frequency in the sixth century has two obvious the same time, "white silver" and explicit references to alloys, in particular the
2598
implications: first, the fineness of silver was increasingly a concern especially in bitqu phrase, nearly disappear from the record. The implications of this fact
the realm of the private economy (whence the majority of references originate; depend on the interpretation of qalU. There is agreement that the word, when
temple archives are more sparing with such qualifications, although they do use referring to silver, designates silver that had been treated in the smelting oven
them), and second, it must not only have been possible (and frequently neces- where it had been refined to a certain degree, but it is less certain that it neces-
sary) to verify these qualifications, 2593 but the need must have been felt to find a sarily refers to silver with a higher degree of fineness than the frequently at-
way to guarantee the quality of circulating silver. Therefore it has been argued tested (white) silver of 87.5 %, as Vargyas believes?599 He points to a few texts
that several of the other qualifications applied to silver refer either to the fine- which contrast "white silver" and qalu silver, but also refers to what he consid-
ness itself or to a guarantee of a certain degree of fineness. ers the contradictory evidence of Dar. 84 (3 Dar), in which "white qalU silver" is
Such is the case for the two most common qualifications of silver, pe$i'l and attested. In other cases, qalu and pe$i'l interchange. 2600 As white silver is not
qalU. However, these words only refer to the silver's colour and to the fact that it necessarily silver of 87.5 %fineness, there is no problem here. However, new
has undergone a process of refinement and examination. "White silver" (kaspu evidence is found in the unpublished Borsippa text BM 103493 (31 Dar): in this
pe$i'l), the most frequent kind of silver in the seventh and sixth centuries, is taken debt note, twelve shekels of silver are described as pe$i'l qalu sa ina isten siqli
by Joannes simply to be non-oxidised silver, in contrast to blackish silver which bitqu nuhhutu sa ginni "white, 'refined' silver with one-eighth alloy, nuhhutu,
594
would have been considered less attractive? In contrast, Vargyas assumes that with the ginnu characteristic" (on the two final terms, see below). This confirms
"white silver" is silver of 87.5 % fineness. He bases his argument oil the obser- Joannes' supposition that qalu silver can be of the same fineness as bitqu silver
vation that whenever the fineness of white silver is stated explicitly, it is 87.5% (which is frequently white, as in this case). 2601 On the other hand, qalU is also
2595
2589 Vargyas 2001: I8ff. For references to bitqu silver post-dating Xerxes see below Vargyas2001: 18f.
2596 •
note 2598. This silver is intended by the one attestation of "Akkadian," i.e., 'native,' "Remes Silber ist weiB": Meyers GrofJes Konversations-Lexikon 18 (6 1909): 463b.
2597
silver, according to Vargyas 2001: 38. Note however that the text in question (Nbk. 38,4 Jursa 2006a. Note also CT 57, 98 (Sippar, reign of Darius, date lost): I ma.na 14
Nbk) antedates the regular use of indications of fineness. The distinction between qa/U gin ku.babbar r.sa gin-nu1 [a-na I] ma.na 3 gin ku.babbar babbar-u ma-nu "one mina,
silver and akkadu silver may simply be between "refined" silver (or "silver treated by a fourteen shekels of ginnu silver are reckoned as [one] mina and three shekels of white
silversmith"; for the exact meaning of the word see below) and "common," i.e., un- silver." This seems to indicate that in this case white silver was refined, pure silver
treated, silver. whereas ginnu silver had 85 % fineness.
2598
2590
Vargyas 2001: 35f. See Joannes 1994: 139+6 , Vargyas 2001: 14f. and 18 81 for references to "white
2591 silver" from the late fifth century (e.g., BE I 0, I; ace Dar II). A late attestation of sa ina
See below p. 598f. 105
2592 For all this see Vargyas 2001: 35f. (and note 2587 above). iSten siqli bitqu appears in BE IO, 77 (3 Dar II); Vargyas 2001: 20 , another, dating to 8
2593 The most obvious way of testing silver for impurities is to refine it; this is amply
Art (which?) in Stolper 1990a: 582 no. 17 (contrasted to gi?u silver of 95.8 % fineness;
see Joannes 2006: I 09).
attested in the temple archives (see, e.g., Bongenaar 1997: 362f., Vargyas 2001: 40f.), 2599 v ' 1994: l39f. on the other hand disagrees; for him qa/U
but this of course could not be done every time a silver quality was in need of verifica- argyas 2001: 13ff. Joannes
tion. It is probable, but not provable, that Babylonian silversmiths in the sixth century silver, as used for exchanges, has a fineness of 80-90 %.
26oo E .m the pair
.g., 'BM 92734 (=ABC 33 = Waerzeggers 2000/2001 no. 30) and BM
used the touchstone; the most likely Akkadian term for the touchstone is only attested in
lexical lists: aban tasnfqti "stone of verification." Note that an Old Babylonian touch- 79641 (= Waerzeggers 2000/2001 no. 24). The latter text (6.3 Dar) is a debt note for 28.5
stone was in all likelihood found in the context of the Larsa goldsmith hoard, see Huot shekels of qa/U silver and seventeen shekels of ginnu silver which is said to have been
1995: 268f. (reference courtesy of D. Charpin), where the original interpretation of the paid in the former text (10.3 Dar): there however the 28.5 shekels are said to be "white,"
Larsa stone is defended against the the counter-arguments of J. Bjorkman cited by Var- pe$U.
2601
gyas 2001: 36 .
173 But it should be noted that it might be advisable to treat this text with caution as it
2594 contains another terminological oddity in addition to the collocation of pe$U and the
Joannes 1994: 139.
