accuracy in representing a person's features. More recent studies, especially on the recutting of images, assimilation, and the production of grave reliefs with portraits of the deceased, indicate that a need to produce accurate facial features was not always a motivating force, even for so-called veristic images. [64] For some ancients, accuracy in portrayal mattered; for others, it clearly did not. [65] Ultimately, the identity of the individual represented in any image in antiquity was established by the accompanying inscription and/or context. Some of these images, as we saw, could even be called symbolic portraits.

A study of the facial features and context clearly shows that some sculptors were not slavishly copying the iconographic hairstyles of the princeps, even though hairstyles were obviously easier to copy than three-dimensional facial features. The series of relief panels from the Sebasteion at Aphrodisias representing members of the Augustan and Julio-Claudian families [66] give ample testimony to the range of interest and/or ability of sculptors in reproducing physiognomic features and iconographic hairstyles. Some of these portraits are more recognizable by their facial features than by their hairstyles; others, more by their hairstyles than by facial features. Still others, who are obviously members of the Julio-Claudian family because of attributes and context, are scarcely recognizable by either portrait features or hairstyle. To cite a few specific cases from the Sebasteion, a figure that is clearly identifiable as Augustus by his facial features (cat. no. 103, pls. 185.1-2, 218) has an atypical hairstyl e, In this case, the sculptor was obviously capable but did not follow closely the typical configuration of locks over the forehead. Although the matter is not discussed by Boschung, the sculptor seems to have been trying to render at least the distinctive pincer formation of hair locks more or less correctly for the approximately three-quarter optimum viewing angle of the head. Another panel clearly shows a figure of a velificant Augustus (with billowing mantle) bringing the blessing of the Pax Augusta over land and sea, [67] a theme closely associated with Augustus in his Bildprogramm (representational program) and specifically mentioned in his Res Gestae: "cum per totum imperium populi Romani terra marique esset parta victoriis pax. ..." (Monumentum Ancyranum 13). [68] In this case, Augustus's hairstyle is even more unconventional (for which reason Boschung probably does not include this image in his catalogue, apparently because he rejects this figure as an image of Augustus). [69] Augustus's facial featu res, which are rather crudely carved, resemble him more in optimum view and from below [70] than straight on in frontal view, with foreshortened face. [71] In carving his relief figure of Augustus, the sculptor was probably taking into consideration the viewing angle. A third panel presents a heroic nude male with a captive northern barbarian nearby. [72] The heroized male appears to represent Augustus, but his crudely carved features bear only a superficial resemblance to him, although his hairstyle seems to be a simplified form of Augustus's Type V (Prima Porta type).

The variability in interests and ability of sculptors in reproducing physiognomic features and hairstyles is not limited to representations of Augustus in the relief program of the Sebasteion from Aphrodisias. For example, one relief sculptor working on this monument carved both the facial features and iconographic hairstyle of Claudius fairly well, [73] while another sculptor, who represented Claudius subduing a fallen personified Britannia, was far less successful in carving both. [74] Such departures in facial features and hairstyles from established prototypes are commonly found in relief sculpture because relief carvers were generally not trained as portraitists. Similar deviations are also to be found in sculpture in the round for a variety of reasons--the sculptor's lack of ability as a portraitist, artistic license, the tediousness of reproducing all the intricacies of hair patterns and lock formations, and other reasons that we cannot know.

Because of deviations in hairdos Boschung questions several fairly well-preserved heads as being portraits of Octavian/Augustus, for example, cat. no. ?218, pl. 206; cat. no. ?219, pl. 207; cat. no. ?220, pl. 208; cat. no. ?224, pl. 212. All of these portraits seem to me to represent Octavian/Augustus. One head, in particular, in Vienna (cat. no. ?219, pl. 207) clearly represents his facial features, although his hairstyle is slightly anomalous. A head from Fermo (cat, no. 95, p1. 157.2) shows a similar tendency to represent stringy hair locks over the forehead, although the locks are not as long. The formation of a rather odd pincer on the right side of the Vienna portrait is at least conceptually like that found on the right side of a head of Augustus in Istanbul (cat. no. 106, p1. 122.2). Boschung totally rejects as a portrait of Augustus a sculpture in the Rilley Collection in Cedar Rapids, Iowa (cat. no. [284.sup.*]). This image clearly portrays him physiognomically, but the hairstyle is atypical, resem bling Type V in the pincer effect but in reverse. [75] A colossal marble statue of a Genius from Puteoli in the Sala Rotunda in the Musei Vaticani is also rejected by Boschung as a portrait of Augustus (cat. no. [276.sup.*]), although this work appears to freely interpret Type IV.A (Stuttgart replica series). [76] The colossal size, cornucopia, dress, and artistic style of the figure argue for an image of Augustus's Genius, whose features have been more idealized, obviously to convey the concept of his divine procreative spirit. Also questioned by Boschung as an image of Augustus is a boyish-looking head in the Museo Archeologico Oliverano in Pesaro (inv. no. 3831, cat. no. ?224, pl. 212) that was probably recut from a lion-capped head of Hercules. Although not pointed out by Boschung, the facial features are very similar to those of a boyhood portrait of Augustus's adopted son Gaius in the same local museum (inv. no. 3294). [77] In this case, either the anomalous head is intended to be a portrait of Augustus based on a portrait of Gaius (Boschung's view) or a portrait of Gaius loosely imitating a hairstyle associated with Type IV.A. Of the two possibilities, the latter seems more likely because it would have been more reasonable for Gaius to have imitated Augustus than vice versa.

Another common phenomenon in Roman portraiture that can account for variability is assimilation, that is, making one individual resemble another, usually to suggest visually some sort of kinship--real or fictive--between two individuals or even between man and gods. [78] In portraiture of the leaders of the state, such visual assimilation--imitatio or aemulatio--could be politically motivated to promote a successor or even fictively claim kinship with a previous leader or dynasty. Boschung includes in his catalogue and discussion a number of images that show Augustus with assimilated features not only of his deified father Julius Caesar but also his Julio-Claudian successors (pp. 66--82). I would add to Boschung's list of assimilated portraits of Augustus the portrait bust from Neuilly-le-Real in the Louvre (inv. no. N 3254, cat. no. 55, p1. 63) that shows Augustus with assimilated facial features of Julius Caesar, especially when seen in frontal view. [79] Conversely, images on the reverses of aurei and den arii of the moneyer Marcus Sanquinius [80] produced in connection with the Secular Games show the deified Julius Caesar with the Sidus lulium (Comet of Julius Caesar) over his head and with features assimilated to those of Augustus. [81] We also find portraits of Augustus's grandsons and adopted sons Gaius and Lucius created during the lifetime of Augustus that look like him [82] and even somewhat recall his earlier hairstyles. It would appear that Gaius's later portrait types (compare, for example, "Corinth 135") [83] recall in the fringe of forehead locks Augustus's Type III (Actium/Alcudia type), while Lucius's later portrait type, with its tightly closed pincer (compare, for example, "Corinth 136"), [84] recalls Type II of Octavian (Lucus Feroniae type), with its tightly interlocked forehead pincer. [85]

Interesting for the problem of the portraiture of Gaius and Lucius Caesar are three portraits that Boschung (p. 50) believes form a "replica string" (his "Genf-Koln Replikenstrang") of portraits of Augustus that are variants of the Prima Porta type. These portraits he believes were influenced by the portraiture of one of his grandsons and adopted sons (Lucius). [86] Of the three portraits, only a head in Geneva (cat. no. 102, p1. 94) bears a very strong resemblance to Augustus, and this one only in profile view. In frontal view he looks quite different, more idealized and youthful. Boschung does not note, however, that the Geneva head resembles physiognomically in profile and even somewhat in frontal view a portrait, now in the Uffizi in Florence (cat. no. 100, p1. 106), which Boschung takes as a variant of the Prima Porta type. The pincer effect and forking of the hair in relation to the eyes and nose are similar in these portraits, even if the reverse-comma-shaped locks over the right eye are reduced in nu mber as compared with the Geneva head. Nevertheless, the pattern of hair locks on both sides of both heads is remarkably similar. One distinctive hair pattern found in both, as well as in other examples of Augustus's portrait types, is a forking of the locks above the right ear. Interestingly, too, this feature is found in some of Gaius's mature portraits, no doubt because of the influence of Augustus's hairstyle on that of Gaius. [87]

The Geneva portrait is seen as being closely related to a poorly preserved head in Cologne (cat. no. 109, p1. 95) and a much better-preserved one in Patras (Fig. 12), which Boschung does not illustrate (cat. no. 154). [88] The heads in Cologne and Patras are obviously related to one another not only in general facial features but also in hairstyle. In the better-preserved Patras head, there is little physiognomic resemblance to the Geneva head, and the patterns of locks on the sides of the head are also different. And although there is a general resemblance to the Geneva head in the fringe of locks over the forehead, the pincer is tightly interlocked in both the Patras and Cologne heads, in contrast to the open pincer of the Geneva head. Is this hairstyle really meant to reflect the Type V (Prima Porta) hairstyle, as Boschung suggests? Perhaps it was meant to recall Augustus's Type II hairstyle, especially as seen in the London head (cat. no. 3, p1. 4), under the influence of Type V. Supporting this supposit ion is the youthfulness of the Patras and Cologne heads (as well as the Geneva head). However, it is not clear that at least the Patras and Cologne heads were in fact meant to represent Augustus. These heads may have been intended to portray one of his adopted sons, most likely Lucius, on the basis of a comparison with his portrait statue in Corinth ("Corinth 136"). [89] Lucius's portrait features, including the deep-set eyes and the fringe of locks over the forehead, find a close comparison in the Patras head. In short, the Patras and Cologne heads may show not the influence of Lucius's portraiture on Augustus's portraits, as Boschung suggests, but vice versa.

A further issue to be considered is whether some of the portraits that Boschung takes as Augustus with assimilated features are really portraits of Augustus at all. In studies ruled by the Lockenzahlmethode, it is generally assumed that the hairstyle rather than the portrait features establishes who is represented. However, if we allow for the possibility that at least an occasional portrait of one of Augustus's successors might have been produced with a hairstyle that imitated that of Augustus, a few works included in Boschung's catalogue as images of Augustus become questionable. Three portraits, in particular, may very well be images of Tiberius with hairstyles like that of Augustus: a small bronze bust from Herculaneum in the Museo Nazionale di Napoli with some classicizing features (cat, no. 52, pl. 62), a head in L'Aquila (cat. no. 116, p1. 107), and a somewhat restored marble head in the Museo Capitolino in Rome (cat. no. 164, pl. 160.1-2).

Another phenomenon in ancient portraiture involving assimilation is Zeitgesicht, in which, as mentioned above, a private individual's portrait was made to resemble the princeps in physiognomy as well as hairstyle to some degree. Such private portraits were executed in the artistic style of the times. A good example of Zeitgesicht is a portrait in the Galleria degli Uffizi in Florence that was identified as Augustus [90] but most likely portrays a young private individual (Fig. 13); the portrait features show some resemblance to those of Augustus, with a reverse Type V pincer effect of the hair locks over the forehead. In some cases examples of Zeitgesicht may be difficult to establish. [91] Boschung cites two portraits as "unsure" images of Augustus because their hairstyles seem too freely interpreted. To me, both are images of private individuals, one (cat. no. ?222, pl. 210) with slight assimilation of facial features to those of Augustus (more in profile), the other (cat. no. ?223, p1. 211) to those of Caligula. Two other portraits that Boschung classifies as images of Aug ustus I believe may also be portraits of private persons with assimilated hairstyles and/or some facial characteristics of Augustus: cat. no. 62, pl. 65 (the caption of the plate gives the incorrect cat. no. as 52) and cat, no. 61, pl. 66.

