AIACNews 39-40

Re-use has been a central topic in archaeology and art history the last decades, especially regarding spolia in the architecture.[1] I have focused on re-use of sculpture, and re-carving of portraits in late Antiquity in particular. My hypothesis is that re-carving of portraits was common in late Antiquity; even more so than the making of new likenesses.

Firstly, I will give a definition on re-used and re-carved sculpture: Sculptures moved from their original position, and sculptures with another function or message than they were originally given, are re-used. The examples abound. They are, however, not always easily detectable since they are seldom found in their original context. Re-carved sculpture can most easily be distinguished as they often have physical traces from the operation, or remnants from previous phases. The original form is always altered.

Physical evidence of the re-carving of a portrait is normally visible by the ears, see fig. 1,

that could not easily be reduced in size, and by the hairline, where a marked line often indicates the removal of a layer. Sometimes the chin or the nose has become too small, and the curves around the mouth of the first portrait might have been difficult to conceal. A re-carved portrait can also be identified by the eyes, which tend to be enlarged and closer each other as the face has been diminished.

Fig. 1) Examples on re-carved ears. From right to left, top: private female portrait, ca. AD 250, Capitoline Museum. Detail from photo in Fittschen/Zanker III 1983, 113 no. 170 Taf. 197-199; private male portrait, AD 3^rd Cent., Torino Museum. Detail from photo in Cantino Watagin 1984, 232 fig. 3-6; private male portrait, beginning of AD 4^th Cent., Sperlonga Museum. Detail from photo in Bergmann 1977 pl. 43 no. 5-6. Bottom: private male portrait, AD 4^th Cent., Carettoni 1963, 159-160 pl. 58-59; private male portrait, AD 3^rd Cent., Bergmann 1977, 25 pl. 16 no. 2; Trajan (AD 96-116), re-carved from Domitian (AD 81-96), Archaeological Museum, Split. Photo by author.

Remnants from previous phases might be a curl at the back of the head, or, if it is a recognizable imperial portrait: the features of the original emperor. Examples are Nero's hair-style on Domitian's portraits, as seen in a portrait from Stuttgart[2], fig. 2, or Domitian's on Nervas, as in the Cancelleria-reliefs, fig. 3, one of the first Roman sculptures identified as re-carved[3]. Sometimes there are no recognizable features, but a stylistic discrepancy in the portrait that makes it possible to identify it as re-carved.

Fig. 2) Left: portrait of Domitian (AD 81-96) re-carved from portrait of Nero (AD 54-68), front; left: same portrait, right profile. In Stuttgart, Württembergische Landesmuseum, inv. no. 64/28. Photo from Bergmann/Zanker 1981, 356 fig. 28a-d.

Some portraits are re-carved more than once, like the head of an imperial statue in Jupiter-posture from Lucera[4], fig. 4.

In the end there was nothing left but the neck, and thus the last time it was reworked, the face was added in stucco.

Fig. 4 So-called Jupiter from Lucera. Unidentified emperor in Jupiter-positure, AD 1^st Cent. Re-carved several times. In Lucera Museum. Photo from Jucker 1981, 318-319.

Of course, portraits could be re-carved from other kinds of sculpture than portraits, and from architectonic elements, like a female head from Ny Carlsberg Glyptotek in Copenhagen, fig. 5.I have concentrated on portraits re-carved from portraits only, and portraits with the face in the same place as the original face, unlike for instance a female portrait which must have been re-carved in the back of a head with a long helmet like Minerva’s[5],

fig. 6, and a male portrait from Turin[6], where you can see the first persons ear at the back of the head, fig. 7.

Fig. 5) Female portrait re-carved from architectural element, AD middle of 3^rd Cent. In Ny Carlsberg Glyptotek, Copenhagen, inv.no. 1860. Photo from Johansen 1995 III, 206-207 fig. 91.

Fig. 7) Portrait of an unknown male, AD 3^rd Cent. In Turin, Museo di Antichità, inv. no. 138. Photo from Cantino Wathagin 1984, 232 fig. 3-6.