 5. Silver, silver money and money-based exchange 5.2. Silver qualities 479
478

used in the late period for Greek coins of much higher fineness. Therefore all not refer to a particular degree of purity, it seems simplest to suggest "cut,
that one can safely conclude is that qalu refers to silver that had been manifestly clipped" on the basis of the literal meaning of the verb nuhhutu, reverting to a
treated by a silversmith. 2602 It is unclear whether the term referred to silver of a position held (e.g.) by Powell in 1978. 2607
certain purity at any given period; it certainly did not denote just one kind of Pieces of cut-up silver items - i.e., Hacksilber - are found for instance in the
silver throughout the period under discussion. The terminological shift regarding Babylon hoard (Reade 1986). This explains the double character of sibirtu. This
silver, i.e., the replacement of most common tenns for silver by qalU, which word, which frequently is used to designate (large amounts of) silver, means
occurred after the reign of Xerxes, surely is owed to an intervention by the cen- "block, piece, lump," according to the definition given by CAD S/2. The dic-
tral administration under Artaxerxes I, but it has to be acknowledged that it re- tionary however frequently translates "(silver) in pieces," as indicated by ety-
mains uncertain of what these royal measures consisted. There was most likely a mology and sometimes context. sibirtus in the first millennium can be counted
standardisation of the forms in which silver was allowed to be used as a means and said to have a uniform purity, which suggests "ingots" (or, following Pow-
of exchange, but there is no proof that the fineness of silver was increased by ell, coins sealed in a bag2608 ). They also frequently are said to be nuhhutu - "cut
royal decree from the (probably) common 87.5% to over 90%. up."2609 This does not mean an ingot formed of scrap silver,2610 but rather the
The adjective nuhhutu qualifies silver either alone or together with the adjec- opposite: an ingot (of known purity) cut up for use (and conceivably sealed in a
tive "white" and/or the phrase sa ina isten siqli bitqu "which contains one-eighth bag). 2611 This is illustrated particularly well by Ner. 15 from the Nur-Sin ar-
alloy" and/or the statement that the silver lacks the ginnu characteristic (sa lii chive,2612 in which 20,200 strings of onions are owed by (presumably) a farmer
ginni)_2 603 The meaning of the word is debated. It is certain that nuhhutu does to a businessman representing a business company (harriinu). One would as-
not mean "alloyed" and should be understood together with the sa ina iSten siqli sume that this was one of the common pre-paid purchases found in the archive
bitqu phrase which frequently precedes it (and of which nuhhutu, when used were it not for two unusual guarantee clauses. The second states that the creditor
2604 guarantees for the payment of thirty-one shekels of silver of 87.5 % fineness-
alone, would then be an abbreviation), as was suggested by Powell, who is
followed by Vargyas. The verbal adjective can precede the bitqu phrase and is this is a part of the price of the onions which was credited by the seller to the
therefore not syntactically connected with it.
2605
Furthermore, as was already buyer. 26!3 The first guarantee clause states: pu-ut nu-uh-hl-tu
.
4 (collated by Geers
noted by Landsberger, the silver characteristic expressed by nuhhutu is incom- and Wunsch) sa [sli-bir-tu 4 sa ina 1 gin bit-qa sa ina igi (debtor) (creditor) {era-
patible with' the ginnu characteristic, although the latter silver type is frequently sure} na-si "(the creditor = buyer) guarantees for the cutting-up of the silver
said to contain one-eighth alloy (sa ina iSten siqli bitqu). nuhhutu silver can only ingot of 87.5 %fineness which is already at (the debtor= seller's) disposal."2614
be sa lii ginni - excepting the unique text BM 103493 cited above, which also This means that the seller of the onions had received an ingot of a certain size in
combines pe$U and qa!U. This distribution cannot be a coincidence and proves one piece as payment in part for his produce. As he could not use the ingot in
that nuhhutu expresses a different feature of silver than the bitqu phrase. On the this form in his rural enviromnent (it being too valuable) he required that it be
basis of the observation that nuhhutu silver could not be ginnu silver (and on cut up. For this he preferred to rely on the expertise and/or contacts of the city-
somewhat dubious etymological grounds) it was suggested that the word refers
to silver of mediocre, or medium quality, although it remained undetermined of 2607
See CAD N/2 s.v. nuhhutu "to trim, clip" also appears in Late Babylonian. Note
2606
which physical characteristic this supposed mediocrity consists. However, as that the dictionary cites one Nco-Babylonian reference to linen, or in any case a textile,
ginnu silver is not particularly "fine" silver (see below) and also nuhhutu does which is said to be Ia nu-hu-tu. Powell 1978: 225 suggested "clipped, struck off, cut off''
for nuhhutu but later (1996: 232f.) abandoned this interpretation in favour of connecting
nuhhutu with sa ina muhhi bitqu, which, as was pointed out above, is impossible for
bitqu phrase on the one hand and qa!U on the other: ginnu silver is otherwise never quali-
syntactical reasons.
fied as nuhhutu. See below. 2608
2602 It is therefore a synonym for "smelted" (petqu) silver, which occurs rarely in non- Powell 1999: 22f., against his reasoning in 1996: 237. The argument presented here
administrative contexts. BM 102291 (Borsippa, Achaemenid period) is a debt note for makes this assumption unnecessary.
2609
one mina of silver of 87.5 %fineness (sa ina isten siqli bitqu); the silver is said to con- See CAD S/2 s.v. Note also Jursa 1999: 145 for a reference to sibirtu silver which
tain sixteen shekels of 'smelted' silver of the same fineness (ina lib-hi 16 gin ku.babbar is not only nuhhutu and of 87.5 % fineness, but also explicitly non-ginnu.
2610
pe-et-qu sa ina 1 gin bit-qa). The characteristics by which this petqu silver was recognis- Powe111996: 237.
2611
able must have been the same as those of qalu silver; they do not seem to have consisted See Joannes 1994: 142f.
2612
of a particular degree of purity. Wunsch 1993: no. 65; collations for lines 10 and 24 (by F.W. Geers) reported in
2603
See, e.g., Vargyas 2001: 18ff.; Joannes 1994: 139. CAD Sill: 380a.
2613
2604
Powel11996: 232f. Wunsch 1993: I 31.
2614
~605
Examples are in fact cited by Vargyas 2001: 18 sub 2. Because of its syntactical position, nu-uh-hi-tu must stand for nuhhutu the infini-
2606
See the references given by Vargyas 2001: 21 •
108 tive, rather than for nuhhutu the adjective.
480 5. Silver, silver money and money-based exchange 5.2. Silver qualities 481

2620
based businessman rather than on his own means (it may be that he wished to anteeing its quality. This quality was not particularly high, as silver intended
avoid even miniscule losses as a result of cutting up the ingot). Read in this way, for exchange was generally of lesser purity than silver used for silversmithing-
the text not only helps clarifying a point of detail in the silver terminology, it this is an essential distinction for Joannes' interpretation of the sixth century
also offers anecdotal evidence for the permeation of silver money into the coun- s1.lver termmo . logy. 2621 An altematlve. was suggested by Vargyas. Accordmg . to
tryside and for the problems created by its high purchasing power. him, ginnu silver was coined, that is, the term refers to the sigloi minted by the
Ginnu silver (kaspu sa ginni) is perhaps the most controversial and most of- Achaemenids after the monetary reform of Darius I (which then would have to
615
ten discussed type of silver attested in Neo-Babylonian records? It appears be placed at the very beginning of the king's rule); the earlier references to
frequently in the records from the beginning of Darius' reign, and virtually dis- ginnu silver supposedly refer to the originally Lydian Croeseids which were still
616 2622
appears at the beginning of Xerxes' reign? In particular in Sippar texts, it is minted by the Persians. This conclusion, accepted enthusiastically by Pow-
2617 2623 . . .c .
