Beryl Barr-Sharrar is an acknowledged expert on ancient bronzes, whose published work has highlighted Greek and Macedonian metal vessels (1982, 1999), bronze lamps and decorative busts from the Mahdia shipwreck (1994), and—in a splendid book—the spectacular Derveni Krater (2008). Most recently, she gave a paper at the 19th International Congress on Ancient Bronzes held in California (October 13-17, 2015) in conjunction with the Exhibition "Power and Pathos", which allowed her a close look at some of the exhibits. I, on the other hand, have yet to see that important show; I only reviewed the hefty Catalogue that accompanied it, which prompted Barr-Sharrar’s response.
I could therefore take the easy way out and claim that I was simply paraphrasing the editors’ comments on the points to which Barr-Sharrar objected: the "replication" and "reproducibility of bronzes through casting" because of their technique (Essay I, p. 33 and passim), and the subdivision of selected items in the final display section, V, entitled "Apoxyomenos and the art of replication." But this response would be inadequate, especially since I personally subscribe to the theory of seriality in antiquity, as I hope to demonstrate by citing some of my own publications through the years.
In 1985, The Center for Advanced Study in the Visual Arts (CASVA) and the National Gallery sponsored a Symposium: "Retaining the Original. Multiple Originals, Copies, and Reproductions." I was asked to address the Greek period1 and mentioned several examples of duplication in architecture and sculpture (e.g., stone waterspouts, "Kleobis and Biton," Telemachos Relief) but did not include major bronzes because of the lack of known replicas at the time. I was already intrigued, however, by the evidence of statue bases and of ancient sources mentioning commemorative monuments that comprised great numbers of figures. My concerns on the subject were not about iconography but stylistic homogeneity—how to explain the juxtaposition of works by different masters who came from different places and schools to collaborate on a (supposedly) unified presentation.2
The first example to catch my attention was the Granikos Monument. Erected at Dion, in Macedonia, it represented 25 mounted companions of Alexander who died at the Battle of the River Granikos in 334 B.C.E. It was allegedly made by Lysippos and some modern commentators suggested that the sculptor must have followed the conqueror in his Asian campaigns since the statues were said to have portrait-like quality. I objected to this last theory in view of the fact that Lysippos (to judge from datable signed bases) would have been quite old at that time. Moreover, a large group of that nature would have required a well-established foundry, hardly to be encountered by an army in motion.3 But back "home," with an established school of Lysippos (?) contributing, the feat was at least theoretically possible. The basic model? A horseman with individual head but similar armor, legs cast separately, spread out across his mount, as exemplified by some recent finds from the sea near Kalymnos—two headless, cuirassed torsoes, a right leg and two left legs— illustrated in Power and Pathos (pp. 76-79, figs. 5.3-5.7)
More complex (ca. 40 figures) was the Lysander Monument (also called "of the Nauarchoi"), erected at Delphi to commemorate the Battle of Aigospotamoi (405/4 B.C.E.). The major master was Antiphanes of Argos, but he was helped by Athanodoros and Dameas of Kleitor (Arkadia); Alypos, Patrokles, and Kanachos of Sikyon; Theokosmos of Megara, Pison of Kalaureia; and Teisandros—that is, sculptors belonging to different "schools" (of Polykleitos, of the Athenians Kritios and Pheidias) who would have been heirs to different styles. At Delphi stood also the Arkadian Dedication: nine figures commemorating Theban victories over the Spartans (370/69 B.C.E.), made by Antiphanes of Argos (again), with Daidalos of Sikyon, Pausanias of Apollonia, and Samolas of Arkadia. A semicircular base for the Monument of the Kings of Argos (after the foundation of Messene, 369 B.C.E.) held 10 bronze images.4 Comparable examples from Olympia and other sanctuaries could also be cited.
