by Christopher K. Lezak

SFANS, 2001

Though it is generally not understood that Greek workers were employed at the mint of Rome in an artistic capacity in the First Century, a distinctive variation of lettering on the early denarii of Domitian provides die evidence that this was the case. This variation appears as a misshapen figure "V," with the tail of the letter extended to the rim and its arms slightly concave, so that it more closely resembles the letter "Y." (1) It will be shown that this is most likely a phonetic spelling error by a Greek-speaking die engraver. The variation is found on at least two dated issues of late 81 and on an undated denarius of Domitia, which is traditionally assigned to 82–83:

1. IMP CAES DOMITIAN AYG PONT: laureate head right. (2)
    PP COS VII DES VIII: wreath on curule chairs.

2. IMP CAES DIVI VESP F DOMITIAN AYG P M: laureate head left. (3)
   TR P COS VII DES VIII P P: dolphin on tripod.

3. DOMITIA AVGVSTA IMP DOMIT: draped bust right. (4)
    CONCORDIA AVGYST: peacock right.

The variant coins are official issues of the mint of Rome. The cited denarius with the Wreath on Curule Chairs reverse is die-linked to a denarius with a normal legend, showing that the variant denarii are not contemporary counterfeits, nor the products of a branch mint. The occurrence of the variation on coins of different reverse types also shows that it is not the mark of an officina.(5)

It can next be established that the variation was consciously engraved onto each die. On an individual coin the variation resembles a die break, but the coincidence of three examples suggests that the variation was systematic in nature. Because the lettering on these dies was entirely cut by hand, there was no chance for the mechanical reproduction of the variation. The variant letter was therefore a deliberate act of the engraver, as opposed to a failure in striking or an isolated slip of the tool.

We can also assume that all examples of the variation are the work of the same hand.(6) First, the variation does not occur elsewhere on Imperial coinage of the First Century, and the three known examples are contemporary. Second, this sort of mistaken depiction of lettering is not the type of variation that might be copied from one engraver to the next. An engraver of figures would meticulously copy the details of the model, so, for example, Minerva would not be depicted with a cornucopia unless the model itself was defective. In that case it would be likely that others without a good knowledge of deities might reproduce the same error, but with the legends, it is critical to distinguish between the idea and the physical shape of the text. It would have been easier for the engraver to remember short phrases and engrave them as such, rather than refer to the model for each figure. A Greek-speaking engraver might thus mentally substitute the Greek figure in place of its Latin equivalent in his haste.

The shape and positioning of the variation suggest that we see the Greek figure upsilon in place of the Latin "u".(7) In its capital form, as used in the legends, "u" is written as "Y," which approximates the shape of the distorted "V." In each of the cited examples, the variant letter occurs in vowel positions. "Y" is substituted on two occasions for "u" in the diphthong "au" of AVG, and once for "u" of AVGVST. No distortion is made of the consonant "V," which would not be spoken as "u". This systematic deformity of the vowel is a phonetic spelling error, indicating that the engraver was more familiar with the Greek alphabet than the Latin. Therefore, the legends of these particular dies may be assigned to a Greek or provincial worker, employed at the mint of Rome under Domitian.

It does not necessarily follow that the portraits of these coins are also attributable to this cutter of legends. On later denarii of Domitian, the legends are sometimes poorly spaced, resulting in the crowding of the final letters into the point of the bust. This suggests that the legends were generally engraved after the portraits. Though it seems that a master engraver might be reluctant to have an amateur engrave legends upon a finished portrait, the massive amount of coinage produced suggests that efficiency would have been the prime concern of mint directors. Allowing master engravers to concentrate on portraits by delegating repetitive tasks to others would be efficient. It is therefore likely that different caelators cut the portraits and legends, so that the work of one engraver of legends is not necessarily correlative of the work of an engraver of portraits.

The identification of this caelator does help to show the validity of the DIVI VESP F and DOMITIAN legend variations. DIVI VESP F dies other than the cited example do not bear the upsilon variation. Roman standards of quality were generally high, so it is probable that this was corrected quickly. If the cutter was corrected for this anomaly, it is certain that the rare DIVI VESP F and DOMITIAN legends would have also been corrected, had they been errors. Furthermore, the DIVI VESP F legend was standard on the aes of 81 and presumably adopted concurrently on the denarii(8), though it was probably found unsuitable for their small flans. Die evidence also confirms a date early in the third issue,(9) in agreement with the idea that this was a short-lived experimental type. At this point, towards the end of the second issue and just before the massive third issue of 81, it may have been necessary to take on extra labor, hence our upsilon engraver. Therefore, the DIVI VESP F legend variant was a legitimate form of the third issue legend, and not the invention of a single engraver.

