Die Marks and the Organization of the Roman Mint: An RRDP Case Study
by Alice Sharpless and Lucia Carbone
The purpose of control marks in Roman Republican coinage is not well understood. Beginning as early as 112 BCE, the Roman mint began experimenting with control marks. Sometimes these control marks were unique to the die and sometimes they were not. Control mark systems employed a very wide use of numbers, letters, and symbols. So far, the issues released to RRDP include 40 with control marks. Much has been written about the potential function of such marks. Back in 1987, A. Burnett suggested that die marks “were apparently [originally] intended to keep track of the production of the coinage and the dies used to strike it (pp. 22–23).” Building upon previous studies, most notably those by C. Hersch and T. Buttrey on Crepusius, R. Witschonke argued for a direct connection between the use of die marks and the control of the monetary production in the Roman mint, suggesting that a precise amount of bullion was allocated for a specific issue and that the coins produced from that bullion were accounted for with die marks. Yet, the fact that control marks are never used across all issues at any given time suggests that if control marks were used to track bullion or dies, the mint never felt the need to do so consistently.
The present data release for RRDP, the fourth since the beginning of Phase II, offers further insight on this important question. This release involved the analysis of 6,297 specimens and added 1234 new dies to the RRDP database for the following Crawford types:
|335/1a 335/1b 335/1c 335/2 335/3a 335/3b 335/3c 335/3d 335/3f||335/3g 335/9 335/10a 335/10b 336/1a 336/1b 336/1c 382/1b 383/1|
The recent analysis of RRC 335/3 in particular has offered new insights into the function of “subtypes”—smaller units within an issue often distinguished by different symbols or control marks—both with regards to their use by modern scholars as well as to their purpose within the Roman mint. Crawford identified seven subtypes of 335/3, but there are no known specimens of 335/3e. Crawford’s description of the type may be a misinterpretation of a small group of unusual dies that likely do not form a particular subtype at all. Even more interesting, however, is the evidence this issue provides to our understanding of the organization of the Roman mint. One subtype (335/3c) has four dies that were recut to add a die mark. The presence of these recut dies seems to confirm previously suggested theories that die marks were used primarily as ways to control production volume.
RRC 335/3 is a joint issue of L. Caecilius Metellus and C. Publicius Malleolus that was struck in 96 BCE. Crawford divided 335/3 into seven subtypes (a–g) distinguished by their reverse designs. The basic reverse design shows a naked warrior standing left, holding a spear in his right hand and placing his right foot on cuirass with a trophy on the left. The design of the right field varies but always includes the legend C·MAL. Crawford identified five subtypes (a–e) which have a prow below the legend (Fig. 1), while c–e are further distinguished by symbols above the prow: 335/3c has a caduceus, 335/3d has a grasshopper (by far the largest subtype of the issue), and 335/3e has a tablet.
According to Crawford, 335/3c should have a caduceus between the legend and prow, but only 10 dies precisely fit this description (Fig. 2).
There are an additional 5 dies, which currently can best be associated with 335/3c but do not clearly have a caduceus at all. Three of these dies (Schaefer 335/3c Reverse F, Schaefer 335/3c Reverse H, and Sharpless 335/3c Reverse 1002) do not have anything between the prow and legend (Fig. 3).
Rather, they have marks located between the warrior’s back leg and the legend. The marks could possibly be intended to be the top of a caduceus represented without a handle, but they look more like letters. A fourth die (Schaefer 335/3c Reverse J) has both a caduceus located between the legend and prow and the marks between the warrior and legend (Fig. 4).
On this die, the marks look fairly clearly to be the letters CO. On Reverse F or Reverse H the letters could be CP (Fig. 3). They could be interpreted as standing for C(aius) P(ublicius), though this seems somewhat unlikely given the presence of C(aius) in the main legend. Regardless, the presence of the caduceus in addition to these marks suggests that the marks are not meant to depict the head of a caduceus and should, therefore, not be used to distinguish the typology for a particular coin.
But two dies complicate this picture. Sharpless 335/3c Reverse 1002, which is known only from one specimen, has marks only between the warrior and the legend, but in this case the marks look most plausibly like the head of a caduceus (pointing left) and can not easily be distinguished as letters (Fig. 3). Even more intriguing is Schaefer 335/3c Reverse I, also known from only one specimen (Fig. 5).