Two possible explanations can be advanced for this rapid production in multiples: either the conception of personal style(s) is a purely modern construct, or the various collaborators—simultaneously—made use of a basic model and created minor variations on it to identify specific individuals, while retaining a fairly uniform appearance. I cannot refrain, in this connection, from citing an eloquent example of a "forest of statues": the 42 life-size bronze figures of the historical American Founding Fathers who, on September 17, 1787, signed the Declaration of Independence in the Assembly Room of the Pennsylvania State House, now called Independence Hall. They have been created by artists at the Studio EIS in Brooklyn and now stand in "Signers’ Hall" of the National Constitution Center in Philadelphia. Although numerous historical sources, including portraits and written descriptions, were consulted in producing the individual images,5 it is obvious that (few) basic models were used to produce the bronzes: notice especially the three types of footwear which repeat consistently among the images, and several other details of the attire. Piece casting indeed favors such procedure.
On May 5-August 24, 2015, the Fondazione Prada, in Milan, Italy, held the exhibition "Serial Classic. Multiplying Art in Greece and Rome" curated by Salvatore Settis and Anna Anguissola.6 Although "seriality was an essential characteristic" of Roman marble copies, as demonstrated by many of the same types (e.g., the Diskobolos by Myron, Cat. nos. 2-8; the "Pouring Satyr" by Praxiteles, nos. 25-32; the "Pothos" by Skopas, nos. 29-30) assembled for comparative purposes, the show explored also Classical Greek art, its function, and "its highly conventional serial language." The Greek examples included the frequently copied Karyatids of the Erechtheion, the famous Fifth-century Penelope statue from Persepolis and its later replicas (nos. 52-58; obviously derived from a second "original" that had remained in Greece), and, at the very entrance to the display, a series of large (lifesize) terracotta busts from Medma (South Italy; nos. 19-24) dated between 500-400 B.C.E. These were "produced in molds" and made "after the same original type" although "varying in hairstyle, decoration, narrative detail…and, in ancient times, their color" now lost). Note that such details were of parts that were applied separately and easily replaced individually, creating an apparent variety that disguised a basic identity.
This eclectic procedure is widely accepted for terracottas, since we generally consider them relatively cheap objects of mass production. Yet bronzes are equally dependent on molds, and piece-casting from a single initial model could easily allow for variations in the type of head, position of arms and legs, and other additional elements (so-called narrative attributes) to create apparently, but superficially, different statues.
Let me briefly address the specific instances that Barr-Sharrar cites in her Response, regarding some objects exhibited in section V as exemplifying the "Art of Replication." We would all agree with her that molds were taken at all times from opera nobilia, either to produce marble replicas or to cast newer bronzes: we even know this from ancient sources (on the Hermes Agoraios, for instance) and from extant examples, including the very bronze apoxyomenoi exhibited in "Power and Pathos" (cat. nos. 40-42), to which different manufacturing dates have been assigned, as Barr-Sharrar points out. But only chance has preserved for us the exhibited pieces! Moreover, dating of ancient bronzes continues to depend heavily on stylistic details, since seldom enough elements of the core exist for radiocarbon dating and the chemical analyses of the alloys may produce ambiguous results—as two of the essays within Power and Pathos warn us. Given the wide distribution of the extant Scraper replicas, both in bronze and in stone, it is logical to assume that the "original" creation—an anonymous athlete with no individualizing features, caught in a generic action—was produced in multiples by its workshop and sold throughout the ancient world to serve varied purposes. Think of the comparable modern examples of Rodin’s works!