The imitation by private individuals of hairstyles of leaders and members of their family is hardly unique to the Romans. In a late nineteenth-century painting by Jean-Georges Vibert, In the Image of the Emperor (Fig. 14), [92] a cleric stands before a bust of Napoleon holding a mirror as he combs his hair to imitate the hairdo of the emperor, while on the wall behind him hangs a painted portrait of Napoleon in the same three-quarter pose as the cleric. Even the cleric's face and expression have taken on a Napoleonic look. During the Kennedy years, a number of women copied the First Lady's hairstyle and outfits, while men have imitated Elvis Presley's hair and look. In more extreme cases, individuals have even undergone plastic surgery to make themselves look like various celebrities. Although it may have been considered bad taste, there were no Roman laws governing a private individual's appropriation of a princeps'iconographic hairstyle.

Because of the subjective nature of portrait studies and the constant addition of new archaeological material, no such work can ever really be considered a definitive publication. Added to the corpus in the future will be sculptures like the relatively recently discovered marble portrait of Augustus from the theater at Troy, which could not be included in Boschung's work. Other pieces that have been little published in the past should also be included, like the sculpture in Lowther Castle in Lowther, England. [93] Given the rapid developments in new technologies, I can even see in the not too distant future computer databases replacing bound catalogues. In this way, all newly discovered and attributed portraits could be easily added to a corpus. Eventually, too, such a database would allow a researcher to rotate portraits at will on screen to facilitate three-dimensional comparison.

Notwithstanding the possible future use of new technologies in portrait studies, as well as my own criticisms here, it cannot be stressed enough that Dietrich Boschung has provided a much needed, well-organized, well-documented, well-argued typological study of the portraits of Augustus. For this excellent volume, we owe him, as well as those who laid the groundwork for this publication, an enormous debt of gratitude.

Postscript

On a recent visit to the University of Koln (Cologue, Germany), I learned that the university's Archaeological Institute has undertaken to put on the web their extensive photograph archive (Forschungsarchiv fur Antike Plastik), including photos of Augustus's portraits based on Boschung's catalogue. The website is www.uni-koeln.de/phil-fak/ai/mar/forindex.htm.

Notes

(1.) Many of Zanker's own recent views are to be found in Klaus Fittschen and Paul Zanker, Katalog der romischen Portrats in den Capitolinischen Museen und den anderen Kommunalen Sammlungen der Stadt Rom, vol. 1 (Mains: Philipp von Zabern, 1985), 1--10.

(2.) Boschung includes 211 of Augustus's sculptural portraits in his catalogue.

(3.) For this material, see most recently the general study of Wolf- Rudiger Megow, Kameen von Augustus bis Alexander Servers Antike Munzen und Geschnittene Steine, vol. 11 (Berlin: Walter de Gruyter, 1987).

(4.) See, for example, the image of Augusts on the following: Gilded silver cup from Meroe in the Boston Museum of Fine Arts: John Pollini, Studies in Augustan 'Historical Reliefs," Ph.D. diss., University of California, 1978, 281--82, with further bibliography; Boscoreale cups: recently, Ann L. Kuttner, Dynasty and Empire in the Age of Augustus (Berkeley: University of California Press, 1995); Possibly the gilded silver cup in the British Museum: Pollini, 282--83, with further bibliography; Stucco decoration in cubiculum E in the Villa Farnesina in the collection of the Museo Nazionale Romano: Pollini, 325--26, with further bibliography; cf. Irene Bragantini, in Museo Nazionale Romano, le pitture. Le decorazioni della villa romana della Farnesina, ed. Irene Bragantini and Mariette de Vos, vol. 2, pt. 1 (Rome: De Luca, 1982), 292 and n. 14, p1. 200. For a detail of the head, see Klaus Vierneisel and Paul Zanker, Die Bildnisse des Augustus: Herrscherbild und Politik in kaiserlichen Ron (Munich: Glyptothek Mun chen, 1979), 15.

For possible images of Augustus in other media, see further Pollini, 256--342; and Erika Simon, Augustus: Kunst und Leben in Rom um die Zeitenwende (Munich: Hirmer, 1986).

(5.) These and other issues not dealt with in this review will be discussed in a book on which I am working entitled The Image of Augusts: Art, Ideology, and the Rhetoric of Leadership.

(6.) For a comprehensive treatment of imperial statue type in various contexts, see Thomas Pekary, Das romische Kaiserbildnis in Staat, Kult und Gesellschaft: Dargestellt Anhand der Schriftquellen, Das romische Herrscherbild, vol. 5 (Berlin: Gebruder Mann, 1985).

(7.) See, for example, Fittschen and Zanker (as in n. 1), pl. 2.

(8.) For the meaning of a portrait in ancient Rome, see esp. Sheldon Nodelman, "How to Read a Roman Portrait," Art in America 63 (1975): 26--33. For its meaning in the context of Western art, see Richard Brilliant, Portraiture (Cambridge, Mass.: Harvard University Press, 1991).

(9.) See, for example, an atypical portrait of Augustus in the Rilley Collection (below, with n. 75 below.

(10.) For this phenomenon, see Paul Zanker, "Herrscherbild und Zeitgesicht," Wilssenschaftliche Zeitschrift der Humboldt-Univecsitat zu Berlin, Gesettschaft und Sprachwissenschaft, 31st ser. (1982): 307-12.

(11.) Mus. inv. 605: see P.J. Riis, "An Aeneas in the Ny Carlsberg Glyptotek?" in In Memoriam Otto J. Brendel: Essays in Archaeology and the Humanities, ed. L. Bonfante and H. von Heintze (Mainz: Philipp von Zabern, 1976), 165-70, fig. 1, pl. 41a,b; Fulvio Canciani, in Lexicon Iconographicum Mythologiae Classicae, vol. 1 (Zurich: Artemis, 1981), s.v. "Aineis," 394 (no. [214.sup.*]). Although possibly created during the Hadrianic period, this head more likely dates to the principate of Antoninus Pius, whose epithet "Pius" recalls that of Pius Aeneas (not mentioned by Riis). Antoninus Pius's admiration for Aeneas is evidenced, too, by a commemorative bronze medallion series (also not mentioned by Riis) on which Antoninus appears, like Aeneas, sacrificing the Sow of Lavinium, as on the tight front panel of the Ma Pads. For an example of this medallion, see J.P.C. Kent et al., Roman Coins (New York: Harry N. Abrams, 1978), 297, no. 313, p1. 91. Cf. also the Ny Carlsberg head and the head of Aeneas in the "Aeneas " panel on the Ma Pads: Riis, p1. 41d. A similar, contemporary colossal marble image from Tivoli in the J. Paul Getty Museum (inv. 76.AA.10) would likewise appear to be an invented portrait of Aeneas; however, in this example of Zeitfrisur, the fringe of forehead locks combines both Augustus's so-called Prima Porta type and the old so-called Actium type, renamed by Bosehung the Alcudia type. For this little-known head, see J. Chamay, J. Frel, and J.-L. Maler, Le monde des Cesars (Geneva: Musee d'Art et d'Histoire de Geneve, 1982), 59, p1. 1.1-3.

(12.) For this temple, see Cyril Aldred, The Temple of Dendur (New York: Metropolitan Museum of Art, 1978). For other similar images of Augustus, see the following monuments: gateway of the Temple of Augustus from Kalabsha in the Agyptisches Museum, Berlin: Jurgen Settgast, Joachim Selim Karig, and William Brashear, Agyptisches Museum Berlin (Staatliche Mnseen Preussischer Kulturbesitz) (Braunschweig: Georg Westermann, 1981), 112-13; Temple of Chnum in Elephantine: Simon (as in n. 4), 222; Temple K. at Philae: Lily Ross Taylor, The Divinity of the Roman Emperor (Ann Arbor, Mich.: University of Michigan Press, 1931), 144, fig. 22; Possibly a granite stale from Girge near Abydos in the Agyptischen Museum, Berlin: W. Iwas, "Augustus als Pharao-Eine Stele im Agyptischen Museum Berlin," Wissessschafihiche Zeitschnft der Humholdt-Universital zu Berlin, Gesellschaft und Sprachwissenschaft, 31st oar. (1982): 221-22.

(13.) Georg Lippold, Kopien und Umbildungen griechischer Statuen (Munich: Oskar Beck, 1923), 2-3, passim.

(14.) Zanker, in Fittschen and Zankar (as in n. 1), 1-2, no. 1; and Klaus Fittschen, in "Die Bildnisse des Augustus," in Soeculum Augustum, ed. Gerhard Binder, vol. 3, Wage dar Forschung, 632 (Darmstadt: Wissensehaftilehe Buchgesellschaft, 1991), 163-64. Both speculate that this type might have been used for the equestrian statue set up in Rostris in 43 B.C.E. On this statue, see further below.

(15.) The dating of this sesterius with the legend DIVOS IVLIVS to Ca. 40 B.C.E. is based on the inscriptions containing the words DIVOM IVLIVM on lead slingshot bullets (glandes plumbeae) used by Octavian's troops in the Perusine War during the winter of 41-40 B.C.E.; see Andreas Alfoldi and Jean-Baptiste Giard, "Guerre civile et propaganda politique: L'emission d'Octave au nom du Divos Julius (41-40 avant J.C.)," Numismatica e Antichita Classiche 13 (1984): 147-53. This sort of evidence, however, is hardly conclusive. DIVOS IVLIVS could have appeared in any medium at any point after the official deification of Caesar in 42 B.C.E.

(16.) The phrase in Rostris need not mean on the Rostra; it could also mean in the area of the Rostra. For the varying numismatic versions of this statue, see John Pollini, "Man or God: Divine Assimilation and Imitation in the Late Republic and Early Principate," in Between Republic and Empire: Interpretations of Augustus and His Principate, ed. Kurt A. Raaflaub and Mark Toher (Berkeley: University of California Press, 1990), 343, with further bibliography.

(17.) Velleius Paterculus, 2.61.3; Appian, Belle Civilia 3.51; Cassius Dio, 46.29.2; Cicero, Epistulae ad Brutum 24.7.

(18.) Cf. the Munich Glyptotek "Marius" and "Sulla" portraits: Luca Giuliani, Bildnis und Botschaft: Hermeneutische Untersuchungen zur Bildniskunst der romischen Republik (Frankfurt: Suhrkamp, 1986), figs. 48-52, with page references.

(19.) For this toning-down effect (Pathosdampfung), see also the comments of Klaus Fittschen, "Pathossteigerung und Pathosdamfung: Bermerkungen zu griechisehen und romischen Portrats des 2. und 1. Jahrhunderts v. Chr.," Archaologischer Anzeiger (1991): 253-70.

(20.) John Pollini, The Portraiture of Gains and Lucius Caesar (New York: Fordham University Press, 1987), cat, no. 5, p1. 7. Otto Brandel, Ikonographie des Kaisers Augustus (Nuremberg: Buchdruckerei Eduard Kreller, 1931), distinguished five types, two of which do not represent Octavian/Augustus.

(21.) See, for example, the recent book by Pat Southern, Augustus (London: Routledge, 1998), p1. 1.

(22.) Pollini (as in n. 20), 59-75, passim.

(23.) See also Bosehung, 59 n. 241, 61 and n. 260.