The CONCORDIA AVGVST issue of Domitia is traditionally dated to 82-83, but this does not fit the proposed scheme. If the engraver of the upsilon letter variation was employed with any regularity and was not corrected until 82 or 83, we should see many examples of the variation in the abundant third issue of 81. This is not the case, so a date contemporary to the other two upsilon varieties, at the transition of the second and third issues of 81 is expected. Association with the other undated issues,(10) some of which are die linked to the dated second issue of 82,(11) argues against this earlier date, but the variation ought to be given some weight because it provides the only die evidence available to precisely date this type.(12) Unless the engraver lapsed into his old habit a year later, it must be concluded that this die was at least prepared in 81, and therefore contemplated at that time, even if striking was not accomplished until the next year when the group of family issues is generally thought to have been struck.

The significance of a slightly earlier date is that Domitian uncharacteristically chose to honor his wife before his own well-being.(13) Domitian also retained the pulvinar(14) types of Titus on the early denarii, and did not devote the bulk of his coinage to Minerva, his patron goddess, until late 82. This continuation of the reverse types struck under Titus may be seen simply as an initial lack of attention to the mint by Domitian. However, if types of Domitia were also introduced in 81, then Domitian was already influencing the coinage and the preservation of the pulvinar types may have been his conscious choice. It is possible then, that some issues of Domitia belong to a more dutiful period for Domitian, immediately following his accession as Augustus.(15)

The identification of the engraver of the upsilon variation has implications for the workings of the mint in general. First, the nature of the variation as a Greek phonetic error suggests that Greek engravers were employed at this time by the Roman mint, which also precludes the use of a die signature by an artist of portraits. Second, this particular engraver is not the originator of the DIVI VESP F type, supporting its status as an official legend. Third, it presents evidence that the CONCORDIA AVGVST type of Domitia ought to be attributed to the second or early third issue of 81, rather than the traditional date of 82-83. This in turn suggests an earlier date for at least some of the family issues, which, in conjunction with the pulvinar issues, appears to be an initial period of piety for Domitian, or perhaps propaganda with that intended effect.

I would like to thank Mr. Ben Damsky for guiding my research, Ms. Rhys Cheung for editing the text, Mr. Curtis Clay for answering my initial questions, and Mr. John Jencek for photographing the coins.

(1) Flavian denarii in general often show slight distortions of the lettering. However, the three cited examples are set apart by the extreme length of the tail on the figure and the apparent deliberateness of execution. Other examples of blundered lettering certainly exist. See the New York Sale IV, lot 360, 2002, for a denarius of Titus with obverse legend beginning IMP TITVS. However, it is not yet clear whether such coins are of an Eastern mint, or whether they present evidence of a more widespread employment of Greek workers at the mint of Rome.

(2) C 58, a variant of RIC 7 Domitian. The obverse legend substitutes DOMITIAN for the usual DOMITIANVS. This obverse is die-linked to a denarius with the normal obverse legend IMP CAES DOMITIANVS AVG PONT.

(3) Variant of RIC 21 Domitian. The obverse legend adds DIVI VESP F and substitutes DOMITIAN for the usual DOMITIANVS. RIC II, p. 156 note. There is one obverse die of this type, paired with two reverse dies. Milan; Stockholm; Vienna; Lanz 72, lot 562, 1995; NAC Sale E, lot 2899, 1995; others in trade.

(4) RIC 212 Domitian, BMC 61. This particular die appeared in Triton IV, lot 521, 2000. The Triton catalog notes that this die is shared by an aureus.

(5) The mint was arranged into separate officinae based on reverse types from late 81 to early 82. Carradice, Ian. Coinage and Finances in the Reign of Domitian, A.D. 81-96. BAR International Series No. 178, 1983, p. 146. The system of officinae was abandoned in mid 82. Id. It is possible that the system of officinae existed under Titus, then passing to Domitian. A denarius of Domitian with the unique obverse legend IMP CAES DIVI VESP F DOMIT AVG PONT (Carradice, p. 13) and the dolphin and anchor reverse has a very unusual die axis of 12:00. This axis also occurs on two other denarii of the same reverse type: RIC 26 Titus, in trade; and RIC 20 Domitian, Leu 61, lot 251, 1995 (the catalog date of 82 is in error), but does not appear among other reverse types. Perhaps this officina had the tendency to produce coins with the 12:00 axis and was established prior to January 1st–July 1st of 80, the time of RIC 26 Titus.