One of the best and most famous re-carvings from one face to another is the colossus of Constantine the Great from the Palazzo dei Conservatori, which is re-carved from Hadrian, fig. 8.

Fig. 8) The Colossus of Constantine (AD 306-337) re-carved from Hadrian (AD 117-138). In the Conservatory Palace, the cortile, inv. no. 1622. Photo from Fittschen/Zanker 1985, 147 cat. no. 122 pl151-152.

This was discovered by Cècil Evers in 1991[7], and corresponds with the retrospective sculpture program in the spolia at the Arch of Constantine, which is the prime example of re-used Roman sculpture. It has lately been claimed that the arch itself is Hadrianic[8], but irrespective of this, the tondi were re-carved in the Constantinian period, like in example, where Hadrian is re-carved into Constantine, fig. 9.

Fig. 9) Relief-portrait of Konstantine (AD 306-337) re-carved from Hadrian (AD 117-138), detail from tondo at the Arch of Constantine. Photo from L’Orange/Von Gerkan 1939 .

The portraits of Trajan, Hadrian, and Marcus Aurelius re-carved into Constantine at the Arch of Constantine, are famous examples on re-carved Roman portraits. This indicates the significance of re-carving of portraits in late Antiquity.

So far, the most important research that has been done on re-carved portraits, is Eric Varner's phD dissertation, Damnatio memoriae and Roman portrait sculpture[9]. The study concentrates on the re-carving of imperial portraits, and no later than Constantine.

To explain what was extraordinary about the re-carving of portraits in late Antiquity, I had to study re-carving of Roman portraits from the republican period till the time of Justinian I, who died in AD 565, as well as re-carved portraits in relation to the portraits that were made from scratch. The year of emperor Justinian I’s death, AD 565, was set as the conclusive year for my period of research since he was the last Roman emperor with an identified three-dimensional portrait known today. Even this is re-carved, based on the image of Lucius Verus, Marcus Aurelius' co-emperor from in AD 161-166, fig. 10[10].

Fig. 10) Portrait of Justinian I (AD 527- 565) re-carved from Lucius Verus (AD 161-169). In the Glyptotek, München. Photo from Aurea Roma 2000, 583-584, cat. no. 273.

The conclusion is that far more portraits were re-carved in late Antiquity than earlier. What we know as the late Antique portrait style, on which Hans Peter L'Orange is still considered an authority; this style, may in fact be a consequence of the re-carving.

The diagram in fig. 11, based on study of more than 2000 portraits, illustrates the amount of re-carved portraits in the three different periods in which the Roman history is normally divided: the imperial period from 31 BC to about AD 250, the time of the soldier-emperors ending with Diocletian's establishment of the Tetrarchy in AD 284, and late Antiquity from AD 284 to 565 AD.

Fig. 11) Diagram based on the study of more than 2000 Roman portraits, illustrating the amount of re-carved portraits in the three different periods in which the Roman history is normally divided: to the left, in gold, the imperial period from 31 BC to about AD 250; at the upper right, in violet, the time of the soldier emperors ending with Diocletian in AD 284, and the lower part to the right, in blue, the late Antiquity, AD 284-565.

I am aware of the possible faults in statistic investigations of archaeological material. Considering, however, the amount of portraits plotted in this diagram it is fair to assume that the result is indicative of a tendency.

The diagram indicates the number of re-carved portraits that are preserved from each period, but it tells nothing about how many they were compared to new-made portraits from the same periods. It is only when we know how large the total production of portraits was in each period, that we may say something about the ratio of re-carved portraits.

From the approximately 1500 portraits in the largest, published portrait-collections in Europe, the Capitoline Museum, Louvre, Ny Carlsberg Glyptotek and the Archaeological Museum in Istanbul[11], I have made a statistic table of the total of portraits that were re-carved compared to the amount of new-made images in each period. The result is illustrated in the graph in fig. 12.