(24.) The notion of the "heard of mourning" is based on a late source, Cassius Dio (48.34.3), who claims that Octavian did not shave off his heard until 39 B.C.E. Octavian is, however, still shown with facial hair on coins until 36 B.C.E.: Michael H. Crawford, Roman Republican Coinage (Cambridge: Cambridge University Press, 1974), nos. 534/3, 538/1, 540. Although Octavian may have let his beard grow during the period of mourning for Caesar, the type of long sideburns and/or narrow, neatly trimmed beard following the jawline onto the chin that shows on his coins suggests a "heard of vengeance," appropriate for one who as Caesar's legal heir was to avenge Caesar's murder. After the assassins had been dealt with, Octavian's continued use of the beard remained essentially "military" to compensate for his youth and to highlight his being an Ares/Mars Ultor-like military leader. In short, I think Dio, our only source, is somewhat confused about the "heard of mourning." When a Roman male was in mourning, he purpos ely neglected shaving as an expression of grief. Hence, the heard would have been stubbly and covered the entire lower part of his face. It would not be the neatly trimmed side-whiskers or narrow beard that appears in the coin images of Octavian. On this matter, see further Pollini (as in n. 20), 63 n. 108, 71-73, 91.

(25.) Crawford (as in n. 24), 499-500, no. 490.

(26.) Ibid., 502, no. 494/3a, 506, no. 494/ 25.

(27.) Cf. both numismatic images: ibid., nos. 494/3a, 494/25.

(28.) This recut portrait may have been fashioned out of a head of Hercules wearing a lion-skin cap. Cf. Boschung, 107, cat, no. 2.

(29.) See Suetonius, Divus Augustus, 86. For Caesar and Augustus and their association with the neo-Attic movement, see George A. Kennedy, The Art of Rhetoric in the Roman World: 300 B.C-AD. 300 (Princeton: Princeton University Press, 1972), asp. 378-84. For the neo-Attic style in general, sea also 69-70, 241-46, passim.

(30.) Kennedy (as in n. 29), 297-99.

(31.) This style is exemplified by the Prima Porta statue of Augustus; see John Pollini, "The Augustus from Prima Porta and the Transformation of the Polykleitan Heroic Ideal," in Polyskleitos, the Doryphoros, and Tradition, ad. Warren G. Moon (Madison, Wis.: University of Wisconsin Press, 1995), asp. 262 and n. 3, for further bibliography on the subject of the polemics of style in the late Republic and the beginnings of Augustan Classicism.

(32.) Hans Jucker, in Gesichter: Griechische und romischse Bildnisse aus Schweizer Besitz, ad. Hans Jucker and Dietrich Willers, 3d ad. (Bern: Archaologisches Seminar der Universitat Bern, 1983), 53, fig. 4.

33.) Simon (as in n. 4), 66, attributes the creation of this type only to Augustus's becoming Pontifex Maximus in 12 B.C.E. Moreover, she believes that the Capitoline head of Augustus from the Via Merulana on the Caclian Hill in Rome (fig. 79; Boschung, cat. no. 45, p1. 38) may possibly be one of the rare examples of an actual prototype ("Archetypus") surviving-a hypothesis I find unlikely.

(34.) Jucker (as inn. 32).

(35.) First proposed by Ulrich Hausmann in a lecture in Bochum in 1971, as noted in Hausmann, "Zur Typologie und Ideologie des Augustportrats," in Aufstieg und Niedergang der romischen Welt: Geschichte und Kultur Roms im Spiegel der neuren Fonschung, Principat, Kunst, ed. Hildegard Tamporini, vol. 2, pt. 12, no. 2 (Berlin: De Gruyter, 1981), 567-75, esp. 569-70 and n. 216, pl. 20, fig. 39. Cf. also Paul Zanker, Studien zu den Augustus-Portrats: I, Der Actium-Typus, 2d ad. (Gottingen: Vanderhoeck und Ruprecht, 1978), 44 n. 120.

(36.) Pollini (as in n. 31), 266.

(37.) For the mistaken notion of official works being made for the "average" viewer, see ibid., 263-64.

(38.) Most of these extant portraits were set up by private individuals or groups of people. There may also have been cases in which only a particular model was available or a certain earlier type was preferred. This situation is somewhat analogous to anniversary coin issues. See in general Michael Grant, Roman Anniversary Issues (Cambridge: Cambridge University Press, 1950).

(39.) Cf. Dietrich Boschung, Die Bildnisse des Caligula, Das romische Herrscherbild, pt. 4, vol. 1 (Berlin: Gebruder Mann, 1989), pls. 2-4.

(40.) See Pollini (as in n. 20), 3-5, 16-17, passim.

(41.) The modern technical term damnatio memoriae is unattested in the ancient written record but has often been used to characterize the idea of damning one's memory. On the other hand, memoria damnata was a phrase used in antiquity. On the subject of damnatio memoriae, see Pollini (as in n. 20), 13 and n. 46, for further bibliography. For a more recent comprehensive treatment, see Pekary (as in n. 6), 134-42; and Eric Varner, "Damnatio Memoriae and Roman Imperial Portraiture," Ph.D. diss., Yale University, 1993, which is currently being revised for publication.

(42.) See Robert Amy and Pierre Gros, La Maison Carree de Nimes, 2 vols., "Gallia" Supplement, 38 (Paris: Centre National de la Recherche Scientifique, 1979).

(43.) Augustus, for example, ordered that silver statues set up to him in Rome be melted down (Monumentum Ancyranum 24).

(44.) See, for example, Paul Zanker, Provinzielle Kaiserportrats, Bayerische Akademie der Wissenschaften, Munchen, Philosophisch-histotischa Klasse, Abbandlungen, vol. 90 (Munich: Verlag der Bayarischen Akademie der Wissenschaften, 1983), who deals with this issue in a general way, especially for the second century A.D.

(45.) For the Prima Porta statue and its postulated bronze prototype, see recently Pollini (as in n. 31), with further bibliography.

(46.) For variability in the degree of the pathos formula in portraiture, see in general Fittschen (as in n. 19).

(47.) Already after his victory over Sextus Pompey in 36 B.C.E., Octavian had wanted to put to rest memories of the Civil Wars (Appian, Belle Civilia 5.132).

(48.) See John Pollini, "The Gamma Augustea: Ideology, Rhetorical Imagery, and the Construction of a Dynastic Narrative," in Narrative and Event in Ancient Art, ed. Peter J. Holliday (Cambridge: Cambridge University Press, 1993), 263, 266-67.

(49.) See Pollini (as in n. 31).

(50.) Cf. Boschung, 34-37, 63.

(51.) Ibid., 33-34, 37, 63.

(52.) It has generally been assumed on the basis of Strabo's use of the word eikon (5.3.8) that this image was a statue pedestris. For the proposed quadrigate image, sea Pollini (as in n. 48), 285, fig. 86.1 discuss in detail my reasoning for this quadrigate imagery in a forthcoming book entitled The Rhetoric and Poetry of Visual Imagery and the Creation of Dynastic Narratives in Augustan Art and Thought.

(53.) Boschung (125) takes this image of Augustus as only a Paris Louvre MA 1280 type (Type IV) portrait. Although the hair is damaged in part, the bulge in the hair locks over the right eye (unlike the prototype of Type IV) and their down-turned tips, which are abraded, are more suggestive of Type IV.A (cf. a head from Terracina [cat. no. 59, pl. 58]).

(54.) For a detailed treatment of this event, see Pollini (as in n. 20), 32--35, pl. 14.1. For more recent bibliography, see Boschung, 125--26 (cat. no. 36). Gaius's departure was probably to coincide with the dedication of the Forum Augustum and his being the first military commander to set out from here. See Frank E. Romer, "A Numismatic Date for the Departure of C. Caesar?" Transactions of the American Philological Association 108 (1978): 187--202.

(55.) Pollini (as in n. 4), esp. 87--89, 95--99, 126--32; and idem, "The Acanthus of the Ara Pacis as an Apolline and Dionysiac Symbol of Anamorphosis, Anakyklosis and Numen Mixtum," in Von der Bauforschung zur Denkmalpflegue, Festschrift fur Alois Machalschek, ed. Martin Kubelik and Mario Schwarz (Vienna: Phoibos, 1993), esp. 214--17, for the wider implications of this augural act. Although I had proposed in my dissertation that the augural act represented had to do with marking out the augural templum terrestre, on which the monumental altar was to be built, it is equally possible that the augural act was a maximum augurium salutis Rei Publicae. See further ibid., 216; and idem, "Frieden-durch-Sieg-Ideolog und die Ara Pacis Augustae: Bildrhetorik und die Schopfung einer dynastischen Erzahlweise," in Interdisziplinares Kolloquium Historische Architekturreliefs yam Alten Agypten bis zum Mittelalter (forthcoming).

(56.) See Cassius Dio, 37.24; and Kurt Latte, Romische Religionsgeschichte (Munich: Beck'sche Verlagsbuchhandlung, 1960), 140--41, 289--300.

(57.) Erika Simon, Ara Pacis Augustae (Greenwich, Conn.: New York Graphic Society, 1967), 9, first noted the similarity between the two structures. To my mind, the Ma Pacis was in effect a monumental counter response to the Forum Shrine of Janus Geminus, which Augustus restored. Although bifrontal emphasis is evident in Greek temples and sacred structures, the double doors and bifrontality of the Ara Pacis was probably intended to recall more directly the Shrine of Janus, since a Roman janus is a double-gated passageway. This relationship between the architectural form of the Ara Pacis and the Shrine of Janus cannot have been coincidental because we know that opening or closing the doors of Janus signified (respectively) a state of war or peace. Marks for doorposts in its doorsill indicate that the Ma Pacis also had closing doors.

(58.) Cassius Dio, 51.20.4; Augustus, Monumentum Ancyranum 13. See also Suetonius, Divus Augustus 31.4; and Livy, 1.19.2, who says only that it was closed after Actium. The first closing formally marked the cessation of the Civil Wars; the second took place after the Cantabrian War in 25 B.C.E.; the third, of unknown date, presumably occurred after Augustus's return from Spain and Gaul in 13 B.C.E., the year in which the Ma Pads was "constituted" through an augural act.

(59.) This is true also for most portraits of Type IV with sure provenance; Boschung, 86, 90.

(60.) In the case of the posthumous altar image in Palestrina (cat. no. 63, pl. 67.1-2), which Boschung classifies as belonging to his Kopenhagen 611 replica series, the hair is too damaged to indicate to what type this portrait belongs. Also, unlike the other three portraits, the Palestrina head turns slightly to the right, rather than to the left.

(61.) See Crawford (as in n. 24), 499, nos. 490/1--4, 740 and n, 4.

(62.) See Pollini (as in n. 20), 2--3 and n. 11, for further bibliography on this subject.

(63.) See, for example, graffito from the Domus Augustana (81--96 C.E.): Vierneisel and Zanker (as in n. 4), 15 (sketch).

(64.) For veristic-style funerary images, see esp. Paul Zanker, "Zur Rezeption des hellenistisehen Individualportrats in Rom und in den italisehen Stadten," in Hellenismus in Millelitalien, Abhandlungen der Akademie der Wissenschaften in Gottingen, vol. 97 (Gottingen; Vanderhoeck und Ruprecht, 1976), 581--619, esp. 592--605.

(65.) For the evidence, see Pollini (as in n. 20), 11.

(66.) For these reliefs, see R.R.R. Smith, "The Imperial Reliefs from the Sebasteion at Aphrodisias," Journal of Roman Studies 77 (1987): 88--138; and more recently Charles Brian Rose, Dynastic Commemoration and Imperial Portraiture in the Julio-Claudian Period (Cambridge: Cambridge University Press, 1997), 164--69, cat, no. 105, passim.

