The Arch of Trajan at Beneventum

This section treats the passage reliefs of the Arch of Trajan at Beneventum, which depict respectively an alimenta distribution by the emperor and some kind of inaugural imperial sacrifice (figs. 91–92). The panel that truly parallels the cups is the alimenta panel.^[2]

The arch can be dated from the terms of the dedicatory inscription;^[3] it was dedicated in A.D. 114, between 19 August and 9 December. W. Gauer observed that the earliest date for the arch's inception is immediately after Trajan's Dacian triumph of 107,^[4] which is portrayed in the "small frieze" just under the attic story; he suggested, sensibly, that the arch's decorative program might well depend from the theology of imperial worth and legitimacy worked out for Trajan's anniversary celebration of decennalia in A.D. 108. Thus the arch was begun between ca. A.D. 108 and ca. A.D. 110/111 and finally dedicated in A.D. 114.

The arch marks one end of the Via Traiana, the new stretch of road laid by Trajan to extend the old Via Appia to Brindisi to connect that great port directly with Rome and west-central Italy. Brindisi was the traditional port of embarkation for Western legions moving to Greece and Asia; the road itself was made ready by A.D. 109 to transport Trajan's troops to war with Armenia and Parthia. The choice of Beneventum as a site for this arch, voted to Trajan by the Senate, clearly points to a connection between its erection and the dedication of the great new highway. Thus the sacrifice in the arch passage (fig. 92) is often identified with the dedicatory sacrifice of the Via Traiana. Certainly it shows neither a profectio from Rome (in the same narrow sense as the nuncupatio depictions discussed in chapter 5) nor the conventional sacrifice ending a triumph.^[5] Most likely it is a sacrifice at Beneventum itself to open the Via Traiana, in connection with Trajans profectio to Parthia; or it may be connected with Trajan's decennalia celebrations.^[6]

An obvious point of correspondence between these passage reliefs and the BR cups occurs in this very sacrifice or profectio panel: the use of the BR victim group (pl. 9), the "Pausias motif," for the victim slaying.^[7] However, this is not the point on which I will build my case; the way in

which this victim group is incorporated into the composition depends on Domitianic workshop stylization. The evidence is worth presenting, as this particular point will be of use later on (see chapter 5 for individual monuments cited).

I refer to the coin (fig. 90) issued for Domitian's ludi saeculares ; it is likely itself to depend on a monumental painting or relief. The points of resemblance between the arch passage panel and the Domitianic coin type are: the group now consists only of the axe-swinging popa and the victimarius who kneels at left to twist the bull's head down, while only the forepart of the bull projects from the massed figures immediately right of the victim group; these in turn consist of two tall figures oriented toward the right and overlapping very slightly (the left behind the right), and from between them, as if flattened to paper-thinness, projects the bull; finally, the group is set at left of the main imperial sacrifice scene, so that its lines of force diverge instead of converging on the sacrificial altar; the group as a whole is reduced in size relative to the other figures, and the bull itself is now very small in relation to its slayers, and so even smaller compared to the altar group celebrants.

The onset of this process of "insertion" is observable on the Temple of Mars Ultor relief (fig. 9a). Here, however, the victim group is still to the right, and part of a unified and centralized composition; the bull slayers are still three in number and of the same size as the other figures in the relief, and the victim is still, realistically, of massive size. The Beneventum relief, on the other hand, goes even farther than the Domitianic coin in treating the victim group as a kind of pictograph whose physical "realism" is a matter of indifference—the left of the two figures who cut off the bull actually stands with his forearm loosely resting on the animal's back (at his waist level) as if the bull were an inert lump of balustrade and not a heaving animal maddened by fear under violent restraint!