Fig. 12) A statistic table illustrating the amount of re-carved portrait compared to the amount of new-made portraits in each period.

In spite of all the damnatio memoriae portraits, there are all together fewer re-carved portraits in the first period than in the second, compared to new-made ones, and most re-carved portraits appear in the third period, in late Antiquity. The change is remarkable and occurs in the second period, in the time of the soldier emperors. Why is this so?

There are several reasons for re-using sculpture. It is not only because of damnatio memoriae, nor is the answer to be found in the late Antique economy alone, as proposed by Horst Blanck in Wiederverwendung, still the most important monograph on the re-using of ancient sculpture, from 1969[12]. (Horst Blanck has told me that he as well is more moderate on that materialistic point today.)

Not even the retrospective politics can entirely explain the popularity of re-use. Sometimes sculptures were re-used for practical reasons, for conservation purposes, for aesthetic values, or for ideological reasons. The illustration in fig. 13 shows a syncretism of economy, politics and religion, factors that influenced one another reciprocally.

Fig. 13) Illustration of the syncretism of reasons for re-use of sculpture.

Re-use because of damnatio memoriae is clearly political. The politics and the religion of the period create the basic fundament for the attitudes, or the ideology, which again has an impact on the aesthetic values. Re-use of practical reasons has to do with economy, while re-use of sculpture in a new context is based on both economy and aesthetic values.

When it comes to portraits, there is another aspect: the ancient view of a portrait. It was not just a portrait; it contained an abstract form or archetype of the pictured[13]. This form is called vehicles by Methodius of Olymphos[14]. In German it is called Wesen[15], and Italian animo is suggested[16]. Already Heraclit wrote, around 500 BC, that the Greeks did not distinguish the gods from the statues of gods[17], and Origines wrote approximately the same 700 years later[18]. Other sources, like Suetonius[19], tell that it was the same with the emperor and imperial portraits, and although there are few sources on the form in private portraits, it seems to have been equally imperative.

It is mainly because of this form that portraits of emperors culprited damnatio memoriae were destroyed by law; the emperor's form could continue to act through the portrait. This corresponds with when the emperor's portrait presided courtrooms as a substitute for the emperor, who could not attend every case. The portrait represented him, and secured the justice.

Imperial portraits contained an imperial form that could not be mixed with portraits of normal individuals. This must be the reason why there are no imperial portraits identified as re-carved into private persons, but there are also other groups that do not interfere with each other, like female and male portraits. No rule without exception; a few portraits have been re-carved into the opposite sex[20].

There were certain rules for what could be re-carved and what could not, and until the middle of the 3^rd Century AD, there was no example of portraits re-carved from a popular emperor. Altering imperial portraits not exposed to damnatio memoriae could lead to consequences. However, as the graph indicates, something happened in the 3^rd Century. Now let s take a closer look at the events that led to this.

The change started in the Severian period, that is AD 193-235. Eric Varner calls the Severian Dynasty "a critical juncture in the history of damnatio memoriae"[21]. Until then, certain emperors had suffered damnatio memoriae, and their portraits had been re-carved. In the Severian period, there was an accumulation of culprits like Geta, Plautilla, Plautus, Julia Mamea, Elagabal and Alexander Severus. Given the large number of damnatio memoriae, the verdict lost its significance. There was no attempt in trying to conceal that portraits were removed, and there were few re-carvings. Examples of the few re-carvings from this period are the four known Elagabal-portraits re-carved into his cousin Alexander Severus[22]. One of them, from Ostia, now in Palazzo Massimo, is depicted in fig. 14.

Fig. 14) Portrait of Alexander Severus (AD 211-218) re-carved from a portrait of Elagabal (AD 218-222). In Museo Nazionale Romano, inv. no. 329. Photo from Sapelli 1998, 24 fig. 5.

Alexander Severus wanted to be identified with so-called good emperors before Elagabal, but his likeness with his cousin made it possible to produce portraits quickly and at low expense, something that was necessary in the turbulent circumstances in the period.