(67.) See Smith (as in n. 66), 103--6, pls. VI--VII.l-2; and John Pollini, "The Tazza Farnese: Principe Augusta 'Redeunt Saturnia Regna!"' American Journal of Archaeology 96 (1992): 285-86, fig. 1.

(68.) Smith (as in n. 66), esp. 261--62, with reference to this relief from Aphrodisias.

(69.) Boschung does not even include this relief in either of his catalogue sections: "Unsichere Augustusportrats" (196--97) and "Falschlich als Augustus benannte Portrats" (202--4).

(70.) Smith (as in n. 66), pl. VI.

(71.) Ibid., pl. VII.2, an angle that the ancient viewer would not have been able to see from the ground.

(72.) Ibid., 120--21, no. 8, pls. XVIII, XIX, takes this figure as a possible Tiberius. Unlike this portrait image, however, Tiberius generally is represented with a smallish mouth with thin lips, and his forehead tends to bow forward. Normally, too, the locks that Tiberius wears high over his forehead are small and numerous, although in some portraits they take on a more Augustus-like effect, probably for dynastic reasons. The facial features of this relief portrait are more like those found in other portraits from Asia Minor; cf., for example, a head from Ephesos in the Selcuk Museum (Boschung, cat. no. 26, p1. 24.2--4) and a head in Konya (Boschung, cat, no. 111, pl. 190.1-2). Cf., also, the overlife-size head in the Museo Gregoriano Profano (Boschung, cat, no. 174, pl. 139). The left arm of the pincer formation of hair locks of Type V is shoved into the right pincer, as is sometimes found in various replicas of this type; cf., for example, Boschung, cat, no. 89, pl. 169; cat, no. 78, pl.90.

(73.) Smith (as in n. 66), pls. VIII, IX.2; cf. with the portrait of Claudius in the round in the Ny Carlsberg Glyptotek, pl. IX.4.

(74.) Ibid., pls. XIV, XV.1-2.

(75.) See Adrian Stahli, in Jucker and Willers (as in n. 32), 74--75, no. 27, who correctly identifies this as a portrait of Augustus. Analogously, in his volume in the Herrscherbild series on Caligula (as in n. 39), Boschung also questions a bust in the Metropolitan Museum in New York (cat. no. ?46, pl. 37) as representing Caligula because its hairstyle is markedly different from any of Caligula's known iconographic hairstyles. Although Boschung thinks this could be a portrait of a private person (an extreme case of Zeitgesicht!) ends up favoring the possibility that this may be a workshop variant (Werkstattvariante) or a subtype (Nebentypus) because its facial features resemble those of Caligula. This atypical portrait, in my opinion, definitely represents Caligula, although it portrays him in a more idealized manner to recall images of Augustus. In fact, the small, closed central pincer effect of locks over his forehead, which differs from the hairstyles of Caligula's main types, seems to me intended to r ecall Augustus's Type V hairstyle, the pincer effect of which is normally more open. For a somewhat similar division of the locks of the central, more open pincer of an Augustus Type V portrait, see the head from Fermo (cat, no. 95, pl. 157.2 [incorrectly given in the plate caption as cat, no. 1001). Cf. also the image of Caligula on a cameo in the Kunsthistorisches Museum in Vienna (cat. no. 34, pl. 30.1-2), which was long taken to be a portrait of Augustus but was correctly shown by Helmut Kyrieleis ("Zueinem Kameo in Wien," Archaologische Anzeiger [1970]: 492--97) to be that of Caligula. Here, too, Caligula's facial features have been slightly assimilated to those of Augustus. Another example of an atypical portrait of Caligula is a small bronze bust (cat. no. ?48, pl. 39.1-4), which Boschung also finds questionable but which, based on facial features alone, surely must represent him.

76.) In addition to the bibliography cited by Boschung in cat, no. [276.sup.*], see also Hans Jucker, "Iulisch-claudische Kaiser- und Prinzenportrats als `Palimpsete,'" Jahrbuch des Deutschen Archaologischen Instituts 96 (1981): 244, fig. 11. Some locks forming the fringe of hair over the forehead in Boschung's Stuttgart replicas also show some variation in arrangement.

(77.) Sea Pollini (as in n. 20), 95, cat, no. 1, pl. 3.

(78.) On this phenomenon in the portraiture of historical personages, see in general Anne-Kathrein Massner, Bildnisangleichung: Untersuchungen zur Entstehungs- und Wirkungsgeschichte der Augustusportrats (Den romische Herrscherbild, pt. 4) (Berlin: Gebruder Mann, 1982); for assimilation to divinities, especially with regard to the Augustan period, see Pollini (as in n. 16), 334--63, with further bibliography.

(79.) Cf. another head of Augustus whose facial features likewise appear to have been assimilated to those of Julius Caesar: Boschung, cat, no. 155, pl. 154.24.

(80.) Harold Mattingly, Coins of the Roman Empire in the British Museum, vol. 1, 2d ad. (London: British Museum, 1976), 13, nos. 69--73, pl. 2.19-20; and C.H.V. Sutherland and R.A.G. Carson, ads., Roman Imperial Coinage, vol. 1 (London: Spink and Son, 1984), 66, nos. 337--42, pl. 6.340. See further Pollini (as in n. 16), 352--53, pl. 26 (the name "Sanguinius" in the article should be corrected to read "Sanquinius").

(81.) This would represent the opposite effect of assimilation that I see in the bronze bust from Neuilly-le-Real in the Louvre, Boschung, cat. no.55, pl. 63.

(82.) Pollini (as in n. 20), 10--11,93, passim.

(83.) Ibid., cat, no. 14, pl. 16. Cf. also others of Types III through V, with three or more dominant comma-shaped locks over the center of the forehead forming an open pincer with the reverse-comma-shaped locks over the right eye.

(84.) Ibid., cat, no. 38, pl. 39.

(85.) See esp. the head in the British Museum, inv. no. 1878: Boschung, car, no. 3, pl. 4.

(86.) Ibid., 153, cat. no. 102, pl. 94, as represented by his "Cassino-Karthago" group, which he takes to be Lucius. Cf. also idem, "Die Bildnistypen deciulisch-claudischen Kaiserfamilie: Em kritischen Forschungsbericht," Journal of Roman Archaeology 6 (1993): 52--54.

(87.) See, for example, Pollini (as in n. 20), pls. 15.3, 16.3, 19.3, 21.3, 27.3, 31.3, 32.3.

(88.) Deutsches Archalogisches Institut (Athens), nags. 70/480--70/484.

(89.) Pollini (as in n. 20), cat. no. 38, pl.39.

(90.) See, for example, Guido A. Mansuelli, Galleria degli Uffizi: Le sculture, vol. 2 (Rome: Istituto Poligrafico dello Stato, 1961), 59--60, no. and fig. 47.

(91.) See, for example, a bead taken by Zanker (as in n. 10), 300, fig. 210, as a case of Vespasianic Zeitgesicht that is, in fact, a portrait of Vespasian recut from a portrait of Nero: John Pollini, "'Damnatio Memoriae' in Stone: Two Portraits of Nero Recut to Vespasian in American Museums," American Journal of Archaeology 88 (1984): 548--52.

(92.) See Brilliant (as in n. 8), 83--85, fig. 33.

(93.) Sea Paul Arndt and Georg Lippold, ads., Photographische Einzelaufnahmen antiker Skulpturen, 9th ser. (Munich: F. Bruckmann Aktiengasellschaft, 1929), no. 3074. In these photographs, the portrait appears to be authentic.

1 Portrait of Octavian, Beziers-Spoleto type (Boschung), or Type I (Pollini). Perugia, Museo Archeologico Nazionale deli' Umbria (from Zanker, Der Actium-Typus, pl. 18b)

2 Portrait of Octavian, Lucus Feroniae type (Boschung), or Type II (Pollini). Lucus Feroniae, Antiquario

3 Portrait of Octavian, Alcudia type (Boschung), or Type III (Pollini). Alcudia, private collection (from Zanker, Der Actium-Typus, pl. 1)

4 Portrait of Octavian, Paris Louvre MA 1280 type (Boschung), or Type IV (Pollini). Paris, Musee du Louvre (from Boschung, Die Bildnisse des Augustus, pl. 36)

5 Portrait of Augustus, Prima Porta type (Boschung), or Type V (Pollini). Musei Vaticani. Braccio Nuovo

6 Portrait of Octavian/Augustus, Stuttgart replica group (Boschung), or Type JV.A (Pollini). Stuttgart, Wurttembergisches Landesmuseum (from Boschung, Die Bildnisse des Augustus, pl. 52)

7 Portrait of Octavian/Augustus, Kopenhagen 611 replica group (Boschung), or Type IV.B (Pollini). Copenhagen, Ny Carlsberg Glyptotek (from Van Poulsen, Les portraits romains: Republique et dynastie julienne, vol. 1 [Copenhagen: Ny Carlsberg Glyptotek, 1974], pl. 51)

8 Invented portrait of Aeneas. Copenhagen, Ny Carlsberg Glyptotek

9 Relief image of Augustus as pharaoh on the Temple of Augustus from Dendur. New York, The Metropolitan Museum of Art (from Aldred, The Temple of Dendur, fig. 39)

10 Image of Octavian on sestertius (from Kent et al., Roman Coins, pl. 33, no. 115)

11 Portrait of Gaius Caesar, Augustus's adopted son. Florence Galleria degli Uffizi

12 Portrait of Octavian or his adopted son Lucius, in Patras, Greece, Museion

13 Portrait of an unknown individual made to resemble Augustus. Florence, Galleria degli Uffizi (from Mansuelli, Galleria degli Uffizi: Le sculture, vol. 2, fig. 47)

14 Jean-Georges Vibert, In the Image of the Emperor. Location unknown (from Brilliant, Portraiture, fig. 33)

Die Bildnisse des Augustus, Das romische Herrscherbild, pt. 1, vol. 2
(Review used with permission of Prof. Pollini)

Berlin: Gebruder Mann Verlag, 1993. 252 pp.; 239 b/w ills., 9 foldouts. DM 290.

This volume on the sculptural portraiture of Augustus, arguably the most important in the Romische Herrscherbild series (which currently numbers ten volumes), was long in the making. First conceived for the series by Max Wegner in the 1930s, a comprehensive study of the portraits of Augustus was originally to be published by Walter Gross. After the latter decided in the 1970s to focus only on the coin portraiture of Octavian/Augustus, Paul Zanker took over the project, [1] but in the end he passed it on with photograph documentation to Dietrich Boschung, who brought this magnum opus to completion within a remarkably short time.

The essential goals of any such modern iconographic portrait study are, first, to assemble all known portraits of a given personage; second, to determine the appearance and style of each of the presumed lost prototypes on which all of the known surviving replicas are based; third, to attempt to date the creation of the lost prototype and the surviving replicas and other portrait versions; and fourth, to try to determine the reason(s) for the creation of each type. Because no ancient author discusses the nature of portrait production, aside from some passing references and anecdotal comments, we must depend to a large degree on the evidence provided by the portraits themselves in addressing questions of the nature, ideology, replication, distribution, reception, and redefinition of an individual's portraiture. In Augustus's case, that body of evidence is substantial, numbering well over two hundred surviving sculptural portraits [2]--more than exist for any other Roman leader.