The victim group on the arch, then, follows a line of adaptation of the BR victim group that passes through a Domitianic reworking documented by the coinage. This dependence on Domitianic workshop tradition is not surprising; mere common sense predicates a good deal of continuity between the workshop staffs employed first by Domitian and then, after the single of year of Nerva's reign, by Nerva's successor, Trajan. Such continuity has already been established for the purely architectural decoration of the arch, linked now to the Domitianic Arch of Titus.^[8] The autonomy of this line of transmission for the victim group is confirmed by the contrast with the so-called Uffizi relief (fig. 94). This

Trajanic or Hadrianic relief returned to a fuller and more "classical" rendering, which is much closer to the line of adaptation seen on the Julio-Claudian Temple of Mars Ultor relief (fig. 9a).

This brings us to the real focus of this section. Given the pictographic condensation of the BR victim group on the arch's sacrifice passage relief, it is especially noteworthy that the facing panel in the arch passage does borrow from the BR sacrifice panel composition, directly and in a far more sophisticated fashion. This other passage panel, the alimenta relief (fig. 91),^[9] borrows as well from the "son-giving" composition of BR I:2. In both cases, the borrowing seems to be the product of a search for formal inspiration, that is, for guidance in the composition of massed figure groups, rather than an effort to evoke the kind of historical events celebrated on the cup panels and their original prototype.

The alimenta passage relief divides formally into two halves, left and right; I discuss the left section first. This portion of the relief is dominated by a triad of paludate figures (heads now missing), whose placement can be compared to that of the figures of a carousel as it revolves before a spectator. Each figure steps forward with one leg trailing on a circular path, first coming toward and then receding from the spectator: at left fully frontal; in the center in a higher, "nearer" plane of relief in three-quarter profile; at right in lower relief, again seen from behind and moving back into the relief ground. Each figure extends one or both arms forward to a greater or lesser degree, thus enhancing the sense of forward motion along the circular path established by the grouping of the triad. They circle toward and past a low tripod table on which stand two vaguely triangular lumps, which represent food or money to be handed out as alimenta to the children to the right of the table. The right-hand triad figure, moving past the table, still has an arm out over it and presumably has just handed out largess to the child now moving away with loaded cloak at right center; the central figure, who can only be Trajan, is moving toward the table in his turn, awaited by the child who right of the table holds out his cloak to pocket the expected dole.

The rite of distribution, then, seems to have been introduced by the right-hand triad personage, plausibly identified as the curator viarum; not only would he have been responsible for the construction of the Via Traiana, but by virtue of his office he was also in charge of the distribution of alimenta^[10] (sensible, as he controlled the road network on which this distribution depended). The emperor is about to take over the presiding office, awaited by the expectant child and the equally expectant figures of city personifications behind and to the right of the table.

Note, finally, the distinctive shape of the tripod table (the front support of which and its braces are now missing): its legs are solid and straight, flat extensions, rectangular in section; they support a very wide, round flat-bottomed tray with high, straight sides. This tray table is very different in form and proportions from the tripod altar on the facing passage panel, which is the table's compositional pendant as one passes through the arch.

Turn to the left-hand portion of the sacrifice panel BR II:1. Here is the same "carousel" of three paludate figures, in this case a lictor and two officers. Corresponding to the right-hand Beneventum triad figure is a lictor three-quarters front, right arm similarly flexed at the elbow and brought slightly forward, left arm hidden under his cloak balancing the fasces on the left shoulder; corresponding to the Beneventum Trajan is a figure in high relief in the nearest foreground plane, in three-quarter profile, left arm bent and brought forward and grasping something small, the right arm brought forward from the shoulder and bent sharply up; corresponding to the Beneventum curator is a third figure, in lower relief, seen from behind with his head turned, so that we see it in profile, his cloak similarly draped from a point on each shoulder so that catenary arcs fall down his back in succession between a cluster of vertical folds on either side. These three figures circle just behind Tiberius, immediately before whom stands a tripod altar of exactly the same shape and proportions as the Beneventum tripod table.