very frequently contrasted with simple "white silver." According to Vargyas, e ll , seems attractive smce reterences to the early (Lyd1an and Achaemenid)
the earliest reference is from the fourth year of Cyrus: GC 2, 101 from the Eanna coinage would otherwise be completely missing in the cuneiform record al-
archive states that ginnu silver may not be melted down by temple silversmiths though at least some coins must have circulated in Babylonia in this period. 2624
on the pain of incurring the "punishment of the king," hftu sa sarri. From However, there are serious problems.
roughly the same period comes CT 22, 40, a private letter from Sippar in which The archaeological record does not fit Vargyas' interpretation very well: not
it is said that as per royal order (amiit sarri), ginnu silver is not to be accepted only are there no coin finds in sixth-century contexts from Babylonia - this
(innassi) as payment. 2618 The most common interpretation sees ginnu (perhaps a could perhaps be considered a coincidence2625 -, the quality of the sigloi is
loan from Sumerian gi.na, "regular, certified"2619 ) as a stamp on the silver guar- higher than that of the Babylonian ginnu silver, which is usually of 87.5 % fine-
ness: the coins found in Babylon contain 96.4-98.1 % silver. Vargyas acknowl-
2615 edges this objection but assumes that sixth century sigloi may have had a lower
See esp. Vargyas 1999; 2001: 2lff. 2626
2616 silver content. The second problem is chronology. If ginnu silver were iden-
The earliest reference post-dating Cambyses actually comes from the period of the
tical with the sigloi, one would expect more references post-dating the reign of
revolt of Nebuchadnezzar IV, Nbk. 12, all other tablets are from 2 Dar or later. One
reference dates to the first year of an Artaxerxes (probably Art I): 113 ma.na ku.babbar Darius (or rather, post-dating the revolts in 2 Xer), even post-dating Xerxes
ina 1 g[in] rgfrl-re-e bab-rbal-nu sa ge-en-nu (BM 79000, Babylon). Otherwise there is altogether, i.e., from the time after the monetary reform - if reform it was - of
to the best of our knowledge no attestation of ginnu silver from after the revolts against Artaxerxes 1: sigloi manifestly continued to be used in Babylonia as shown by
2627
Xerxes. During the reign of Darius however it appears quite frequently, often with the Babylon hoard. However, apart from BM 79000 there are no such late
87.5 % fineness. The ubiquity of silver of this kind prompted scribes sometimes to point
out explicitly the absence of the ginnu characteristic: kaspu sa Iii ginni. This is an exclu-
262
° CAD G assumes that the entire surface of sheets of silver was stamped whereupon
sively 'negative' expression. Even if ginnu silver were identical with the sigloi as sug- these sheets could be cut up as necessary.
2621
gested by Vargyas (see below), there would be no material basis for the assumption of See Joannes 1994: 140. Vargyas 1999: 256f. argues against this assumption, and in
Vargyas that Greek coins hide behind the 'non-ginnu' silver. Note that the supposed fact it is difficult to verify this distinction. The two texts referring to the prohibition to
"owls" (ginnu and non-ginnu silver displaying a characteristic designated as ha-al is-sur smelt ginnu silver or to use it for certain types of exchanges have no bearing on this
and hal-la is-sur respectively, Vargyas 1999: 263 with further references) are actually issue, we believe (see below), and smelting of silver in the temples produced silver
most likely very small pieces of silver resembling "bird droppings" - this is the literal which was used for manufacturing objects as well as for the purposes of exchange; the
meaning of the expression, see most recently Stol 2000: 628 (reference courtesy of M. terminology used is the same (see note 2635 below).
2622 • 65
Weszeli). For the Croeseids see Vargyas 1999: 259+ .
2617 2623
See, e.g., ABC 33, 38, 39. The widespread use of ginnu silver in this period is par- Powell 1999: 2lf.
2624
ticularly well illustrated by the tablet Dortmond no. 32 (7 Dar, courtesy of A.C.V.M. See, e.g., Powell 1996: 234 (and 1999: 22: "Like markets, coinage was there before
Bongenaar; the tablet will be published by C. Waerzeggers), in which a man is suspected the Greeks. Neither are mentioned [in the written record, MJ], because it was not neces-
of having stolen silver implements from the temple because he had been caught with sary to state the obvious"). The earliest coins found in Babylonia date to "about the mid-
eighteen shekels of white silver in his possession and "you do not have an income of dle of the first quarter of the fourth century," see Moorey 1994: 238.
white silver: how can you try to give/sell this silver for ginnu silver" (erebka sa kaspi M~ II . . .
Powe pomts out that Mesopotamia was not explored as extensively and intensive-
pesf yiinu mfnii kaspu anna ana kaspi sa ginni tanandin). If the "white silver" in question ly as many areas of the Mediterranean (Powelll996: 235).
resembled the offcuts from bowls and other silver vessels that were found in the Babylon 2626
Vargyas 2001: 30.
hoard (Reade 1986: Plate III) it is understandable that the temple was suspicious. ginnu 2627
Joannes 1994: 144 connects the disappearance of ginnu silver in the fifth century
silver on the other hand could not have been mistaken for fragments of stolen silverware.
2618 with the increased use of silver of greater purity, especially of Greek origin, and the
Joannes 1994: 140, Vargyas 1999: 256, Jursa 2006a (for CT 22, 40). disappearance of the distinction in fineness between silver used for silversmithing and
2619
See Powell 1996: 233. silver used for exchange. But as pointed out above, the existence of this distinction is
482 5. Silver, silver money and money-based exchange 5.2. Silver qualities 483

2632
references to ginnu silver. It virtually disappears after 2 Xer, even though the of Addaru (XII)."
628
silver terminology remains otherwise unchanged during this king's reign? Sumu-iddin was certainly not content with the silver he had been sent: he
There is an even graver problem regarding the early attestations of ginnu sil- asks for more money. As can be shown on the basis of some more or less con-
ver. Pace Vargyas the earliest dated reference to the use of ginnu silver in a temporary texts, tuhhfi. silver was at least sometimes considered unacceptable as
. ' .
commercial context only comes from the s1xth year of Cambyses.