Boschung's primary focus in Die Bildnisse des Augustus is the creation of an elaborate taxonomical schema of Augustus's principal portrait types based on the extant portraits themselves and the rather limited literary and epigraphic evidence for his appearance. Although comprehensive, the present study is not all-inclusive. There is little discussion of the evidence provided by cameo and gemstone images of Augustus, [3] which was felt to be of marginal importance in establishing a portrait typology. Also omitted from discussion are possible images of Augustus in other media, especially vessels. [4] And because of the nature and goals of the Herrscherbild series, relatively little will be found in this volume with regard to the perceptual images of Octavian/Augustus or the psychological and sociopolitical needs that prompted their creation. [5] With regard to the numismatic evidence, it would have been helpful if coinage were treated in a more comprehensive way, even if that part of the study were written by another individual, as in the case of Boschung's volume on Caligula in the Herrscherbild series. As for the literary evidence for Augustus's physical appearance, it would have been more appropriately presented at the beginning of the book, rather than just before the catalogue.

Like others before him, Boschung accepts that there were three principal portrait types of Augustus (p1. 1.3-5; Figs. 3-5), to which he adds two earlier ones (p1. 1.1-2; Figs. 1, 2), with two subtypes (p1. 1.6-7; Figs. 6, 7). All of these could be employed with various body types representing him as imperator, priest, hero, divinity, or deified leader. Although there is general scholarly agreement as to the dating of one of the types (the so-called Prima Porta type), other matters are more problematic. Particularly difficult is establishing the earliest of Augustus's portrait types, as well as dating the prototype of the so-called Forbes type (after a head in the Boston Museum of Fine Arts), which Boschung rejects as the best replica of the lost prototype, preferring instead a head in the Louvre (his "Paris Louvre MA 1280" type). In Boschung's study, as in all such scholarly endeavors based in part on incomplete evidence and subjective interpretation, a number of points will continue to be contested and will need to be further clarified and modified in the future.

Before addressing the rather complex issues involved in establishing a portrait typology, I would like to offer a few words on the book's format. The three chapters after the introductory one constitute the core of the study: chapter 2 (pp. 11-50) deals with the establishment and categorization of the different types of Augustus's portraits; chapter 3 (pp. 51-65) reviews past scholarship on the successive types and their dates, together with Boschung's own conclusions; and chapter 4 (pp. 66-82) discusses the dating of the various individual images. The final two chapters deal with broader issues: chapter 5 (pp. 83-91) attempts to explain the distribution of Augustus's different portrait types, while chapter 6 (pp. 92-103) briefly discusses the copying of portraits, presents literary evidence for the appearance of Augustus, and gives a very useful thumbnail sketch of other issues pertaining to the images of Augustus, including various statue types, honorific inscriptions, and reasons for erecting images. [6] Boschung's discussion of Augustus's sculptural images in chapters 1 through 6 is followed by a catalogue of extant individual portraits, arranged according to types and, in some cases, subtypes. Under each catalogue entry, he gives basic information: museum, type of image, measurements, provenance (if known), condition, description, suggested dating, concise selective bibliography, and page references to the portrait in his text. With only two exceptions (cat. nos. 154, 166), he provides one or more illustrations of each of the portraits in his catalogue. In addition, Boschung presents a very short section on portraits of Augustus on several important cameos that represent him in frontal view with the Prima Porta hairstyle (pp. 194-95, cat. nos. 212-17). There are also brief catalogue entries of doubtful (pp. 196-97) and modern portraits (pp. 198- 201), as well as of those he takes as incorrectly identified as Augustus in the past (pp. 202-4)--by no means an all-inclusive list. After the catalogue of portrait s are several helpful line-drawn maps (pp. 206-13) showing the known provenance of portraits for each of Octavian/Augustus's portrait types. At the end of the study are three indexes (general, museum, and provenance). Besides the many photographic illustrations, a pocket attached to the back cover of the book contains useful foldouts (Beilage) with line drawings of key portrait heads (views of frontal, profile, and back of head) of the various portrait types. In these line drawings individual locks are selectively numbered to facilitate comparison.

The most important part of any typological study of this sort is the photographic documentation. Ideally, there should be a minimum of four views of each portrait (front, back, and both profiles), all shot at the same angle. Extremely desirable also is a photograph of each portrait from the optimum view; that is, the principal angle at which it was intended to be seen (often with face averted to the right or left). For a variety of reasons beyond the control of the portrait typologist, it is often not possible to obtain photographs of all these views, or even photographs of good quality, because of the inaccessibility of some images or the way in which portraits/portrait statues are displayed in museums and collections. Such qualifiers aside, Boschung should have obtained additional views or better photographs of a number of the portraits in his catalogue. Given the importance of Augustus to our understanding of Roman portraiture, the impact of his portraiture and portrait ideology on subsequent ages, and th e fact that this volume in the Herrscherbild series will remain the principal catalogue for some years to come, more of an effort should have been made to obtain the best possible photographic documentation. In a number of cases, Boschung uses photographs of plaster casts of extant portraits rather than of the original work itself. In certain instances, this might be understandable if a portrait is impossible to photograph because of its location in a modern setting, but not when there exist good-quality photograph of the original work, as in the case of a head of Octavian in the Stanza degli Imperatori in the Museo Capitolino in Rome: only photographs of a plaster cast are represented (pls. 14, 28.1), with no photograph of the optimum view of the original sculpture, even though excellent photographs of the original head are available. [7] When photographs of casts are used, the physical characteristics of the sculpture itself (such as restorations, breaks, discoloration) are difficult, and sometimes impossible, to detect.

In comparing photographs of different portraits and consulting Boschung's catalogue, I discovered in a few instances that the caption under the reproductions gave an incorrect catalogue number (for example, p1. 61 should be not cat. no. 65 but 51; p1. 65, not cat. no. 52 but 62; p1. 157, not cat. no. 100 but 95). In the citation of sources, more precise page references would have been preferable to the "if." typically used in German scholarship. Also, the inaccurate and anachronistic vocabulary of kingship or "emperorship" (for example, "Prinzenportrats") used to characterize Augustus, members of his family, and the form of government that he established should be given up. This sort of vocabulary (including, in English and American scholarship, the use of emperor and empress), which has been so prevalent, projects false notions onto the past, especially in terms of leadership and governance. Although Rome had acquired an empire (imperium) already under the Republic, Augustus was not an emperor, a word that, of course, derives from imperator but had a quite different meaning in antiquity. Augustus's civic position in the state was that of princeps ("first citizen" or "leader"), a term already in use under the Republic. The Roman historian Tacitus (Annales 1.9), writing in the 2nd century C.E., pointed out that Augustus had established neither a kingship nor a dictatorship but a principate (governance by a princeps): "Non regno tamen neque dictatura, sed principe nomine constitutam rem publicam."

In his "Introduction" (pp. 1-10), Boschung discusses some methodological and general issues regarding Roman portraits and their production. He sets up four general principles governing portrait studies: (1) "Konstituierung der Typen" (establishment of the types); (2) "Replikenrezension" (replica critique); (3) "Rekonstruierung des Entwurfs" (reconstruction of the [portrait] design); and (4) "Interpretation der Typen" (interpretation of the types). This methodological approach is a well-established one, based on a strong Germanic tradition going back to J. J. Bernoulli, who catalogued some ninety-seven heads of Augustus in his fundamental work Romische Ikonographie, vol. 2 (1886). In establishing a given type, this approach places a great deal of emphasis on the number, form, and arrangement of hair locks, especially (but not exclusively) over the forehead--the so-called Lockenzahlmethode (method of counting locks). To he sure, the Lockenzahlmethode is a useful diagnostic tool in establishing portrait types, b ut the almost all-consuming emphasis placed on it in many portraiture studies can lead to erroneous identification, as well as leave us at times wondering what is meant by a portrait. [8] Is a portrait a likeness of an individual or simply of a hairstyle (a "Portratfrisur")? It seems to me that an image of Augustus that does not reproduce one of his known iconographic hair types but closely resembles him in facial features might in some cases appropriately be designated an atypical portrait or a portrait with an atypical hairstyle [9] rather than simply excluded altogether as representing him. Such images have sometimes been too quickly dismissed as examples of Zeitgesicht (temporal visage), that is, a portrait of a private individual made to resemble the princeps or some other member of his family. [10] Conversely, how might we refer to an image of an individual who resembles the princeps in hairstyle but not in facial features? Such a hairstyle might be considered an example of Zeitfrisur (temporal hairstyl e), to coin a term. Providing an instance of such a Zeitfrisur is a colossal marble head of a mature bearded male from Rome in the Ny Carlsberg Glyptotek (Fig. 8), which is most likely an invented portrait of Augustus's legendary ancestor Aeneas. [11] This image can be identified as Aeneas because of its colossal size, beard, mature features, and, most important, the arrangement of locks both over the forehead in a mirror reverse of Augustus's Prima Porta type (cat. no. 171, pl. 69) and at the nape of the neck. This image, which stylistically appears to date from about the Hadrianic to early Antonine period, may have been based on an Augustan model.

And how are we to regard representations of Augustus that can only be identified by inscription? Among such images are those found on reliefs from Roman Egypt (not mentioned by Boschung) representing Augustus in a stereotypical Egyptian style. On the sides of a Temple of Augustus from Dendur now in the Metropolitan Museum of Art in New York (Fig. 9), [12] Augustus is represented as pharaoh, so identified by his cartouche. These representations might appropriately be regarded as symbolic images. The same might be said of certain statues in the round in pharaonic dress, which some scholars have identified as portraits of Augustus but which Boschung categorically dismisses as images of him (see, for example, cat. nos. [268.sup.*], [281.sup.*], [285.sup.*], [287.sup.*]).

In categorizing portraits that are identifiable as Augustus on the basis of hairstyle and, to a lesser extent, facial features, Boschung seeks ultimately to determine the appearance of the model for a particular type. Portraits that are of high quality, reflect the style of the city of Rome, and show a great degree of correspondence among themselves are considered to constitute replicas of a presumed lost model (Urbild) or, for Boschung, a lost design (Entwurf). Those few portraits that show the closest affinity to one another constitute Boschung's Kerngruppe (core group); others that show substantial affinity with one another form the remaining "replica series." The concept of a Kerngruppe works fairly well, but where to make a division between the Kerngruppe and the remaining less close versions of the type can be very subjective, as is also the matter of establishing what constitutes the best one or two examples within a Kerngruppe of the presumed lost original model. Within a given replica series there m ay be portraits that show an affinity among themselves but deviate from the core group in various ways. Such a group of portraits Boschung calls a Repliken-strang (replica string). Portraits that are of essentially the same portrait type but differ significantly from the core group could be regarded as Varianten (variants). Whether considered part of a replica string or a variant, some portraits appear to combine elements of more than one type and so constitute Klitterungen or Typenklitterungen (contaminations). The reasons for these deviations among the surviving portraits are varied and can only he postulated. Some portraits do not fit into any established type and should be considered independent creations. In his terminology, Boschung uses Kopie (copy) and Replik (replica) interchangeably, although others, from Georg Lippold on, [13] have attempted to differentiate between the two. Some distinction should at least be made between a replica (that which shows strong affinities with others of the same type) and a version (that which does not show strong affinities with a core group and can be considered an adaptation, variant, or new creation).