The Severian period anticipated the time of the soldier emperors in destroying unwanted imperial portraits rather than re-working them. At the same time it became more usual to re-carve private portraits. In the time of the soldier emperors, the population of the Roman Empire declined by almost 50% as a result of numerous wars and plagues[23]. The availability of portraits was highly increased. This, combined with the recession and lack of artists, formed the background for the re-carving of portraits, and most importantly, the view on mankind was altered during these hard years, also referred to as “the military anarchy”. The individual was no longer held in high esteem, and the understanding of portraits as vehicles for the abstract form was changed. The form of the portraits, previously treated with respect, lost its importance as the view of mankind changed, and portraits could be re-carved on a grand scale.

A lot has been written about the style of the 3^rd century portraits, which to most students is coherent with the short-cut soldier male portrait type. Alois Riegl led later scientists to believe that the decline of style begun with the making of the portrait of Decius from around AD 250, Andreas Alföldy introduced the term "Gallienian classicism", Eugenie Strong followed by others claimed that portraits of the soldier emperors had a realism that released them from the dynastic style of the earlier portraits, Gerhardt Rodenwaldt wrote that the style of the soldier emperors showed the real Roman realism, and Hans Peter L'Orange claimed that the portrait style of the 3^rd Century AD was experimenting, and that deeper knowledge could only be reached by dating the portraits before the theories were established[24]. This led to the idea that Roman portrait-style history became more complex in the 3^rd Century. As it is now clear that re-carving of portraits increased remarkably in the 3^rd Century AD, it might be easier to understand why the analysis of style appeared more complicated to the above-mentioned scholars.

Few new imperial portraits were made in the time of the soldier emperors; but only a small number of sculptures of any genre were executed in this period. It is symptomatic in periods of wars and conflicts that scarce attention is paid to art and culture due to financial priorities as well as the obvious lack of means. In particular few new imperial portraits were made. The emperors changed rapidly, and inherited portraits from each other. The respect for the emperors and their imperial form decreased.

Jale Inan and Eliasbeth Rosenbaum remarked, when investigating Roman portraits in Asia Minor for the first volume of the catalogue of portraits in The Archeological Museum in Istanbul, that among the material there were no imperial portraits from the era between Caracalla and Diocletian, that is the 3^rd Century AD[25]. Meanwhile, this period is particularly rich in portraits of private individuals in Asia Minor. Clearly, portraits of most of the soldier-emperors exist, but not nearly as many as those of the previous emperors.

Speaking about the soldier-emperors and their portrait-style, one needs to bear in mind that there was another trend in this period, namely that of the philosopher. In contrast to the short-cut hairstyle of the soldier, the philosopher had long hair and beard, inspired by the old classical philosophers. This trend seems to have been geographically limited to Greek areas, Athens in particular. The portraits were re-carved from the bearded portraits from the 2^nd Century AD. Among these are found the portraits re-carved in the shortest span of time after the execution of the originals. Some of the portraits of the cosmets were made only few years before they were re-carved, like the one illustrated in fig. 15, and shortly after the re-carving, they were re-used as building-material in the walls around the Agora.

Fig. 15) Portrait of a cosmet, AD ca. 250. In the National Museum, Athens, inv. no. 327. Photo from Lattanzi 1968, 53-54 pl. 20a.

The cosmets were a particular kind of official teachers or scholars of philosophy in Athens[26]. They were originally, in Hellenistic time, teaching Aristotelian philosophy. In the 3^rd Century AD, when the number of pupils was at its greatest with an average of 300 per year, the institution had become more universal, like a university.

The activity of the cosmets, and the production of cosmet-portraits and the re-carving of them, probably reached its end by the invasion of the Herules in AD 267, when Gallienus was emperor.

More portraits have come down to us of Gallienus than of any other 3^rd Century emperor. Gallienus was first emperor with his father Valerian from AD 253 to 260, and alone from AD 260 to 268. The particular stylistic features in the portraits of Gallienus, the "Gallienian classicism", have been investigated by several scholars.