Reverse I has not one but two caducei, but neither are located between the legend and prow. One caduceus is instead placed upright between the warrior’s legs. The other has the head located between the warrior’s back leg and the legend, while the handle, which is angled downwards, sits between the prow and legend. It appears, in fact, that the second caduceus was not originally made as a caduceus at all. It seems instead that the die originally had marks (letters?) like those on Schaefer 335/3c Reverse J (Fig. 4) but the engraver added a handle to turn these marks into a caduceus, resulting in the unusual placement and the awkward angle of the handle.
Crawford also identified another subtype with the prow: 335/3e, which he describes as having a tablet between the prow and legend. Crawford cited a specimen of this type from the Bellicello hoard but we have found no other specimens. In a 1957 publication of the Bellicello hoard, Cutroni Tusa lists one coin from this issue in the hoard (no. 18) but describes the reverse as having a “cavalletta” (locust) above the prow, in which case the coin is actually 335/3d (Fig. 6).
It is possible that when Crawford described 335/3e he was actually thinking of one of the specimens with the marks between the warrior and legend that could potentially be interpreted as an open tablet, in a smaller version of the tablet that appears on types 335/3f-g. This interpretation seems possible for Schaefer 335/3c Reverse F or Sharpless 335/3c Reverse 1002 (Fig. 3), but seems unlikely for the other dies. We must conclude, therefore, that 335/3e, at least as Crawford describes it, is a “phantom” type that does not really exist.
Even more relevant to the question of shedding light on the function of “subtypes” is the presence of four dies of 335/3c that were clearly recut from dies that had already been in use minting 335/3b. It seems at some point a die engraver simply added a caduceus between the legend and prow on an older die. The following dies were recut:
- Schaefer RRC 335/3c Reverse A recut from Schaefer RRC 335/3b Reverse C
- Schaefer RRC 335/3c Reverse B recut from Sharpless RRC 335/3b Reverse 1007
- Schaefer RRC 335/3c Reverse E recut from Schaefer RRC 335/3b Reverse K
- Sharpless RRC 335/3c Reverse 1000 (Fig. 7) recut from Schaefer RRC 335/3b Reverse B (Fig. 8)
Dies are sometimes recut for use in a new year or for use by a new moneyer. Examples of dies recut to fit a different civic year, month, or moneyer are quite rare, but not unknown in the Greek world, as made clear by Figures 9 and 10.
In Figure 9, the reverse die of this cistophorus from Tralles, probably dated to 145–140 BCE, has the letters AY (Audnaios, the Macedonian month of December/January) cut over AΠ, clear reference to month of Apellaios, which immediately precedes Audnaios in the Macedonian calendar. In this case, the die, which was not exhausted, was used for a different month, since the Lydian city of Tralles briefly adopted the month indication on its cistophoric coinage. Figure 10 shows a New Style Athenian tetradrachm— originally dated by M. Thompson (pp. 232–237) to 147–146 BCE, but likely produced a few years later—whose reverse die has been recut to write ΔΗ/ MOΣ/ ΘE over MIΛ/ TIA/ ΔHΣ, ME. The recutting of the reverse die allowed for the substitution of the more generic ΔΗMOΣ to MIΛTIAΔHΣ, the moneyer who had signed the rest of the issue (Fig.11).
For reasons unknown to us, MIΛTIAΔHΣ could not sign the die anymore, so the city had to hastily step in while the production was in progress. Recutting a die was thus a way to prolong the life of a die under different issuers or in different years.
In the case of RRC 335/3c, however, we have dies that were still in usable condition being recut for use by the same moneyer. The addition of the caduceus to these four dies confirms Crawford’s proposed chronological sequence of subtypes 335/3b and 335/3c, but they also appear to confirm the suggestions of Burnett and Witschonke that dies marks were used to control production. The fact that these dies were not simply used until exhaustion but were instead recut to form a new subtype within the issue suggests that the production size for particular types, marked in this case by symbols or variations on a base design, was established before striking began. The size of each type was not simply determined by how many coins a die with a certain die mark could strike before it was spent, but by a decision made prior to production, possibly in accordance with a specific decree of the Senate, as suggested, among others, by the presence of S.C. on issues related to emergency funding for Rome’s grain supply in 56 BCE (Carbone and Yarrow 2019, pp. 16–19; Fig. 12).
These production levels could be controlled by the use of changing control marks within a type, but the recut dies of 335/3c show that the division of issues into “subtypes” could serve a similar function. This suggests that the mint could have tracked bullion more consistently than was previously thought based on the number of control marked issues alone.