This "carousel" group serves to frame the action and lead the viewer into the narrative, at the same time enhancing the viewer's perception of deep "real" space. It is a clearly demarcated and highly sophisticated compositional group that functions as a dynamic visual device—what can be called a compositional unit. I have not seen this particular device used elsewhere in Roman relief (one is hampered by the fact that figure groupings and precise spatial devices are seldom discussed); note the identity of structure and placement, with the basic iconographic elements that the triad is paludate near the shared element of the distinctive tray table. In his rendition, the arch designer has used "carousel" and altar conflated, omitting the sacrificant Tiberius, who on the cup intervenes between these two compositional elements; retaining the placement and gestures of the BR figures, he has made the central triad figure in the foreground in high relief serve as protagonist of the panel and has put an enveloping fringed cloak on the back right triad member, who seems here, as on the cup, to be a bearer of something, probably a standard or vexillum. Also, the "carousel" has been rotated back a notch, so that the right-hand figure is

buried less deeply in the relief ground and is on the same plane as the bearer at left, while the central figure is more fully displayed in three-quarter view. This kind of quotation and adaptation of compositional units occurs in Greek battle friezes, Roman sacro-idyllic landscape paintings, and Late Antique/Early Byzantine manuscript illustrations, among other examples; the reuse of such units, which serve both an iconographic and a formal function, is the hallmark in antiquity of a continuous artistic tradition in a given medium.

Now for the right-hand portion of the arch panel. At farthest right a man moves away with a (female) child on his shoulders whose arm curves down to lead away an older (male) child on foot;^[11] beyond the older child a slightly shorter child comes up to the central tripod table; this child holds his arms up and looks upward at the emperor/attendant, who is making a gesture of reception and bestowal. The descending curve formed by the linked arms of the father and his son, highlighted by the curving gesture of the girl child on his back, is continued in the next plane of relief back by the curved body of a child in the arms of the leftmost female personification, and by the outstretched arms of this female and the personification immediately next to her. The father and two children on foot form, as it were, the front plane of highest relief, as shown below.

Behind them is a second plane constituted by the massed curtain of female figures, three personifications with turreted crowns who stand frontally with their heads turned toward the center; beyond this basic group are, behind the altar table from right to left, another father with his child on his shoulders (the father of the child at the table, presumably) and a larger female personification (probably Beneventum herself). This second plane is illustrated below.

The group as a whole consists of fathers with children who move in procession in stages to the center, taking their turns, to supplicate a benefit

that state personages benevolently aquiesce in extending, meeting the supplicants with an answering gesture.

Significantly, this group, at least up to the altar table and the third female personification from the right, is a mirror image of the left-hand portion of the hostage scene BR I:2. There, a group of fathers usher their children forward in turn, two children on foot, framed at far left by a father with his child on his shoulders and in their movement highlighted by the stretching out of arms forming a descending arc. This descending arc is met by the answering curve of the gesture with which Augustus acquiesces in the children's supplication for protection and sponsorship. This curving sequence of foreground figures, highlighted by curving accents in the relief plane behind, is set against a backwall of massed figures of Romans and barbarians.

The Roman (Drusus) physically succors the children, as do the Beneventum personifications (the cities from which the children come). Above. (p. 107) I showed how the BR "flap" composition turned up on the Ara Pacis, used in either case to frame the young general Drusus set in its midst (fig. 78). At Beneventum (fig. 91) it occurs in its own right, facing, not enclosing, a standing imperator and benefactor. It is plain, too, that the Beneventum panel derives not from the Ara Pacis version but from the BR version of this compositional structure, in figure types, narrative context, and formal location (framing one end of a centralized panel narrative composition). This is what occurred with the left-hand section of the Beneventum panel and BR II:1 (carousel group + table unit); a triad of paludate figures was involved in an imperial offering from a special table (to the gods on the cup, to men on the arch). In the unit described here, young children are brought by their fathers as loyal subjects/citizens to ask for and receive imperial sponsorship in the form of the extension of a father's nurturing role by the emperor (guardianship on the cup, material nourishment on the arch).

Compositional analysis thus establishes that the alimenta passage relief composition was constructed by combining two compositional units that each make up roughly one-half of two separate panels from the Boscoreale Cups. These formal compositional correspondences are indisputable; how is one to comment on them? It is true that very little panel relief survives from the years between the reigns of Augustus and Trajan; it is possible that these compositional units came down to the Trajanic artist mediated by separate lines of transmission, to converge in the arch by accident. However, it seems to me too much of a coincidence that the two

units quoted should stem from a single assemblage and be reused in a single assemblage and that in each case not only formal but contextual reference has been maintained to a significant degree. The arch panel is based on the observation of a single assemblage containing the originals of the two cup panels BR I:2 and II:1.