I argue here that the main reason why the portraits of Gallienus are special is that they are re-carved. As Marianne Bergmann has proved, the Gallienian coins are retrospective, referring to Alexander the Great and Augustus[27], but Gallienus did not refer to them, only. There are examples of sculptural portraits of young Gallienus re-carved from portraits of Augustus, see fig. 16,

Fig. 16) Portrait of young Gallienus (AD 253-268), re-carved from Julio-Claudian emperor, beginning of AD 1^st Cent. In the Pergamon Museum, Berlin. Photo from Bergmann 1977, 127 pl. 12 fig. 5-6.

and later portraits of Gallienus that are re-carved from Hadrian, see fig. 17. In my opinion the examples are so many it cannot be mere coincidence, and I thus come to the conclusion that Gallienus introduced the retrospective re-use adapted by Constantine on his arch, and in his sculpture program in general.

Fig. 17) Portrait of Gallienus (AD 253-268) re-carved from Hadrian (AD 117-138). In Museo Torlonia, inv. no. 603. Photo from L’Orange 1933 fig. 9-10.

Investigations of Diocletians Arcus Novus and the Arch of Constantine has led to an interpretation of retrospective, propagandistic sculpture programs as a typical late Antique phenomenon[28]. The re-carving of the portraits of Domitian in the Cancelleria-reliefs is a result of damnatio memoriae, made possible by the motifs, that could be transferred to later historical situations. The surviving examples indicate that re-carving of imperial reliefs without damnatio memoriae begun with Gallienus. On the Arco di Portogallo Gallienus re-used reliefs from Hadrian and the Antonines[29]. The most famous is a Hadrianic panel, in which the head was re-carved into Gallienus, fig. 18. Due to modern restorations this is no longer easily visible[30].

From Gallienus onwards re-carving of all kinds of portraits seems to have been accepted, although it was still uncommon to re-carve imperial portraits into private and vice versa. Apart from the belief in the abstract form of the portrait, this might also be due to the fact that most imperial portraits were exhibited in public places, belonging to the society as a whole. If portraits were re-carved, this was the result of consensus.

In Tetrarchic times porphyry was popular in imperial portraiture, but among the marble portraits, there are several examples of re-carving from earlier portraits, like the so-called Licinius in Side, Asia Minor, re-carved from Commodus, fig. 19.

Fig. 19) Late antique imperial portrait, possibly of Licinius (AD 308-324) re-carved from Commodus (AD 180-192). In Side Museum, inv. no. 35. Photo from Inan/Rosenbaum 1966, 86-87 cat. no. 63 pl. 40 fig. 1-2, pl. 41 fig. 3.

Like the soldier-emperors, the tetrachs had short hair, and the heads are cubic or ovoid, thus enabling the artist to reduce the volume with ease. The tetrarchic portrait style is technically easy to re-carve, and one may therefore ask what came first - the style itself or the re-carvings?

Until Constantinian times the techniques for re-carving were not up to the standards of the re-carved damnatio memoriae-portraits in the 1^st and 2^nd Century AD, but at the beginning of the 4^th Century AD, the quality reaches new levels. An example from Museo Torlonia is a Constantine, re-carved from Hadrian, fig. 20, like the famous colossus in Palazzo dei Conservatori, fig. 7. This is a highquality re-working.

Fig. 20) Portrait of Constantine (AD 306-337) re-carved from Hadrian (AD 117-138). In Museo Torlonia, inv. no. 619. Photo from Evers 1991, 806 fig. 9a-b.

The quality of re-carved portraits leads us to the second part of this paper, concerning the methods for re-carving.

The earliest method was to make some adjustments in the face, as seen on the re-carved Ptolemeian head in fig. 21.