Barr-Sharrar devotes her most specific analysis and argumentation to the two Herms of Dionysos, one of which, signed by the sculptor Boethos, came from the Mahdia shipwreck and the other, at the Getty Museum, has an unknown provenience (Cat. nos. 45-46). Replicas of the type in other media produced while the signed herm was underwater prove that other "originals" circulated in Roman times; one of them may be the very Getty Herm which, on the basis of its alloy, has been attributed to the same workshop that made the Mahdia Herm. Barr-Sharrar, however, dismisses the importance of the traces of cobalt in both bronzes as being a feature recurring in other, unrelated, metal statues. She correctly discounts the possibility that the Getty Herm is a Renaissance work because the Mahdia Herm was unknown at the time, but cannot explain the special technique revealed by its interior, which substantially differs from that of the signed piece. She believes, however, that discrepancies in the arrangement of the Getty headdress mean that the entire piece was produced from molds taken not from a prototypical model—as needed for mechanical reproduction— but from the already completed Mahdia Herm, on which the ribbons had been arranged in wax by the sculptor before the one-piece, direct-casting procedure.7
I cannot argue against such specific technical details; I can only contribute something from direct experience. When I had a chance to study the fragmentary "Lady from the Sea" I could notice that the surface of the interior (the chest section of the preserved front) was greatly simplified from the complex arrangements of garments and folds visible on the exterior. In particular, one major fold (cast hollow) was applied separately and must have been made as an independent piece with its own "lost wax" procedure. Only a cuirass-like original form must have been made and cut into pieces for ease in casting, to be later "dressed" with the appropriate folds. Indeed, such a two-time possibility had been raised about the manufacture of the headdress of both the Mahdia and the Getty Herm, but it was challenged by Barr-Sharrar in view of the above-mentioned discrepancies. Isn’t it possible, however, to have both "master" and "pupil" working virtually side by side (or even days/months apart) on two fillet-less heads, the master producing the more elaborate and skillful arrangement and the pupil imitating him in a less successful manner?8
One final objection concerns the two Archaistic figures, Cat. nos. 47-48. I agree with Barr-Sharrar that they both reproduce a generic kouros type, differentiated by the addition of diverse attributes—in the case of the Pompeian youth, objects that transformed it into a "servant statue." But this very imitation of a style current several centuries previously is a form of replication, whether to produce a forgery ( passing it as an "antique," as I still believe for the Piombino Apollo), or to enhance the concept of "statue" as an inanimate object, or even to impart a sense of religious ambiance to its location. The many Severizing/Polykleitan/Classicizing examples of "pretty boys" suggest that there was a Roman market for such creations well beyond the Archaic/Archaistic imitation. I never saw the two "Apollos" together; I only was given a glimpse of the Pompeian bronze’s back in 1982, when it was at the Istituto del Restauro in Rome, lying face down undergoing examination. At that time, Dr. Fausto Zevi suggested that the modeling of the back recalled that of the Apollo in the Louvre, and I agreed that it had the kind of fleshy/feminine quality I had noticed in the Piombino statue. Profile photographs of both figures enhanced the similarity of the pose, but I did not suggest production from a single model. The Rhodian provenience of the bronze in Paris, however, may help in another instance.
The famous Praying Boy in Berlin arrived in Venice before September 28, 1503, as it was described in a letter to Isabella d’Este by Lorenzo di Pavia, a local artist. It had been excavated outside the city walls of Rhodes and belonged to Fra Andrea de’ Martinis of the Knights of St. John of Jerusalem, whose order had a base on Rhodes. Although virtually complete, the bronze lacked both arms and the front part of one foot. Pietro Bembo, a Venetian cardinal and noted author and humanist, in seeing the statue, donated to its owner a half foot in his collection that had been found in excavations "nel Padovano," that is, the area of Padua. The piece fitted so perfectly that Enea Vico thought it truly belonged to the Boy—yet this is a virtual impossibility, given the distance between the two findspots.9 But another explanation is possible. The statue in Berlin is under lifesize, a scale typical of the adolescent servant figures that embellished rich Hellenistic and Roman villas. Standard forms existed for specific anatomical parts since the fifth century B.C.E., witness the well-known depictions of the Foundry Cup. Workshops in Rhodes produced not only Archaistic statues like the Piombino Apollo, but also works in different styles—e.g., Classicizing figures like the Praying Boy, which might originally have served the same practical function as the lychnouchoi known from North Africa and South Italy. In addition, Pietro Aretino, in praising that statue, remarked that its back was better than the front and appeared feminine rather than masculine—a dichotomy I had noted in the Piombino Apollo.