This still leaves the possibility that the arch designer was looking at an earlier imperial relief that did his work of quotation and combination for him, what I have termed a mediating work. We do have for comparison on the facing sacrifice panel a reuse of a BR cup compositional unit, which can be proved to have been transmitted by intermediate stages of reuse. That unit is the victim group discussed above. There it can be seen how the handling of what was once a formally rich and sophisticated figure group had turned flat and stale, after the group had passed through several stages of mediation. Evidently, the Beneventum master was quite willing to turn to the later and simpler reworking of a given compositional unit, if such a version was available, rather than attempt a more difficult presentation (contrast especially the work of the designer of the Uffizi relief, fig. 94, discussed on p. 158). The alimenta panel, too, when compared with the BR renderings echoed, manifests this tendency toward stylization and dimensional simplification on the part of the Beneventum master. Given this observable tendency, the remaining close correspondence between the alimenta panel groupings and the BR groupings (Gallic fatherson group, and altar group) seems to show that the Beneventum master worked here at only one remove from the BR panels' originals.

To conclude, the designer of the reliefs of the Arch of Trajan at Beneventum (the Beneventum master) belonged to a workshop from the capital known to have carried out other commissions in Rome.^[12] In drafting the composition for the alimenta passage relief, he made notes for himself to carry to Beneventum, based on direct observation of the BR prototype reliefs. In doing so he marked down useful ideas on how to make a stimulating panel rendering of particular kinds of crowd scenes dictated by his commission at Beneventum: "paludate imperial party approaches offering table"; "children supplicate the emperor for his fatherly patronage." For these considerations of structure and content, he reworked for his own commission what seemed to him a useful guiding prototype, the pair of reliefs copied for BR I:2 and II:1. One is tempted to think that the Beneventum master thought that one useful element of his prototypes was that they were Augustan reliefs, that is, that some of his audience were meant to pick up on the Augustan reference as such; this kind of imitatio would fit nicely with Trajan's well-known tendency to portray himself as

a "new Augustus" by evoking Augustan political themes and imitating Augustan monuments (consider the Forum of Trajan vis-à-vis the Forum of Augustus). While it is possible that a reference to specific Augustan monuments was supposed to be legible in the alimenta panel, it cannot be proved, and it is rather unlikely given that the arch at Beneventum had to be viewed at such physical distance from its visual sources, which were located in Rome.

A few closing words on style are in order. The discussion of compositional units here should enlarge the reader's notion of what the lost Augustan monumental corpus was worth in artistic terms. As those who have waded through the literature on Roman relief know, the so-called achievements of Flavian illusionism have dominated discussions of the evolution of Roman relief style, especially as a dating criterion for unattached fragments, ever since F. Wickhoff waxed enthusiastic over the Arch of Titus panel reliefs and associated monuments, to the point of comparing the master of the Arch of Titus (figs. 107–8) to Velázquez. He was particularly struck by Flavian mastery in the manipulation of relief planes "die Wirklichkeit vorzutauschen,"^[13] as other scholars have been ever since. We, like Wickhoff, cannot but respond to this Flavian mastery of the expressive possibilities of relief, which we can see beginning to crystallize into a more stylized deployment in much of the Trajanic corpus. What the BR cups show is that some at least of the most sophisticated visual devices available to the master carvers of the late first and early second centuries of the Empire had already been initiated and formulated at the beginning of the Empire. By sheer accident most of what Julio-Claudian relief is left to us is in the form of procession friezes, which exercise in decorous conformance to their genre a much more subdued style than the narrative historical panels that first surface on the Arch of Titus. The BR narrative panels show us that Augustan response to the demands of this genre can stand comparison with the weightiest of imperial compositions from the so-called Golden Age of Flavian and Trajanic narrative relief.