The Romans might have been inspired by the Ptolemeian workshops, but the difference is that the Ptolemeians had no real portraits, but idealizing representations, like the Greeks. The Greeks were satisfied by changing the base-inscription when re-using statues[31], while the Ptolemeians made some physical changes in the portrait[32]. The invention of individual, realizing portraits is normally attributed to Roman artists, thus the methods of re-carving from one individual into another must also be regarded Roman.

Fig. 21) Ptolemeus IX (88-80 BC) re-carved from Ptolemeus X (107-88 BC). Probably from Alexandria. In J. Paul Getty Museum, Los Angeles, inv. no. 83 AA 330. Photo from Varner et a. 2000: 92-95 cat. no. 3.

In early imperial times the normal method was to remove a layer from the face, like this sketch in fig. 22 illustrates. The sketch is hypothetical, based on a damnatio memoriae-portrait of Caligula re-carved into Augustus.

Fig. 22a) A part of the face is removed. Fig. 22b) The back of the head is removed.

Another method, which seems to have been more usual in late Antiquity, was to remove the back part of the head, and turn the angle of the face slightly upwards. In this way the artist could avoid the problem with extra protruding ears when the face was diminished.

Re-carved portraits, the late Antique in particular, constitute a heterogeneous group of sculptures which cannot easily be systematized. Given the regional differences in Roman portraiture, two main categories remain: Western and Eastern portraits. The West is mainly represented by today’s Italy; the East for the most by Greece and Turkey. Even though the capital was moved to Constantinople in late Antiquity, it seems reasonable to claim that Western portraits were more urbane, while the Eastern have more provincial features. The Western and Eastern portraits can be divided into subcategories, such as Hispanic, Syrian etc.

To be able to say something on workshops and origins, I have chosen to focus on the portraits from Asia Minor. When studying re-carved late Antique portraits, I found that the collection in the Archaeological Museum of Istanbul was the most representative. In the Western museums the portraits were mostly collected in a period when classisistic aesthetic values were paid more attention than today's search for a realistic impression. Besides, the provenance of the monuments in Istanbul is mostly known, which is important in trying to understand different regional workshops and techniques.

Based on the portraits in Istanbul I have introduced some categories based on the method of execution, called the "bandit" -type, the "helmet"-type, the "Efesos-tetrarch"-type, the "Athens-type" and the "Eutropius"-type.

The "bandit"-type is recognized on the shape of the heads, fig. 23. These portraits have no protruding facial features; it is like if the individuals represented have stockings over their heads, like bandits. These compact heads are ovoid, and the nose and the ears are in low profile, almost not breaking the surface. The profile is shortened, the hairline high, and the facial features are centralized by the root of the nose. The mouth is vaguely smiling, giving a gentle look. The eyes are large and almond-shaped, with bags underneath. The pupils are bean-shaped, and the gaze looks upwards. It is as if the original portrait is removed at the upper part of the head only, and evened out downwards. The head is inclined towards the back.

Fig. 23) Examples of re-carved portraits of the ‘bandit’-type. From left to right: 1) Portrait of unknown male, tetrarcic period, re-carved from unidentified head. In Efesos, Selcuk Museum, inv. no. 1675. Photo from Inan/Rosenbaum 1979, 184 no. 148 pl. 119, fig. 1-2; 2) Tetrarcic male head, re-carved from unidentified head, from Alba Fucens, in the National Museum in Chieti. Photo from L’Orange 1984 pl. 8 fig. c-d; 3) So-called Licinius (AD 308-324), re-carved from unidentified head. In Vienna, Arthistorical Museum, inv. no. 932. Photo from Inan/Rosenbaum 1966, 149-150, cat. no. 191, pl. 175 fig. 1-2.

The "bandit"-type is dated to the tetrarchic period. Regarding one of the examples, the one known as Licinius from Efesos, it is said that in late Constantinian times some portraits were made to be reminiscent of the Tetrarchic ones. I suggest here that these portraits were re-carved in Constantinian times, from Tetrarchic portraits.