2629
However, a means of exchange and was therefore refused when offered. 2633 This and the
the oldest reference to ginnu silver has not been recognised as such yet: it is
found in the undated letter YOS 3, 153. This text from the Eanna archive can be
2632
YOS 3, 153: 16-27: 113 ma.na ku.babbar ina su 11 tctin-nin-numun-til ren1 <ki-i> u-se-
assigned with near certainty to the period between 3.13 Nbn and 4.17 Nbn on hi-la ku.babbar it-tan-ni igi li•a.kin sa en-ia ki-i ap-te-e-su 3 gin ku.babbar ina lib-bi tu-
prosopographic grounds. 2630 According to Vargyas' theory we could only be uh" -hu u re-eh-ti gi-in-ni ku.babbar sa suk.hi.a a-na '"erinmc sa ki lugal d+en id-din-ni
dealing with Lydian coins here- but it does not seem very likely that such coins suk.hi.a sa iti.se at-tan-na-as-su-nu-tu. The scribe uses both ct+ en and en for "(my) lord."
The preterite id-din-ni has to be taken as a wish since it stands in a main clause (Streck
would have been in circulation in southern Babylonia a few years after the fall
1995: 127ff.).
of Lydia to the Persians, but before the Persian conquest of Babylonia. 2633
The interpretation of this word is contentious. CAD T: "(qualifying silver)," AHw.:
The passage in question will be cited in full, as we need to know whether it
"dargebracht" (on the basis of etymology only), similarly Zawadzki 1985: llOf. on Nbn.
refers to the actual use of ginnu silver as a means of exchange (something which 119 and Joannes 1982: 240 on YOS 6, 115. Vargyas (2001: 40f.) on the other hand sug-
according to CT 22, 40 may have been explicitly forbidden by royal decree a gests on the basis of Nbn. 119 that the word be translated "refined." (The text is a settle-
few years later). Sumu-iddin, who is anxious about his ability to provide for the ment of accounts with temple smiths regarding a large quantity of silver they had had to
workers under his command, 2631 writes as follows (YOS 3, 153: 16-27): refine.) Vargyas' understanding of tuhhu cannot be correct, as the text, as expected, uses
"<When> (my) lord sent one-third of a mina of silver via Innin-zeru-usabsi, tubbu for "refine." As the summary treatment of the text in Bongenaar 1997: 362f. also
he delivered the silver to me. When I opened (the bag) in the presence of my shows, the tuhhu silver was not reckoned with the refined silver proper. It was returned
lord's messenger, three shekels of silver therein turned out to be tuhhfi., and the to the treasury, either without having undergone the smelting process at all, or as a sub-
rest was ginnu( -silver). Let (my) lord pay silver for the rations of the men who standard result of it. Only once is tuhhu silver definitely used as a means of exchange,
do service with the king. I have already given to them the rations for the month viz. in the earliest reference, TCL 12, 46 (32 Nbk). The silver is described as tu-hu-u sa
'"ku.tim ul-tu pit-qu u-ter-ra "tuhhu silver which the goldsmith had given back from/after
the smelting" - this does not necessarily imply that the silver in question had been re-
uncertain for the sixth century, too (note 2621). Joannes' conclusion is endorsed "with fined, the phrasing rather suggests the opposite. In all later attestations tuhhu silver is
some modifications" by Vargyas 1999: 255, but even if"the standard of value was raised either refused, or at least returned to the temple which had issued it without being used:
from 87.5 per cent to fine silver" (ibid.) under Artaxerxes I, the distinction between PTS 2741 (1 AM, see below note 2635), CT 57, 125 (ace Nbn; the interpretation given in
coined and uncoined silver would have remained as meaningful as before and would CAD Tis wrong; lines 5-6 concern silver issued for blankets, line 7-8 [x g]in tu-hu-u it-
therefore be expected to be reflected in the sources in the same way. Therefore the dis- <te>-hi-is [!u is occasionally attested in nB Texts, e.g., in VS 4, 39, 4 Nbn]), YOS 6,
appearance of ginnu in the record remains unexplained in Vargyas' theory. 115 (Joannes 1982: 240; 7 Nbn), and most importantly YOS 3, 68. This letter is to be
2628
Mostly one hears of "white silver" and of white silver which was also bitqu and/or dated to the reign of Cyrus or Cambyses on the basis of prosopographic evidence (the
nuhhutu silver. temple clerk Nabil-karabu-isme appears, i.a., in the easily datable letters YOS 3, 77, 126
2629
BM 26754 II 26680 (6 Camb, Borsippa); kaspu sa ginni sa nadiini u mahiiri, debt and 173). The passage in question was misconstrued by the CAD, so it shall be cited here
note. Vargyas' dating (1999: 257 and 263; 2001: 22) of the loan contract "S. 138 (Bertin in full. In the context of narrating the events of a trip through the Uruk countryside where
796)" = BM 30416 ("21.5.4 Cyr"), in which ginnu silver is used, is incorrect: the text they purchased agricultural products, the senders state: "when we gave (offered) silver
must be dated to the fourth year of Darius on prosopographical grounds. The main prota- for willow wood, we were told: 'it is tuhhu.' Let (our) lord send us white silver so that
gonist, Marduk-na~ir-apli of the Egibi family, is not active before l Dar (see Wunsch we may pay it for the willow wood" (ku.babbar a-na gishi-le-pi ki-i ni-id-din um-ma tu-
2000a: I 13). Collation by J. Hackl in fact shows that the king's name has to be read uh-hu ku.babbar pe-$U-u en lu-se-bi-la-an-na-si-ma a-na hi-le!-pi ni-id-din). It follows
1
da*-ra-i[a-(a- )musT from the second sentence that the first transaction mentioned by the senders was abortive
2630
The addressee Ill-remanni (a comparatively rare name) is attested as commissioner in that the sellers of the wood refused to cooperate on account of having been offered
(bel-piqitti) of Eanna from 3.14 Nbn-12.16 Nbn; the last reference to his predecessor is tuhhU-silver instead of (the customary?) white silver.