The first of the portrait types of Octavian/Augustus to be discussed (pp. 11-22, 59-65) is the so-called Actium or Octavian type, renamed by Boschung the Alcudia type after a head in Alcudia (Mallorca; Fig. 3). He sees four replicas (pp. 11-13) representing the Kerngruppe (cat. no. 6, pls. 7, 8; cat. no. 10, pl. 9; cat. no. 31, pl. 10; cat. no. 32, pl. 11), with another twenty-four also reflecting this type to a lesser degree. Boschung's analysis here illustrates one of the inherent problems of such typological studies, which make comparisons among extant heads as a means of reconstructing the lost prototype: namely, the self-limiting evidence of the portraits themselves and the quality and available angle of photographic views of each head. For example, within Boschung's Kerngruppe of the Alcudia type (pp. 10-13), the head in Alcudia is capite velatus (head veiled); only the top half of the Zurich head is preserved, and it is very weathered; and there are no strict profile views or back of the head shots fo r the Uffizi and Tripoli heads. Like Paul Zanker and Klaus Fittschen, [14] Boschung (pp. 52, 60-61) pushes the creation of this type (although not of the Alcudia head itself) back to at least 40 B.C.E. His dating is based largely on the coin evidence of Octavian's so-called DIVOS IVLIVS emission (pl. 238.2-3; Fig. 10, which might date anytime between about 40 and 38 B.C.E. (p. 60 n. 244). [15] Boschung agrees (p. 60 and n. 247) with those who see the Alcudia type as also reflected in later numismatic images on the so-called triumphal series (pl. 238.4-7).

Some have felt that the Alcudia type was Octavian's first three-dimensional portrait type and was the one used for the gilded equestrian image that the Senate set up in 43 B.C.E. to Octavian in Rostris [16] in the Forum. [17] But, as Boschung rightly notes, the earliest numismatic images of Octavian dating to this time look very generic. Since die engravers often show as much diversity in creating two-dimensional portrait images as do sculptors carving portraits in the round, such generic images would indicate that they were not based on any real portrait image of Octavian. In short, the earliest numismatic images of Octavian that reproduce the facial features and hairstyle of the Alcudia type appear to belong to the DIVOS IVLIVS issue, suggesting that die engravers for this issue were using as their model an image based on the Alcudia type.

In this type, a heightened sense of physiognomic realism is expressed in an artistic form that derives from the old so-called Hellenistic Pathosbild (an emotionally charged image). Reflective of the old pathos formula is the accentuated twist and inclination of the head, the plastically carved hair locks that appear somewhat agitated over the forehead, and the tension in the brows and forehead. This image of Octavian is, however, a far cry from the Roman Pathosbilder of earlier times. [18] In my opinion, the pathos has been toned down [19] in the Alcudia type and tempered by classicizing elements, especially evident in the surface treatment of the flesh and the more composed and lower-relief hair locks at the sides of the head. In this type we also find a stylistic range from a more academic classicism, as evidenced in a head from Ephesos in Selcuk (cat. no. 26, pl. 24.2-4), to a highly modeled and richly plastic treatment, as in a head in the Palazzo Bardini in Florenze (cat. no. 9, pl. 18.1-3). In the dati ng of individual portrait versions of the Alcudia type, Boschung tends to give weight to whether or not a particular work seems to have been influenced by the strongly classicizing style of the Prima Porta type, which most portrait specialists would see as created in or shortly after the founding of the principate in 27 B.C.E. In certain cases (for example, cat. no. 24, pl. 22, and even more in cat. no. 18, pl. 23), we can see the impact of the strongly classicizing Prima Porta type and pincer lock motif over the forehead.

After discussing his Alcudia type, Boschung (pp. 22-26) takes up the matter of the problematic portraiture of Octavian's earliest years. The old designation of Otto Brendel's Type A (represented by a famous head from Ostia in the Musei Vaticani [20] has long been rejected by Roman portrait specialists as an image of Octavian in his early teens and taken instead as a portrait of one of Augustus's adopted sons--a fact lost on modern historians, who continue to use this head in their historical treatments to illustrate what the young Augustus looked like. [21] Far more debated as being a portrait of the young Octavian is Brendel's Type B, which some have regarded as Octavian's earliest known type, predating the so-called Actium type (Boschung's Alcudia type). Boschung (pp. 51-52, 54-55) and others (myself included) [22] consider Brendel's Type B to be a portrait of Augustus's grandson and adopted son Gains. The facial hair, in the form of long side-whiskers and/or beard, evident in some of these portraits (for example, a portrait in the Galleria degli Uffizi [fig. 11]) and in a number of images of Octavian in his early coin types (for example, Fig. 10) has often been interpreted--incorrectly, in my opinion--as a Trauerbart (beard of mourning). [23] The long sideburns and/or narrow, neatly trimmed beard worn by Octavian in his early coin types (a detail that might also have been painted on some of his marble portraits) was most likely a military beard or "beard of vengeance" to evoke an image of a Roman Ares/Mars Ultor-like commander. [24]

Boschung postulates two new types as having been created before the Alcudia type. One of these he calls the Lucus Feroniae type (pp. 23-24, 59-62), after a head from Lucas Feroniae (cat. no. 4, pl. 5; Fig. 2), just outside Rome. This type, which he identifies in only two other replicas (cat. no. 3, pls. 4, 28.4; cat. no. 5, pl. 6), appears to be related to the Alcudia type. Although these three heads had previously been taken as versions of the Actium type (Boschung's Alcudia type), there are enough distinguishing features shared among them (and apart from other replicas of the Alcudia type) to establish this as a separate type or, at least, a subtype. Besides the tightly locked pincer effect over the forehead, these few portraits display a number of points of comparison in patterns of locks at the sides and back of the head (cat. no. 3, pl. 4; cat. no. 4, pl. 5). Most noteworthy are the very distinctive long, horizontal locks high above the right ear that form a double-stacked fan of down-turned locks. Clea rly different from the postulated prototype of the Alcudia type are the six to seven reverse-comma-shaped locks over the right temple in ail three portraits. These distinctive reverse-comma-shaped locks are also found on an early coin type of Octavian, with seemingly true portrait features, that Boschung correctly associates with his Lucus Feroniae type (pp. 24, 59-60). This numismatic issue, minted by Q. Voconius Vitulus (pl. 238.1), probably dates to the late 40s B.C.E. It is distinctly different from the DIVOS IVLIVS issue (Fig. 10), which, as we have seen, appears to be associated with the Alcudia type. Because of the problematic dating of both the DIVOS IVLIVS issue and that of Voconius Vitulus, it is difficult to know whether the Alcudia or the Lucus Feroniae type came first or to what extent the two may have overlapped in time. In any case, the Lucus Feroniae type appears to be either a distinct type that might have led directly to the Alcudia type, as Boschung postulates, or an early subtype of the Al cudia type that was very short-lived. As for the surviving replicas of the Lucus Feroniae type, Boschung dates all three to the time before the creation of the Prima Porta type. Of the three, the head from Lucus Feroniae, despite its summary carving and lower artistic quality, shows the strongest classicizing features, especially in the linear treatment of the hair.

In individual features and in the shape of the face, the Lucus Feroniae head (as well as others of that type) does not stand far from two other portraits that Boschung postulates belong to yet another early type of Octavian. This type, with physiognomic features rendered in a less realistic fashion than in the Lucas Feroniae type, Boschung calls the Beziers-Spoleto type (pp. 25-26, 59-62) after only two existing replicas: one from Spoleto in Perugia (cat. no. 1, pls. 1.1, 2; Fig. 1), the other from Beziers (ancient Baeterrae) in Toulouse (cat. no. 2, p1. 3). The two replicas are extremely close in formal details, including the number and arrangement of hair locks over the forehead, but differ somewhat in the treatment of hair locks at the sides of the head. Boschung believes that the Beziers Spoleto type, like the Lucus Feroniae type, was in existence in the period between 43 and 40 B.C.E., therefore predating the Alcudia type. He acknowledges that the period between 44 and 40 B.C.E. is a rather narrow range of time for the coexistence of three types and, further, that it is difficult to explain the relationship of the three earliest types, except to see the Beziers-Spoleto and Lucus Feroniae types as "experimental" in nature.

Although known in very few replicas, as we might expect for this early period and limited geographic area, the Beziers-Spoleto and Lucus Feroniae groupings appear to be genuine types rather than subtypes of the Alcudia type. The earliest numismatic evidence, which Boschung might have utilized to greater effect, seems to bear out his hypothesis. The first numismatic portraits of Octavian, dating to 43 B.C.E., present a classicizing image of a boyish youth, in some cases with longish side-whiskers. [25] This image is so generalized and so unlike his portraiture in the round that we must conclude that it is only a symbolic portrait for a coinage that was created in great haste to pay his troops. These and subsequent issues present images of Octavian that are classicizing to a varying degree. This preference for a classicizing style in numismatic imagery is expressed also in the strongly classicizing physiognomic features of the Beziers-Spoleto type.

Although not noted by Boschung, the images of Octavian on the coinage minted by L. Livineius Regulus in 42 B.C.E. [26] are among the earliest to reflect what appear to be his real portrait features. These coin likenesses compare fairly well with the portrait from Beziers (cat. no. 2, pl. 3). Even the fringe of locks over the forehead of the Beziers head, when viewed in profile, and of the numismatic images appears to be comparable. [27] As also in the coin portraits, the preserved part of the hair of the reworked Beziers head [28] is in low relief and is generally more composed than the thicker, plastically carved locks of the head from Spoleto. In this respect, the hair of the Beziers head may more closely reflect the lost prototype, which, based on the coinage, may in fact have been a more classicizing image than Boschung believes. In both the Beziers and Spoleto portraits, strong classicizing traits are manifest in the smooth, idealized structure and planes of the face, which is only somewhat averted, unl ike the more dramatically turned head of the Alcudia type. Evident at this time in the late Republic are both classicizing and nonclassicizing (baroque) tendencies, as well as a combination of the two, in keeping with late Hellenistic classicizing trends in Greco-Roman Ideal-skulptur (nonclassicizing) and adaptations and in Roman portraiture.

Interestingly, Octavian's own literary and rhetorical style, which most likely followed that of Caesar, was of a simple, classicizing type, or what might be called neo-Attic. [29] This was a Roman version of the late Hellenistic Attic style, which in turn looked back to late Classical models. This personal style of Octavian stood in stark contrast to the florid, more exuberant Asiatic style of his chief rival, Marc Antony. [30] In the late Republic, literary and rhetorical styles could be highly politicized and exemplify an individual's character and virtue. It was not until the founding of the principate that Classicism came to be the Zeitstil, the dominant style of the day, commonly referred to as Augustan Classicism. [31] Indicating the power of the princeps's stylistic imprimatur, even the polemical debates between the "Atticists" and "Asianists," which had so characterized the last days of the Republic, simply disappeared.

After discussing the earliest three types of Octavian/Augustus's portraiture, Boschung turns his attention to the old Forbes type, which he renames the Paris Louvre MA 1280 type (pp. 27-37, 60-65), after a head in the Musee du Louvre (cat. no. 44, pls. 36, 37; Fig. 4). This portrait he takes to he the best replica of his Kerngruppe, which includes also three other replicas (cat. no. 45, pl. 38; cat. no. 41, pl. 39; cat. no. 37, pl. 40). Of the thirty replicas in this group, he identifies fifteen as belonging to the lost prototype, with two Replikengruppen, or subtypes: one group of eleven replicas, of which the best representative is a head in Stuttgart (cat. no. 58, pls. 52, 53; Fig. 6); the other, a group of four replicas, best represented by a recut head in Copenhagen (Ny Carlsberg Glyptotek 611, cat. no. 60, pl. 64; Fig. 7). Boschung postulates (pp. 35-37) that the first of these subtypes (his Stuttgart replica group) was created under the influence of the Prima Porta type, while the other (Kopenhagen 61 1 replica group) would be a later (posthumous?) modification of the type. All of the replicas of the Kopenhagen 611 group are dated after the death of Augustus.