The "helmet"-type, fig. 24, is characterized by giving the hair a pronounced line along the forehead. The edges are sharp, as if they were cut into metal. These portraits have hairstyles reminiscent of helmets, thus the name. The ears are cut in low relief. Both the front and the back of the head are shortened, and the eyes are deep in the surface.

Fig. 24) The ‘helmet’-type From left to right: 1) tetrarchic male portrait, re-carved from portrait from the period between Nero (AD 54-68) and Hadrian (AD 117-138). In the Vatican Museum. Photo from Blanck 1969, 54-55 no. A32 pl. 19 fig. b; 2) Portrait of Galerius (AD 293-311) re-carved from unidentified head. From Africa. In Museo Nazionale Romano, inv. no. 4289. Photo from L’Orange 1984 pl. 22 fig. c-d; 3) Late antique male portrait, re-carved from unidentified head, found in late antique wall at the Palatine. In the Palatine Museum. Photo from Carettoni 1963, 159-160, pl. 58-59.

The "Efesos-tetrach"-type is named after the example in fig. 25, which is one of the best.

Fig. 25) The ‘Efesos-tetrarch’-type. From left to right: 1) The ‘Efesos-tetrarch’, portratit re-carved from a Gallienian head (AD 253-268), from Efeos. In Selcuk Museum, inv. no. 720. Photo from Inan/Rosenbaum 1966, 144-145, cat. no. 183 pl. 107 fig. 1-3; 2) Male portrait, tetrarchic, with the inscription ‘M. Sylvius Otho Imp.’, re-carved from earlier portrait, probably from the AD 3^rd Cent. In the Münchener Residenz, inv. no. 41. Photo from Hojer 1987, 187 no. 147; 3) Private late antique male portrait from the tetrarchic period, re-carved from an unidentified head. In Skopje, Archaeological Museum. Photo from Bergmann 1977 pl. 48 fig. 1-2.

It slightly resembles the "helmet"-type, with sharp edges between the hair and the face, but here the edge continues around the chin. The ears protrude from the even hairstyle. It is a high quality technique, but few examples remain, and those that do are presumably from Efesos or Side.

The "Athens"-type is recognized by the fact the portraits in this group are re-carved from portraits with rich hairstyles and beards, that is portraits from the later half of the 2^nd Century AD, fig. 26. Most of these portraits are from Athens, hence the name. The ears are normally untouched, protruding from the new face, while the hairstyles are cut down in a shorter fashion. The face is deep in the profile, but the broad forehead is most particular. These portraits are examples on how the eyes could be stretched between the ears and the root of the nose when re-carved. It is a good technique, but the portraits in this group are normally of a lower quality.

Fig. 26) The ‘Athens’-type. From left to right: 1) Private male portrait, late antique period, re-carved from unidentified head. In Athens, the National Museum. Photo from L’Orange 1933, 124, cat. no. 59 fig. 110-112; 2) Private male portrait, late antique period, re-carved from unidentified head. In Athens, the National Museum. Photo from L’Orange 1933, 126 cat. no. 67 fig. 118-119; 3) Private male portrait, late antique period, re-carved from unidentified head. In Athen, the National Museum. Photo from L’Orange 1933, 126 cat. no. 66, fig. 116-117.

The last group of re-carved late Antique portraits that I want to present is the Eutropius-group, which already had its name when M. Aurehammer wrote an article on them in 1988[33]. The most famous portrait in this group, and the one for which they are named, is a 5^th Century portrait, previously attributed to the court-eunuch and writer Eutropius at the time of emperor Valens, AD 364-378, fig. 27. Beside the problem with the dating, this bearded portrait can not represent an eunuch, but it keeps the name.