from 3.13 Nbn, the first reference to his successor from 4.17 Nbn (see Kiimmel 1979: While it seems clear that tuhhu silver was not normally used for transactions at least
144 and Kleber 2008: 36). In BIN 1, 71, another letter by the sender of YOS 3, 153, from the reign of Nabonidus onwards, it is not obvious what constituted its limited desi-
Sumu-iddin, to Ill-remanni, a Nabonidus date (9.15 Nbn) is actually mentioned. YOS 3, rability. Several hypotheses can be taken into consideration. tuhhu silver is once con-
190 is another letter belonging to this dossier. trasted with white silver, so perhaps it might be oxidised, 'ugly' silver - then however
2631
He works on a building project at some distance from Uruk. This is also the back- one wonders how to interpret the references to tuhhu silver 'returned' or refused from the
ground ofYOS 3, 190 and BIN 1, 71; see also BIN 1, 3 (Sumu-iddin to the satammu, the refining process. Possibly it is the inferior silver slag resulting from the refining process
letter mentions the fifteenth year ofNabonidus; it refers to the same affair as BIN 1, 71). which is attested as hfltum in Old Assyrian (see Joannes 1993-7: 103; he cites a Neo-
 5. Silver, silver money and money-based exchange 5.2. Silver qualities 485
484

context make it appear possible than there was also some restriction to the us- fact that the intended transaction was an entirely private venture while the com-
ability of ginnu silver in this period, but the passage cannot be considered a paratively rare and 'safe' ginnu silver may at this time have been earmarked for
genuine proof of this assumption. In fact, the whole issue of the supposed re- special purposes - for instance for payments in the institutional economy or,
striction, or ban, on the use of ginnu silver is in need of reconsideration.Z
634 more likely, for payments to the crown. 2639
The ban against smelting ginnu silver in GC 2, 101 was taken as an indica- Summing up, it is virtually certain that ginnu refers to some distinctive mark
tion that ginnu silver was not of sufficient quality to be used for the manufacture on, or form of, silver. The mark was intended to convey information about the
of silver implements by the silversmiths of Eanna - but after refining this would silver's purity and/or the institution guaranteeing this purity. At least when it
no longer have been a problem. Vargyas on the other hand assumes that the first came into use, it probably also carried the connotation of being intended for
ginnu coins had to be protected because of their rarity and their ideological im- certain types of transaction only. This connotation is no longer apparent during
port. This is certainly a plausible explanation (if one accepts the premise that the reign of Darius, when ginnu silver clearly circulated freely and was perhaps
ginnu silver was coined silver). However, silver refined (or perhaps better, sim- even considered in general a more 'natural' form of silver money than "white
ply smelted) by the temple smiths was also expended for commercial pur- silver"2640 - which does not preclude it from having been the desired means of
poses,2635 and the prohibition to smelt ginnu silver may simply have been in- payment for particular purposes. The principal arguments against the intriguing
tended to safeguard against unnecessary losses during the refining process, theory that we are dealing with first the Croeseids and later Darius' sigloi are a)
given the fact that there was no reason to doubt or improve its fineness: this was the fact that ginnu silver appears earlier than hitherto recognized, during the
2636 reign of Nabonidus, and b) its disappearance after the revolts against Xerxes,
after all certified by the very fact of its being ginnu silver. As for CT 22, 40:
the tablet states that no ginnu silver was to be accepted as payment for beer in a even though the silver terminology did not change otherwise during the reign of
private transaction; instead, murruqu silver ("cleaned silver"), perhaps simply a this king and even the replacement of the various terms for fineness by qalu
2637 under Artaxerxes I should have not affected the need explicitly to refer to coins
synonym of"white silver," kaspu pe$U, was to be used. Vargyas assumes that
.
at Ieast occas10na IIy. 2641 The chronology of the use of gmnu
. shows that we are
the king's decree protected the coins because of their rarity- which was to be-
come unnecessary under Darius? 638 But what kind of protection would have dealing with a practice for guaranteeing the quality of silver (intended for the
been intended here? Should ginnu silver simply have been hoarded by institu- royal treasury?) which was introduced by the Chaldean monarchy and which
tions? It seems more likely that the problematic point at issue in the letter is the was at first continued by the Achaemenids and may even have been strongly
encouraged under Darius. It possibly depended on the agency of the local tem-
Babylonian reference to (silver from) isdi agarinni as a possible parallel): this would ples2642 and seems to have been more or less abolished like many other local
explain in particular Nbn. 119 very well. Finally, one could suggest on the basis of the traditions after the upheavals during the beginning of the reign of Xerxes.
literal meaning of the word ("that which has been brought close") that tuhhu silver might Whether we are dealing with stamped silver or perhaps even silver in some kind
be silver whose purity had not been established by controlled refining and alloying in a of distinct shape, viz. some form oflocal 'coinage,' cannot be ascertained.
temple foundry, but by using a touchstone (i.e., by rubbing the metal on the stone and The probable link between silver qualities in general and ginnu silver in par-
comparing the traces with those made by silver of a known fineness). Currently none of ticular on the one hand and the royal administration (and its tax demands) on the
these hypotheses can be verified. other is underscored by the references to another silver qualification which has
2634 See note 2618 above for references to the pertinent tablets and the most recent
hitherto received little attention. Unpublished texts have more than tripled the
secondary literature. handful of attestations of the so-called erbu sa aranni, literally "the income of
2635 The terminology referring to the treatment (petequ) of such silver is the same as
the cashbox," that were known so far. The arannu cashbox or treasury was first
that used when silver is intended for silversmithing. See, e.g., from Sippar CT 55, 275, investigated by Oppenheim (1947). In his discussion of the quppu cashboxes in
287, CT 56,454, CT 22, 131, Nbn. 749. See also PTS 2741 (Eanna archive, I AM): 8.5
temples he also makes a reference to Dar. 484 (19 Dar), in which a payment a-
shekels of silver are issued from the money chest (quppu) for smelting (ana pit-qu). One-
third of a shekel is returned as tuhhu (it-te-eh-si), five shekels are given to PN for the
2639 Th'IS hypothesis . IS
. not af"'.ected by the questiOn
. of whether or not ginnu silver is
purchase of twenty minas of iron, and 3.2 shekels are said to have been lost in the smelt-
ing (ina pit-qu in-da-tu). 0.33+5+3.2 = 8.53. coined silver. See von Reden 2007: 29ff. and 84ff. for the close nexus between money,
2636 The crown had a close interest in the economic fortunes of the temples; this is fre- coin issues and tax demands in Ptolemaic Egypt. The rationale for the royal 'sponsor-
quently the background of threats of meting out the "punishment of the king," hltu sa ship' of ginnu silver must have been very similar.
264o Th. . dby the Dortmond text, see note 2617.
IS IS suggeste
sarri, as in the case ofGC 2, 101. See Kleber 2008: 68ff. 2641
2637 Note a new attestation of murruqu silver in BM 31278, an Egibi archive letter that Add to this the discrepancy regarding the purity of ginnu silver and that of the sigloi
can be dated to the reign of Darius on prosopogrpahic grounds: ku.babbar bab-ba-nu-u recovered from Babylon and the uncertainties regarding the range of circulation of the
[m]u-ru-qu [i]n_r na1as-su (we owe this reference to J. Hackl). sigloi and in particular also of the Croeseids.