As in the case of Octavian's earliest portrait types, the dating of the original lost prototype behind the Forbes/Paris Louvre MA 1280 portrait has been much debated. The two principal dates proposed for the creation of the prototype are about 30-27 B.C.E. (therefore predating the creation of the Prima Porta type) and about 17 B.C.E., in connection with the Ludi Saeculares (Secular Games). Some have argued that the old Forbes type (Paris Louvre MA 1280) reflected the need for a new official Redaktion (updating/correction/edition) of the highly classicizing Prima Porta type because that type would have appeared to be too impersonal an image (p. 52 and n. 192). To support this claim, those who favor the 17 B.C.E. date cite the appearance of the Forbes type, with its distinctive lock formation, on denarii of L. Vinicius in 16 B.C.E. (pp. 60-61, pl. 239.2-3). The three-dimensional replica that best shows the configuration of locks over the forehead in the numismatic image is a labeled bronze bust of Augustus fro m Neuilly-le-Real in the Louvre (acc. no. N 3254, cat, no. 55, pl. 63), whose antiquity has been questioned in the past (incorrectly, I believe). A few have associated the creation of the Forbes type with the setting up of the Ara Pacis (13-9 B.C.E.), [32] on which Augustus is represented with a hairstyle associated with this type, and/or Augustus's becoming Pontifex Maximus (high priest of the Roman state religion) in 12 B.C.E. [33] For this type, Hans Jucker preferred the designation Ara Pacis type. [34]

For Boschung, the L. Vinicius issue reflects not the original prototype of his Paris Louvre MA 1280 type, which he dates to about 29-27, but rather an edited version of that prototype, which he sees reflected in a group formed around the portrait in Stuttgart (cat. no. 58, pls. 52, 53). Boschung (p. 63) sees the Stuttgart replica group, unlike the original prototype, as being influenced by the Prima Porta type. I am inclined to agree that the original prototype of Boschung's Paris Louvre MA 1280 type was probably created before the Prima Porta type, quite possibly in commemoration of Augustus's triple triumph in 29 B.C.E. The fact that it would have largely been replaced so soon after its creation by the Prima Porta type would also explain why it is hardly found outside Italy (p. 84). Had the principal type first been produced after the Prima Porta type (that is, about 17 B.C.E.), then we would probably expect to find far more replicas of the Paris Louvre MA 1280 type disseminated throughout the vast Roman Empire.

The last of Boschung's main types is the old Prima Porta type (pp. 38-59, 60-65), which takes its name from the famous statue of Augustus from Livia's country villa at Prima Porta, now in the Braccio Nuovo of the Musei Vaticani (cat. no. 171, pls. 69, 70; Fig. 5). Including the Prima Porta sculpture, a total of 148 replicas are catalogued by Boschung as belonging to this type, with another six cameo images of Augustus reflecting his Prima Porta hairstyle. The number of three-dimensional replicas and versions (adaptations, variants, new creations) of this type (more than doubles the number of surviving versions of all his other types) indicates its great popularity. Among the reasons for the success and endurance of the Prima Porta type are the simplicity and geometry of its very distinctive pincer and fork arrangement of hair locks over the forehead. Although the earliest known replica of the Prima Porta type is a bronze head from Meroe in the British Museum (cat. no. 122, p1. 195), which can be securely dat ed before 25 B.C.E., most portrait specialists see this type as having been created around the time of the founding of the principate in 27 B.C.E. This supposition is further confirmed by Ulrich Hausmann's association of the Prima Porta type with Augustus's numismatic images on cistophori from Pergamon that can be closely dated to 27-25 B.C.E. (pp. 53, 61 and n. 255, pl. 239.4). [35]

In his conclusion to chapter 3 (pp. 61-65), Boschung points out that unlike the Alcudia type, which looked back to the tradition of late Republican portraiture, the Prima Porta type conveys in its design (Entwurf) the physiognomy of the princeps in a classicizing Formensprache (language of forms), which depends especially on Polykleitan forms. By blending with prototypical forms of Greek art, the personality of the princeps, according to Boschung, retreated behind a Kunstfigur, which had to appear unassailable through its aesthetic qualities. Although the more intensive classicizing physiognomic features and ordered hairstyle of the Prima Porta type are indeed oriented toward the high Classical ideal, as expressed in the portrait image of Perikles, the end result points more in the direction of late Classical portraiture, which softens the frozen forms of the ideal Classical stereotype, permitting the personality of the individual to shine through the ideal type. Although an icon (in the modern sense of the word) is created in the process, this image of Augustus remains a portrait, while embodying at the same time a new concept of the heroic ideal. In my opinion, this new Augustan model stands, in a sense, midway between Greek high Classical and late Classical concepts of portraiture, even as the statue of Augustus from Prima Porta stands midway in its proportions between the old Polykleitan and Lysippan canons. [36] In attempting to challenge and outdo earlier prototypical concepts of the heroic ideal, the sculptor created the new Augustan heroic ideal. Although the complexity of these concepts would have been meaningful to only a small elite group, [37] the new ecumenical Augustan portrait image did capture the popular imagination, as attested by the great number of extant images of this type found throughout the empire. Like Octavian/Augustus's previous types, the official image was reinterpreted to varying degrees, with a wide range of stylistic treatments of facial features and hairstyles (compare, for exam ple, pls. 148, 149). And even though Augustus's earlier types continued to be reproduced, some examples of the older types show the influence of the new Prima Porta type, even to the extent of being Typenklitterungen (contaminated types). The continued production of earlier types after the creation of the Prima Porta type may have been related to the commemoration of earlier events from the princeps's life. [38]

Boschung's attempt in chapter 4 (pp. 66-82) to date individual surviving replicas and versions of Augustus's various portrait types constitutes one of the most problematic aspects of his work. Only a terminus post quem or a terminus ante quem can be established for most of the portraits. Contributing to the dating problem is the fact that Augustus, who was deified after his death, played an important role in the dynastic politics of subsequent principes and so his image continued to be reproduced during the principates of his successors. It is often extremely difficult, in any case, to date an individual work on stylistic grounds alone because of eclecticism and variability in workshop practices. Demonstrative of this problem is the case of the handsome portrait of Augustus from Ariccia in the Boston Museum of Fine Arts (cat. no. 80, pls. 119.3, 120), which has been dated from Angustan to Hadrianic times (Boschung places it in the period of Caligula). Similarly, a portrait of Augustus from Fondi in the Museo Nazionale di Napoli (cat. no. 16, p1. 26), combining a softened form of classicizing elements with lively, plastically carved hair locks, has been variously dated because of its stylistic treatment. Boschung (pp. 17, 61-62, 75, 83) classifies this work as a Neuschopfung (new creation) dependent on the Alcudia type (but probably even more so, in my opinion, on the Prima Porta type); he dates it, apparently correctly, to the Caligulan period because a pattern of hair locks behind the left ear compares fairly well with that found in some of Caligula's portraits. [39] Nevertheless, a strikingly close stylistic comparison of the general treatment of facial features and hair locks can also be made with Augustus's numismatic image on a cistophori series from Pergamon (27-25 B.C.E., p. 61 and n. 255, p1. 239.4), showing that such a portrait could also have existed some fifty years earlier. A great range of stylistic possibilities within a classicizing style (from a hard, cold, academic treatment to a softer, more mo deled and plastic one) can also be demonstrated within just the latter half of the Augustan Principate in the extant portraits of Augustus's grandsons and adopted sons Gaius and Lucius. [40] The vast majority of their portraits are datable to the latter part of the Augustan period because these youths were important only to Augustus's dynastic plans and played no significant role after his death in 14 C.E. Augustus's portraits can be dated more securely when they either bear a strong and intentional physiognomic resemblance to his successors or have been recut from portraits of his successors who suffered a "damnatio memoriae," or, more accurately, a memoria damnata (damned memory), after their death. [41]

In chapter 5 (pp. 83-91), Boschung discusses and attempts to explain the distribution (based on known provenance) of each of Augustus's different portrait types in four general regions: (1) Rome and Italy, (2) the western provinces, (3) the eastern Greek provinces, and (4) Egypt. Some of his reasoning is speculative. For example, Boschung assumes that unless carved in a distinctly local style, most of the marble portraits of the earlier iconographic types from the western provinces were imported from Rome or Italy. It is difficult to believe, however, that there were not also local sculptors (at least in the main Roman centers of the western provinces) capable of producing high-quality Rome-style portraits based on imported plaster or clay models. Some of these sculptors may even have originally come from Rome, elsewhere in Italy, or Greece. Although there was relatively little good-quality marble available locally in the western provinces, it does not mean that raw marble could not have been imported and ca rved in workshops in some of the larger western Roman centers. Are we to assume, for example, that the high-quality architectural sculptures of the Augustan Temple of Gaius and Lucius (Maison Carree) [42] were carved in Rome and shipped to Nimes?

According to Boschung, we do not find many examples of the Alcudia type in Italy, which was under Octavian's control in the Second Triumviral period, because such an emotionally charged image (Pathosbild) would have appeared "shocking" to the Italians after the civil war period and would therefore not have been acceptable to them. Boschung postulates that after the creation of the Prima Porta type, the Italians would have replaced these Pathosbild versions. This explanation, however, seems somewhat questionable because the removal of any images of Octavian after his victories at Actium and Alexandria (except by his order, as occurred in Rome [43]) might have been interpreted as an act of disloyalty or even treason. Although there was no law to the contrary, this did not prevent charges of treason (maiestas) from being brought against Granius Marcellus, the praetor of Bithynia, because of his imprudent act of replacing Augustus's portrait head on a statue with a head of Tiberius after the death of Augustus (T acitus, Annales 1.74).

To be sure, more replicas of the Alcudia type have been found in the western provinces than in Italy, but the provinces, too, were under Octavian's control during the Second Triumviral period. Boschung suggests that the inhabitants of the western provinces would not have been as "sensitive" as the Italians about the civil war period when the Alcudia type was in vogue. According to Boschung's hypothesis, there would have been no need, therefore, to replace later on the replicas of the Alcudia type in the western provinces. There may, however, be another explanation for the relative lack of portraits of the Alcudia type in Italy: a number of these portraits may have been destroyed in the factional strife in Italy during the civil war period.

In chapter 6 (pp. 92--103), Boschung briefly discusses the copying of portraits in the early principate, contrasting the copying then with that practiced in the late Republic. Although Boschung notes that Augustus was not the first to have his portrait copied, he makes no mention of the fact that portraits were not merely occasionally reproduced but already replicated in great numbers in the late Republic, as we know from the case of M. Marius Gratidianus, tribune of the Plebs in 87 B.C.E., whose image was set up in omnibus vicis (in all the districts) of the city (Cicero, De officiis 3.80; Pliny, Historia naturalis 34.27). By the Augustan period, the number of such districts (vici) was 265 (Pliny, 3.5.66). However, in the early principate, portrait images of Augustus were set up in great numbers not merely in the city of Rome but throughout the entire Roman Empire. Boschung cites the portrait head of Augustus from far-off Meroe in Nubia (presumably created a few years after the original prototype) as an exa mple of how fast and far the official image of Augustus spread. As Boschung and others have also pointed out, Augustus's official image was not only copied but also commonly altered and transformed in various parts of the empire for a variety of reasons, including the sociopolitical need to translate the Roman concept of the princeps into local or regional perceptions of leadership. [44]

Because any typological study of portraiture of necessity seeks to focus on the similarities among the extant portraits that establish them as replicas of the lost model, Boschung places relatively little emphasis on the question of diversity in so many images of Augustus. In a way, the matter of the great variety in imagery is more interesting than the similarity among replicas because of what it tells us about not only the various workshop practices and traditions current in antiquity but also the nature of ancient reception and response to images of the leader of state. As far as similarity is concerned, especially with regard to establishing the original lost prototype, there are relatively few replicas that make up Boschung's Kerngruppe.