This evidence of seriality across vast geographic spaces should not surprise in Late Hellenistic and Roman times. A much humbler example is again provided by my own experience. During the summer of 1969 I travelled through sites in North Africa, from Cherchel to Cyrene, to improve my knowledge of classical antiquity. In Algeria I picked up a broken terracotta tube of the kind that was used by Roman construction to provide rib-like frames for vaults, walls, and pavements: cylindrical but smaller at one end so as to fit within a similar piece and create a sequence of hollow pipes (tubi fittili). Roman sites which I visited were littered with such fragments. When in Carthage, therefore, I picked up another such broken tube, expecting it to differ from my previous example. But the two fitted perfectly into each other— splendid examples of standardized manufacture across the Roman Empire.
Am I comparing a cheap construction device probably of mid-Imperial date with the expensive artistic production of Hellenistic bronze workshops? Mutatis mutandis I believe the analogy applies. As the contents of the Mahdia wreck and of many other loaded ships that foundered in the Mediterranean have shown, the insatiable demand for luxury generated by different standards of living and self-expression fed into the already practical Greek artisanal mind and produced the means of fast replication, from inexpensive terracottas to what we would consider artistic masterpieces in expensive media. After all, remember that the Greeks did not conceive of a Muse for sculpture and that they called art techne.
1. "Defining the Issue: the Greek Period," in Studies in the History of Art vol. 20 (1989) 13-26.
2. This interest differed somewhat from that of existing publications on ancient groups, which tried to reconstruct the total compositions from the evidence of extant (empty) bases or echoes in other media: Ch. Ioakimidou, Die Statuenreihen griechischer Poleis und Bünde aus spätarchaischer und klassischer Zeit (1997); H. Bunke, Statuarische Gruppen in der frühen griechischen Kunst (JdI-EH 32, Berlin 2004)
3. Besides the riders, the monument perhaps included also some infantry. For the sake of space, I shall cite my works primarily in abbreviated form; in this case, see my review of a 1989 book by G. Calcani, JRA 4 (1991) 206-209; also my Hellenistic Sculpture I (1990) 139 n. 14.
4. For the evidence (inscriptions and ancient sources), see my Fourth Century Styles (1997) 240-43 and ns. 8-12.
5. Information derived from Google, s.v. "Constitution Center."
6. For access to the accompanying pamphlet, in Italian and English, I am greatly indebted to Dr. Irene B. Romano of Arizona State Museum and Arizona University.
7. Such direct casting would obviously eliminate the existence of an "original" wax model because of the "cire perdue" procedure. Yet, regardless of the time frame, would such a mold taken from the completed and cast Mahdia Herm (obviously before it was shipped on its fateful voyage) indicate its existence in the master’s workshop for future serial reproduction?
8. I was able to study the bronze, Izmir Museum No. 3544, in November 1966 when it was displayed at the Museum of the University of Pennsylvania in Philadelphia. See "The Lady from the Sea: A Greek Bronze in Turkey," AJA 71 (1967) 329-34, pls. 97-100, especially figs. 7-8 for interior views and 12-13 for line drawings specifying the separately cast elements and their re-assembly. Cf. also the abbreviated version in Expedition 10.1 (Fall 1967) 2-9 with the same line drawings and additional photographs. A modern parallel? It has recently been demonstrated that the version of the Mona Lisa in Madrid is not a later copy of the original painting in the Louvre but a simultaneous rendering by a painter working side by side with Leonardo, as indicated by the slightly different angle of the woman’s pose and the shifted elements of the background. Martin Bailey, correspondent with The Art Newspaper in London: "… the two pictures had been done side by side in the studio, and it was probably on easels which were two or three yards away from each other." (quoted in "The Mona Lisa's Twin Painting Discovered" (NPR)). Something comparable is now being advocated for the two versions of "The Virgin of the Rocks" in Paris and in London.
9. For references and a more extensive discussion, see my Hellenistic Sculpture I (1990) 227-28, and n. 20 on pp. 241-42. Enea Vico’s statement, in the original Italian, reads: "…e congiunta la parte del detto piede con quella, che alla statua mancava, si conobbe quella essere propria sua" (my emphasis).