(6) The practice of Greek caelators might suggest that the letter variation is the die signature of the portrait artist. This is clearly not the case. Die signatures are common on Greek coins, but were not used on Roman issues. Furthermore, the dies in question are not medallions that might merit an exception to the rule. The head left type is unusual for Domitian, but the group taken as a whole is better characterized as an assortment of regular issue denarii and aurei.

(7) A conclusion suggested (twice) by Ben Damsky.

(8) There are no aes with legends that correspond to the first and second issues of 81. As defined by RIC II for the coinage of 81, obverse legends ending in AVG are of the first issue, legends ending in PONT are of the second, and legends ending in P M are of the third. The corresponding reverse legends are TR P COS VII, COS VII DES VIII P P, and TR P COS VII DES VIII P P, respectively.

(9) First, the legend appears primarily in the third issue, but one example of the second issue is also known (IMP CAES DIVI VESP F DOMIT AVG PONT). Carradice, p. 13. Second, there is a series of die links with a reverse die, TR P COS VII DES VIII P P (type not specified), paired with obverses reading IMP CAES DOMITIANVS AVG PONT, IMP CAES DIVI VESP F DOMITIAN AVG P M, and IMP CAES DOMITIANVS AVG P M. Carradice, p. 53 (note 15). The placement of the DIVI VESP F variants at the transition of the second and third issues minimizes the time span required to strike in all three marriages, thus explaining the apparent longevity of this reverse die.

(10) Domitian issued a small group of coins honoring members of his family. Included are: Domitia, RIC 210 Domitian, RIC 212-14 Domitian; Titus and Iulia Titi, RIC 216 Domitian; Vespasian and Domitilla, RIC 69 Titus; and Domitilla, RIC 71 Titus. The two coins cataloged in RIC II under Titus have been reattributed to Domitian by Carradice, p. 20.

(11) An aureus, RIC 40 Domitian, IVPPITER CONSERVATOR, is die-linked to a dated denarius of the second issue of 82, and a denarius, RIC 41 Domitian, SALVS AVGVST, is also die-linked to a dated denarius of the second issue of 82. Carradice, p. 19. Association with these two types appears to be the principal reason for dating the whole of the family issues to 82-83. Content-wise, these types can be distinguished from the types honoring Domitian's family, and it is argued that this association should be severed in favor of new die evidence.

(12) The legends contain no TR P or COS information. A link between the CONCORDIA AVGVST type of Domitia and a dated reverse die of Domitian may be possible through the hybrid denarius BMC 244 Domitian. However, such a series does not necessarily exist and involves a hybrid, which is by definition coined under unusual circumstances.

(13) The IVPPITER CONSERVATOR type refers to Domitian's escape from the burning of the Capitol at the hands of the Vitellians, years earlier in 69. SALVS AVGVST is a more general statement of the health and safety of the emperor, but it may refer to the same event. RIC II, p. 151.

(14) These are the familiar types showing pulvinar: a wreath on curule chairs, an ornamented throne, a thunderbolt on a draped table, and a crested Corinthian helmet on a draped table. The group and also includes a lighted altar, a tripod with a dolphin above, and a dolphin twined around an anchor. The references of the first two are unclear, but the latter represent Jupiter, Minerva, Vesta, Apollo, and Neptune, respectively. In The Throne and Curule Chair Types of Titus and Domitian, Ben Damsky argues that the traditionally accepted meaning of these pulvinar types, that of supplication to the gods, is incorrect. Noting that the span of the issue is too long for a response to natural disaster, and that supplication is too negative a meaning for Roman propaganda, Damsky instead proposes that the pulvinar represented posthumous honors to Vespasian on the occasion of the completion of the Coliseum.

(15) Domitian did not produce dynastic issues like those that his father Vespasian struck showing his living children. Domitian's own son, who is honored on an issue of Domitia, RIC 213 and 209A Domitian, probably died in 83 (Suetonius gives the date as the "second year after he became emperor," Domitian III, Loeb, trans. Rolfe). This type of Domitia was certainly not struck before the boy's death, so Domitia's coinage seems to span the period of late 81-83. If contemporary, the family and pulvinar series, struck very early in the reign, were most likely an attempt to enhance the image of the always ill-favored brother of Titus.