Collected at the end of chapter 6 is the epigraphic and literary evidence for the physical appearance of Augustus. Boschung tries to discover in the extant sculptural portrait types various characteristics of Augustus's physiognomic features and hairstyle that are mentioned in the ancient written record. He also attempts to reconcile or explain discrepancies between the literary and visual evidence. Despite this necessary endeavor, Boschung, like many archaeologist/art historians, generally accepts the literary evidence at face value. However, the validity of this evidence is often questionable, since certain literary descriptions of physiognomy might be colored by the didactic nature of ancient rhetoric (especially epideictic) and its effects in such literary works as biography and history. Because the ancients believed that physiognomy could reveal the character and virtues of an individual, biographers and historians of the period manipulated their evidence as they saw lit. After all, the primary purpose of biography and history was not so much to tell the truth as to teach and to inculcate moral lessons and values. Accordingly, historians and biographers (most notoriously, Suetonius) used literary and oral accounts, as well as physioguomic "theory," in shaping their literary portraits of the great personages of the past.

Summation of Chronology of Augustus's Portrait Typology

I offer below some further thoughts on each of the portrait types, which are taken up in a hypothetical chronological order and with possible interpretative reasons offered for their creation. Roman numerals will be used to designate the different types rather than names that privilege individual works. Even the old, popular designation Prima Porta type is problematic because it takes its name from the statue of Augustus from Prima Porta, which itself dates no earlier than 20 B.C.E., [45] although the original model of this type was undoubtedly produced about 27 B.C.E. The old so-called Actium type (Boschung's Alcudia type) we now know, too, to have been created a number of years before the Battle of Actium in 31 B.C.E. The neutral designations Type I, Type II, and so on, also allow for the possibility of new portrait discoveries that might better represent a lost model than does the "named" portrait. Any subgroups or subtypes (as in Boschung's Stuttgart or Kopenhagen 611 series) can then be assigned a letter of the alphabet: hence, Type III.A, Type III.B. An "S" can also be assigned if there is only one subtype (for example, Type II.S) or a "V" if there is a variant of a given type (for example, II.V).

As possibly the earliest type, Type I (Boschung's Beziers-Spoleto type; Fig. 1) would presumably have been used for Augustus's gilded equestrian statue in Rostris in 43 B.C.E. The two portraits representing this type show him as very idealized and more youthful than either of his other two early types. Because the statue set up to Augustus by the Senate in Rostris was equestrian, it would have represented him as a military leader, or imperator. The title of imperator was the first to appear on his earliest coin type, which likewise shows the equestrian image. Type I might therefore be called secondarily his Imperator type. In 43 B.C.E., when this type was still current, Octavian became consul and, toward the end of the year, triumvir. Probably only two examples of Type I survive because it was so limited not only in time but also geographically, being found only in the west, Octavian's sphere of influence as triumvir. This type would have been created before the great impetus to copy portraits that came with the founding of the principate and Augustus's inauguration of peace throughout the entire Roman world.

my opinion, little thought went into the creation of Type I insofar as its social or ideological impact was concerned, mainly because it was created out of a sudden need, namely, for the gilded equestrian image to be set up in Rostris in 43 B.C.E. Shortly thereafter, this type was probably deemed unsatisfactory in that it represented Octavian, now imperator, consul, and triumvir, as too young and ineffectual-looking. Such a characterization, in fact, had been part of the anti-Octavian propaganda promoted by his rival Marc Antony and Antony's supporters. Antony's famous political barb, preserved by Cicero (Orationes Philippicae 13.25), "et te, o puer, qui omnia nomini debes" (And you, o boy, who owe everything to a name), must have been an exceedingly painful reminder of Octavian's political inferiority in that it implied he would have been nothing if not for Caesar's name.

In order to appear as triumvir and equal to Marc Antony, Octavian needed a more powerful image that conveyed to the Roman people that he was both the worthy heir of Caesar, who had become a state god (divas) in 42 B.C.E., and the leader of the Caesarean party in opposition to Antony. In Type II (Boschung's Lucus Feroniae type; Fig. 2), which might be considered secondarily Octavian's First Triumvir type, a more mature-looking and emotionally charged image of a leader was created through an increased use of classicizing elements. The portraitist started with the old Hellenistic concept of the Pathosbild, which had already been employed for powerful images of Roman leaders. For Octavian's new image, the old pathos formula with its baroque elements was toned down (Pathos-dampfung), as it was made to harmonize with classicizing tendencies. [46] Like Type I, Type II seems not to have been too successful, since both essentially experimental types were shortly replaced by Augustus's third type.

Type III (Boschung's Alcudia type, formerly the Actium/Octavian type; Fig. 3) was created as an even more evocative image that would both compensate for Octavian's youth and inexperience and better reflect his auctoritas. Type III may even have started out as a subtype of II but later replaced it as the main type. In Type III Octavian has a more simplified hairstyle with a reduction of the number of reverse-comma-shaped locks over the right temple and more agitated locks over the middle of the forehead. Although some have taken this as a reference to Alexander's famous upswept anastole, or upswelling, wave-like, hairdo, any such citation would only have been very indirect. The somewhat agitated locks were probably meant to portray Octavian as more of a man of action. This portrait became the image of choice for about a decade, lasting throughout the rest of the Triumviral period, for which reason it might also be called the Second Triumvir type.

Type IV (Boschung's Paris Louvre MA 1280 type, formerly the Forbes type; Fig. 4) logically followed Type III to satisfy a need for a new image of the leader after the end of the Civil War, a period that Octavian/Augustus wanted to put behind him. [47] Although not noted by Boschung, the pattern of locks at the back of the head of a Type III portrait in the Museo Capitolino in Rome (Stanza degli Imperatori, 2, cat. no. 23, pl. 14.2) is very similar to that of the Louvre MA 1280 head (cat. no. 4-4, pl. 37.2). This closeness helps establish a direct relationship between Types III and IV. This third type would have satisfied a need for a more mature image with a more classicizing, composed hairstyle, in preference to Octavian's more emotionally charged, though still somewhat classicizing image (Alcudia/Actium type), which had been so closely associated with the turmoil of the Second Triumviral period.

Type IV would have served not only to commemorate Augustus's triple triumph in 29 B.C.E., the year in which this type was most likely created, but also to celebrate the closing of the doors of Janus and the peace that Octavian had finally brought to the Roman world, an accomplishment of which he himself proudly boasts in his Res Gestae: "terra marique... parta victoriis pax" (Monumentum Ancyranum 13). Type IV might be called secondarily Augustus's "Triumphator" type. After his triple triumph in 29 B.C.E., he was never to celebrate another triumph, although he had the right to do so when those who served under his auspicia had successfully conducted a war for which a triumph could be voted by the Roman Senate. [48] An important head of Augustus in the Museo Capitolino reflecting this type (cat. no. 45, pl. 38) shows him wearing the corona civica with three gemstones, presumably one for each of his three victories. Even so, the prototype (Urbild) would have been created without the wreath, since any given type would have to serve for various kinds of images.

With the death of Antony and the end of the triumvirate, Octavian was in sole control of the government. He was now ready to turn his attention to stabilizing the political situation and creating a new constitution that would be acceptable to the majority of the Roman aristocrats. To celebrate the founding of a new form of government based on the principle of governance by a princeps, or "first citizen" as well as Octavian's assumption of his new name, "Augustus," with all of its sacral aura, a new ecumenical image was needed that was both retrospective and prospective: retrospective in that it invited comparison with the prototypical ideal of the Classical Greek past and prospective in that it reflected the optimism of the Augustan Principate and transformed Augustus into the new model of the heroic ideal. [49] Although Boschung sees this Type V (his Prima Porta type; Fig. 5) as becoming rather static and sterile as time went on, this view seems to me too modern. The very reason for the success of this type was its symbolic value: it became an icon for the stability and durability of the Augustan Principate. Because Type V initially seems to have celebrated Octavian's taking the new name Augustus and becoming princeps, it might be called secondarily the Princeps type.

Boschung argues convincingly that the prototype of his Stuttgart replica group (what I would call Type IV.A; Fig. 6) is represented on the coins of L. Vinicius in 17 B.C.E., the year of the Secular Games, with which this subtype may have been particularly associated. Type IV.A would represent a new redaction of Type IV under the influence of Type V, [50] which continued to be replicated in great numbers. For the image of Augustus on the Ara Pacis (cat. no. 56; pls. 59.1-2, 225.1), it is a subtype, Type IV.A, that is employed, rather than the earlier prototype, Type IV, as had previously been believed. [51] Boschung assumes that his Stuttgart replica group (IV.A) was employed on the Ara Pacis only because it was the latest official image of Augustus produced. Although this may have been the case, I believe there may also have been an ideological intent. To my mind, the use on the Ma Pacis of IV.A, representing a combination of Types IV and V, was intended to herald Augustus as the triumphant princeps who inau gurates a new golden age of peace. His triumphant return from Spain and Gaul was, after all, the original and official reason for the Senate's voting him the Ara Pacis (Monumentum Ancyranum, 12). This triumphal imagery would also fit the larger context of the Augustan monuments of the northern Campus Martius, especially if Augustus's great dynastic Mausoleum was crowned--as I believe it was--not with a statua pedestris (a statue representing an individual on foot) but with a quadrigate image of him (in a four-horse chariot) as triumphator perpetuus (perpetual triumpher). [52]

Type IV.A may also have been seen as particularly appropriate for representations of Augustus in an augural role. This subtype appears to have been used for the relief portrait of Augustus on the altar from Vicus Sandaliarius (cat. no. 36, pl. 67.4-5), [53] on which he appears holding the lituus (the crook-shaped staff of the augur) as he takes augury in connection with the departure of his adoptive son Gaius, who, under the auspices of Augustus (auspiciis Augusti), is about to set out on his eastern campaign in 2 B.C.E. [54] In the case of augury in relation to the Ara Pacis, I have argued elsewhere that Augustus was originally represented in the south processional frieze performing an augural act in connection with either the inauguration of the area on which the altar was to be built or possibly an augurium or maximum augurium salutis Rei Publicae, [55] which was performed for the safety of the state in years in which peace was renewed. An augurium salutis is known to have taken place in 29 B.C.E. and, gi ven the nature of this type of augury, it is reasonable to surmise that it was also performed in connection with the Secular Games in 17 B.C.E. (when IV.A appears on the coins of L. Vinicius), as well as on the occasion of the inauguration of the Ara Pacis in 13 B.C.E. [56] Augustus's appearance on the Altar of Peace in an augural capacity would have emphasized his role not only as inaugurator of a new golden age of peace and prosperity but also as mediator between gods and man. The form of the Ara Pacis, with its bifrontality and double set of doors, consciously recalled the Shrine of Janus Geminus in the Roman Forum, [57] whose doors had been closed to signify peace only twice in all of Roman history before 29 B.C.E. [58] Following the completion of the Ma Pacis in 9 B.C.E., it is likely that its doors would have been opened in the future when the doors of Janus were closed. The use of IV.A for a representation of Augustus in an augural role on the Ara Pacis and on the altar from Vicus Sandaliarius might ex plain why this subtype is found only in Rome and Italy (p. 84). [59] The Roman religious practice of augury was meaningful primarily in these areas.

Of the two subtypes of Type IV that Boschung identifies, IV.B (Boschung's Kopenhagen 611 replica group; Fig. 7) is the more puzzling. Because of points of comparisons shared among three of the replicas (cat. no. 60, pl. 64;