2642
2638
Vargyas 2001: 16. This is plausible, but cannot be proven.
486 5. Silver, silver money and money-based exchange 5.2. Silver qualities 487

ki-i a-ra-nu sa iti.apin "according to the (rate of exchange of the) arannu in Unfortunately neither can the tablet be attributed with certainty to an archive
month VIII" is envisaged. 2643 This singular reference can now be supplemented nor the protagonists be identified. Prima vista one would be inclined to see this
by one sixth-century text, BM 79128, and two tablets from the late fifth century text as proof of the existence of an arannu treasury within the realm of the tem-
which were published by M. Stolper. In these tablets, the feminine plural of ple (Ezida). However, the temple archives of Sippar and Uruk, where the ex-
arannu IS · use d: payments are to be rnade ak-1 (gis) a-ra-na-a-twta , 1 2644 ( sa
" pression erbu sa aranni appears in several private contracts, do not use the word
2646
2645
biibi!i ) or libbu a-ra-na-tu4 sa arahsamni ... All documents are of a pri- arannu; the temples' cashboxes (including those royal cashboxes that were kept
vate nature, arannu is only invoked for the purposes of determining the ex- in the temple) are called quppu. 2652 It is probable that the same is true for Ezida,
change rate app Iymg. tothe transactwns.-
. 1647 but as the Ezida temple archive has not been recovered there cannot be certainty
Basing his argument primarily on Dar. 484 (and his understanding of Nbn. on this point. In any case, on the basis of the evidence from the Eanna and
807) and on structural parallels from other regions (and periods) of the Near Ebabbar archives, we assume that the payment ana Nabu is a payment "for/on
East, Oppenheim takes arannu 2648 as the cashbox of the fiscus, whose officials behalf of Nabu," rather than "to Nabu" (in which case the arannu treasury
acted also as "assayers," as they regularly had to determine the silver value of would belong to Ezida). If this is correct, we are dealing here with a state treas-
payments made in kind. He adds "it must be assumed that the ever fluctuating ury (which receives a payment made by a representative of Ezida). It is sugges-
silver value of these commodities had to be established periodically by some tive that a goldsmith and an blacksmith, i.e., specialists for the treatment of pre-
central office. This of course, entailed some sort of publication of the rates of cious (and base) metals, are in charge of the treasury: these are Oppenheim's
exchange at regular intervals."2649 (According to Oppenheim, this postulated "assayers." It would have been the task of these men to effect the transactions in
"list of quotations" hides behind the word nishu, which CAD N/2: 267 therefore money and in kind "according to (the exchange-rate of) the treasury" which is
translates as "tariff' in the relevant contexts: see below p. 492). In essence Op- invoked in the three documents cited above. Furthermore, these smiths at the
penheim's argument still stands. It can now be supported by some additional very least evaluated, and perhaps also interfered with, the quality of the silver
evidence, particularly by the references to the "income of the cashbox," an ex- that came into the treasury and that was later brought into circulation again. This
pression which always refers to a given amount of silver. In the table given be- follows from the references to silver designated as erbu sa aranni.
650
low, all known attestations of silver which is thus designated are collected?
The only tablet explicitly attesting a transaction involving the arannu is from Silver qualifications Text Amount Date Origin/Archive
Borsippa. The text, an administrative note, reads as follows: "(Regarding) 3.5 erbu sa aranni BM 94930 24 s [.. ] Camb Bors. (unknown)
shekels of silver, which Bel-ke[sir] paid for (or: to; ana) Nabft at the cashbox on sa ginni erbu sa aran- BM 103583 2653 10 s 1 Dar Dilbat (unknown)
orders of Nabu-sumu-u~ur, son of Bania, on 8.10.acc Camb. In the presence of ni OECT 10, 137 1m 5 Dar (Kis; unknown)
Sin-iddin, the goldsmith, Aplaya, the smith, who is (who are) responsible for the GC 2, 131 1/3 m 9 Dar Uruk (Haba~iru2654 )
2651
cashbox, and Musezib-Nabu, the carpenter." AUWE5, 134 1/2 m 10 Dar Uruk (Haba~iru)
sa ina 1 siqli bitqu sa BM 82662 2655 35 s 5 Dar Bors. (Iddin-P. A)
2643 ginni erbu sa aranni BM 26655 2656 6.75 s 6 [Dar] Bors. (Atkuppu)
The supposed parallel he cites, viz. Nbn. 807: 12 (Oppenheim 1947: ll8), has to be
read akimahiri sa til-gubbi, see Wunsch l995/l996: 45f.
2644
BM 79128 (19 Dar), Jursa 1999: 251f. kur.kur (5) ina na-as-par-tu4 sa ldag-mu-uru (6) a-su sa 1ba-ni-ia (7) ina rugu gisla-ra-an-
2645
YBC 11607 (8 Dar II); Stolper l990b: 165f. "according to the exchange rate pre- nu (8) a-na dag id-di-nu (9) ina du sa ld30-mu (IQ) luku.dim 1a-a lusimug (11) sa ina UgU
vailing at Babylon."
gis
a-ra-an-m. ( 12) u Ikar- dag Iunagar.
2652
2646
NBC 8394; Stolper 2007: 258 no. 14: "at the rate of exchange prevailing in month The difference between arannu and quppu ("vergelijkbaar maar niet synoniem") is
VIII ... " emphasised by Waerzeggers in her discussion of erbu sa aranni in two tablets of the
2647
Even though his understanding was based (in part) on reading aki rnahiri sa quppi Sahit-gine A archive, Waerzeggers 2000/2001: I 102. She also connects arannu with the
in Nbn. 807 (see above note 2643), Oppenheim is surely right in taking aki aranni etc. as staatskas.
2653
a synecdochic abbreviation of *aki rnahiri sa aranni; arannularanniitu is not a simple 10 gin k[u.babbar] sa gin-nu er-bi, sa a-ra-an-ni sa Se.bar u zu.lum.ma; in line 8 the
synonym to rnahiru. silver is only referred to as sa ginni.
On the archive of Haba~lru, son of Balatu, see van Driel 1998b: 66 and Frahm and
2654
2648
And mistakenly quppu.
2649
Oppenheim 1947: ll9. Jursa (in press).
2655
2650
In a brief reference to erbu sa aranni, Vargyas 2001: 31 also cites ABC 36 ("Ste- In this tablet from the Iddin-Papsukkal A archive, sa ina iSten siqli bitqu is missing
venson 1992, 36") as a relevant text. This is erroneous. and sa nadiini u rnahiiri precedes erbu sa aranni. In line 8f., the silver is only designated
2651
BM 26784; transliteration by C. Waerzeggers: (1) 3 1/2 gin ku.babbar sa lden-ke- as sa ginni sa nadiini u rnahiiri.
2656
[sir] (2) ud.8.kam sa iti.ab mu.sag (3) nam.lugal.la 1karn-bu-zi-ia (4) lugal tin.tili lugal Here, sa ina iSten siqli bitqu is missing.
 5. Silver, silver money and money-based exchange 5.2. Silver qualities 489
488

Silver qualifications Text Amount Date Origin/ Archive the actual origin of the silver and thus naming the real creditor (viz. the 'treas-
sa nadtmi u mahiiri BM 26622 [.. ] 10 Dar Bors. (Re\-alpi) ury'), it always would have had to be placed at the very end of the description of
TEBR69 lm 13 Dar Bors. (uncertain) the silver characteristics, before the first mention of the (in this case merely
BM 94872
2657 21? s 13 Dar Bors. (Re'i-alpi) nominal human) creditor and of the debtor. Therefore erbu sa aranni is part of
BM 29716
2658
16 s 14+ Dar Bors. (Belia'u) the silver description. 2664 It implies that the silver in question had at some point
BM 82730
2659
1/3 m 20Dar Bors. (Re'i-alpi) - not necessarily when the transaction in question was undertaken - been in the
BM 95861
2660
178 16 Dar Borsippa (Belia'u) possession of the 'treasury,' 2665 that it was still recognisable as such, and that
§a nadiini u mahiiri
this fact had an importance for the evaluation of the silver qualified in this way.
erbu sa aranni
2661
5m 2 Dar Sip. (Sahit gine A) Silver designated as erbu sa aranni can also be: ginnu silver, silver of 87.5%
qallalu §a ginni erbu BM 63848
5m 2 Dar Sip. (Sahit gine A) fineness, silver "of the kind commonly used for transactions" (sa nadiini u
§a aranni BM 79714
1/2 m ace Camb Bors. (unknown) mahiiri), 2666 and it can come in small pieces (if this is what qallalu means 2667 ). If
qallalu erbu §a BM 29445
gis -2662 erbu sa aranni in RA 97, 101 (BM 54732) refers to the silver described else-
arannz
where in the text as supplementary payment (which is likely, but cannot be
[ ... ] erbu §a aranni
2663
RA 97, 101 10 s (Xer) Cutha (Sangu-IStar-
Babili) proven), we have here an attestation of "white silver" which is probably of
(BM 54732)
87.5% fineness. Furthermore, chronology is of relevance: during the reign of
Table 54: silver designated as "income of the treasury chest," erbu sa aranm
Darius, erbu sa aranni silver is generally ginnu silver (the only exception is the
poorly written text BM 95861 (16 Dar), where this information may simply be
With the exception of the latest text, RA 97, 101 (BM 54732), a land sale
missing), before and after this period, this combination of silver characteristics
contract, and BM 63848 (formation of a business company), all tablets listed
is unattested.
here are debt notes. They come from private archives and concern private trans-
As suggested by Joannes for TEBR 69, the mention of the fact that silver had
actions. Apart from the mention of the erbu sa aranni there is no indication that
been "income of the 'treasury'" must have constituted a guarantee of the value
institutional interests might be concerned. In fact, the list shows that erbu sa · question.
. 2668 It IS
of th e s1.1ver m . however not obvwus . what exactly was guaran-
aranni cannot imply that the 'treasury' is directly involved. The phrase appears
teed and how this guarantee was conveyed. Most likely it was a guarantee of
sometimes within the list of silver qualifications and is followed by another such
silver quality, as the incoming silver of different types and degrees of fineness
expression, such as sa nadiini u mahiiri "(silver) of the kind commonly used for
must certainly have been tested and may have been smelted and alloyed in a
transactions." If the inclusion of the erbu sa aranni phrase were meant to convey
standardised way by the specialised treasury personnel. But the specifics cannot
information of essential legal importance for the transactions as such by stating
be determined with certainty. If ginnu silver was really coins, arannu would be
the local mint, and erbu sa aranni would designate (locally minted) Achaemenid
2657
§a ginni §a nadiini u mahiiri ereb aranni. coins: but the premise is doubtful, as we have pointed out. 2669 If ginnu silver is
2658
Here, §a nadiini u mahiiri precedes erbu §a aranni.
2659
Here, sa ina iSten siqli bitqu is missing.
2660 This is an unusually badly written tablet. The silver in question is first described as
2664
This was seen by Joannes 1982: 275 and Vargyas 2001: 23+ 124, in their respective
17 gin ku.babbar sa na-da-nu u ma-ha-AN (not ri) er-bi sa a-ra-nu; the payment clause discussions ofTEBR 69.
2665
runs as follows: ud.4.udsic sa iti.gan ku.babbar a4 17 gin a-na na-da-nu u ma-ha-ri er-bi It is self-evident that not all taxes collected by the Achaemenid administration in
§a a-ra-nu i-nam-din. ana is undoubtedly a mistake for §a. Babylonia were hoarded; a large part must have been spent (locally, or in the province)
2661 BM 63848 and BM 79714 are edited in Waerzeggers 2000/2001 as no. 18 and 19. for building works, maintenance of the irrigation system, wages of royal personnel, etc.
2666
2662 This is followed by r§a mas-ku1 . The enigmatic "of leather" could perhaps be un- Vargyas 200 1: 21 ff.
2667 .
derstood as a reference to a leather bag in which the silver was kept (i.e., as a synonym Cf. Vargyas 2001: 38f. We understand the passage m YOS 3, 68 as a complaint by
of hindu). See note 2670. the senders that they had to purchase barley for the temple using some of their "small
2663 This is mentioned in a damaged addition to a sale contract which is not part of the change" (which they carried for their expenses while travelling through the Uruk coun-
normal formulary. In the main text, the silver price of the house is described as pe$11 sa tryside) rather than being able to rely entirely on the "white silver" supplied by Eanna for
ina i§ten siqli bitqu nuhhutu and perhaps sibirtu; mention is also made of [x] shekels of this important purpose. CAD Q suggests "of inferior quality" for qallalu. This seems less
"white silver" which are payable as a supplement. Most likely the addition states that the likely, also in light of our references here.
2668
supplement - ten shekels, then - was payable in erbu sa aranni silver. (Since the text Joannes 1992: 275: " ... une garantie supplementaire de valeur de !'argent."
2669
mentions different silver types rather than just qa!U silver, it has to be dated to the reign Simply argumenti causa, if one takes ginnu silver as silver whose fineness was
of Xerxes although it belongs to a dossier terminating in the reign of Artaxerxes I [see guaranteed by an institution in some way, e.g., by a stamp, erbu §a aranni could still
Vargyas 2000: 515].) mean a coin, as just a more specific form of institutional guarantee of value would have
490 5. Silver, silver money and money-based exchange 5.3. Interest rates 491

either stamped silver or silver cast in a certain fonn by, e.g., the temples, the social convention (or administrative decree). 2673 Yet, this long-term continuity
'treasury' might have added a stamp of its own or it might have packed the sil- notwithstanding, there is both a considerable amount of short-tenn fluctuation as
ver in sealed leather bags?670 Alternatively ginnu silver which is also erbu sa well as a distinct trend from lower interest in the seventh century to higher, even
aranni might simply have been the ginnu variety issued by the 'treasury,' in much higher rates, from the late fifth century onwards.
contrast to other types of ginnu st'1 ver. 2671 These uncertamttes
. . notwtt. hstand'mg, In the table given at the end of this section, we have collected the different
the fact remains that the texts on arannu and erbu sa aranni silver constitute the interest rates for silver that appear in the Neo-Babylonian documentation. 2674
most explicit corpus of evidence from Babylonia bearing on the specifics of the Both the variety and the range of these rates may seem large, but there is a clear
active interference of the Achaemenid administration with the silver circulating pattern to the documentation (which is perhaps somewhat obscured by the nec-
in Babylonia. essarily selective citations). As was pointed out above, by far the most frequent
rate is twenty percent p.a.; it is attested throughout the period under discussion,
5.3. Interest rates well into the Seleucid era. It dominates the documentation from the sixth cen-
In the context of the discussion of silver money, interest rates deserve attention tury, but in the seventh it is less frequent than the various lower rates listed in
as they may reflect diachronic changes in money supply, money use and even the table taken together. In the sixth century, rates below twenty percent must be
social attitudes towards money. 2672 The Neo-Babylonian evidence in general is considered exceptional; occasionally they can be explained (see below). Rates
fully embedded in the Mesopotamian continuum in that interest rates are overall higher than twenty percent occur at best sporadically in the seventh century2675
fairly stable. The age-old rate of one shekel per mina per month (20 % per an- and are also rare in the sixth century. From the second half of the fifth century
num) is by far the most frequent rate - a rate whose use as a standard is primar- onwards, interest rates are far less frequently mentioned in general, but enough
ily owed, not to economic factors sensu strictu, but to metrology and thus to material has been recovered to show that rates more often than not exceeded
twenty percent per annum. This not only applies to (potentially punitive) interest
rates demanded for delayed payments/676 but also for 'normal' interest rates:
thirty to forty percent p.a. may have been the norm in this period. 2677
been implied. Compared to the assumption that ginnu designated coins, this hypothesis
Obvious exceptions to a clear general trend such as low interest rates during
would have the advantage of a slightly more convincing chronological distribution of the
evidence (which starts later and ends later), but the Cambyses attestations would still be
the later sixth century are best explained as the result of negotiation between the
troubling. In any case other explanations are also possible. parties, that is, as a result of individual circumstances which are referred to in
2670
This is perhaps suggested by the sa maski "of leather" qualification of BM 29445
2673
(see note 2662). Money bags, and the information that can be conveyed by their seals See, e.g., Hudson 2002: 23f., M. Hudson 2000.
2674
and tags, can be used as a kind of deus ex machina for explaining how information about Most of these rates can already be found in Petschow 1956: 2043 •; see also Wunsch
silver, especially uncoined silver in pieces, might have been conveyed. It is certain that 2002: 234ff. and (for the late documentation), Cardascia 1951: 5, Jursa 2006: 161 93 •
such bags were used far more frequently than the comparatively few explicit attestations References are selective, naturally so for the ubiquitous rate of twenty percent, but also
of hindu and klsu "money bag, purse" suggest. YOS 3, 153 (discussed above) is a good for the more common lower rates of the seventh century, less so for the lower rates from
example: lines 16-19 contain a routine reference to silver sent by a superior as it is found the time after the reign ofNebuchadnezzar and for the higher rates of the late period. We
dozens of times in texts of this type. Only because there was a problem with the silver do have always cited the earliest and latest attestations known to us for each interest rate.
2675
we realise it had been sent in a bag: "when I opened it...," kl aptesu. In the case of the The only exception being debt notes from the Sln-uballit archive from Or/Babylon.
erbu sa aranni silver one notes that more than two-thirds of the attestations involve neat The high interest rates in several of these texts must be owed to the peculiar background
round sums of multiples of a mina or of 'round' fractions of a mina (10, 20, 30 shekels). of these transactions (and the archive in general): as C. Waerzeggers has suggested, this
Such round sums (including also 15 and 40 shekels) also account for the majority of part of the archive probably deals with taxation and military conscription (see Jursa
attestations of ginnu silver (Vargyas 2001: 26ff.), but the margin by which- they out- 2005a: 136+ 1058).
2676
number the odd sums is smaller (fifty-six to forty-four percent). The number of attesta- Pertinent attestations are marked in the table with an asterisk after the year date.
tions of erbu sa aranni silver is however too small for this difference to be of interpreta- 2677
This was first observed by Cardascia 1951: 5; see also van Oriel 1987a: 160 (al-
tive value. On sealed money bags see most recently Vargyas 2005 and 2004: 116f. though his observation that "such stipulations usually [emphasis added, MJ] refer to a
2671
It can be excluded that ginnu silver was issued exclusively by the arannu institu- penalty clause" is inaccurate). The counter-arguments presented by San Nicolo 1954:
tion: the texts show clearly that we are dealing with a specific variety of silver. 280f. cannot be maintained: interest rates exceeding 20 % are very rare before the fifth
2672
A good example of the influence of the money supply on interest in pre-modern century; San Nicolo's examples all come from the exceptional context of the Sln-uballit
societies is offered by the 'commercial revolution' of thirteenth-century Europe, when an archive from Ur. Note also the high interest rates for barley attested in Larsa in the fourth
increasing availability of money for (commercial) loans caused interest rates for such century: 40 % p.a. given as 3.33 % p.m., with even higher interest in penalty clauses
loans to drop from around 20 % to sometimes well under 10 % in the course of the thir- (10% p.m.): Joannes 2001: 257 23 • Interest of 50 % p.a. is stipulated for a debt of dates
teenth century: Spufford 1988: 260ff. after a missed deadline for payment in BE 9, 68 (39 Art 1).