Fig. 27) The ‘Eutropius’-type. From left to right: 1) Late antique male portrait, so-called ‘Eutropius’, AD ca. 450. Re-carved from earlier portrait, most probable a AD 2^nd Century portrait. From Efesos. In Vienna, Arthistorical Museum, inv. no. I 880. Photo from Aurea Roma 2000, 647 cat. no. 355; 2) Late antique male portrait, AD ca. 450, re-carved from earlier head. From Efesos. In Bonn, Akademische Kunstmuseum, inv. no. 295. Photo from Inan/Rosenbaum 1979, 305-306, cat. no. 304 pl. 259 fig. 1-4; 3) Late antique male portrait, AD ca. 450, re-carved from earlier head. From Efesos. In Selcuk Museum. Photo from Aurenhammer 1984, 129-130 fig. 6.

The portraits in the Eutropius-group are long and manieristic, yet not approaching the exaggerated features of the Eutropius-portrait. They are normally high quality re-carvings that can best be identified by the angle and the size of the ears.

The portraits have beards, but they must have been made from portraits with even longer beards, that is portraits from the last half of the 2^nd Century AD, or more probably, as the drill-holes are not that deep, from philosopher-portraits from the 3^rd Century AD. The Eutropius-portraits have long faces because the new chins are where the beard used to be. The faces are slender because the original ears and the hair-locks with deeper drill-holes than the beard had to be removed. The eyes are wide, stretched upwards and outwards to fill the broad space that consequently is a result of this method. It is significant for the 5^th Century AD that the eyes are large.

Of the eight "bandit"-portraits I have found, one is in Chieti[34], three are in Athens[35], two are in Ankara[36], two from Efesos[37], and one, a female portrait in Vienna, is supposedly from Asia Minor as well[38]. There are furthermore six stylistically related portraits in Italy[39], and one in Copenhagen[40]. The Eastern ones have more ovoid heads than the Western ones, and larger eyes. The artists may have used the same methods, but it is easiest to understand how the re-carving is executed from studying the Eastern "bandit"-portraits.

The hypothetical sketch in fig. 28 illustrates re-carving with the back part of the original head removed. The inner portrait is of the "bandit"-type, the outer, black silhouette is from a Greek 3^rd Century portrait. A smaller head, like one of the soldier-emperor style, could have been chosen; the technique does not require very much material, but with this example the operation is more clearly illustrated. The difference between the size of the outer and the inner portrait is realistic, but the sketch is hypothetical.

Fig. 28) Left: hypothetical sketch of re-carving with the back of the original head removed. Right: hypothetical illustration indicating the facial zones in which the new features had to be cut deepest into the original portrait: the eye-area in particular.

The red lines indicate the later head. It is mainly regarding the eyes that it has been necessary to remove material. The late Antique portrait-style did not require large eyes only; they were also supposed to gaze upwards. That is essential to the method: the head is bent backwards to create space for the new eyes, and at the same time keep as much as possibly of the original volume. The ears remain in their original position, but normally adjusted to the new angle. This was done by removing the outer ear-shell.

In fig. 29 are both portraits illustrated on top of each other, in the middle the former portrait, and to the right, the result of the re-carving, a "bandit"-portrait. By changing the angle of the head the artists could exploit the volume to a maximum, while the vertical neck remains the same, giving the "bandit"-portraits their peg-like look: the head appears to be an expansion at the top of the neck. Amanda Claridge’s drawing of heads from the 1^st and 2^nd Century AD illustrates this difference[41], fig. 30. I have put a "bandit"-head beside her sketch, to illustrate the difference.

Fig. 29) Hypothetical sketch of original and re-carved portrait on the top of each other to the left, the original portrait in the middle and the re-carved to the right. The angle of the head had to be changed to exploit the volume to a maximum.

Fig. 30) Amanda Claridge’s sketch of heads from AD 1^st and 2^nd Century to the left, from Claridge 1988, 135-162, and the authors sketch of a ‘bandit’-type portrait to the right, to illustrate the difference of the angle of the head.

With the Eutropius-type a new technique was developed: the upper part of the face was re-carved by the same principals as for the "bandit"-group, and the ears could remain in their original position, or be moved inwards, but to alter the neck was a new invention, fig. 31. By removing a little bit more material from the top of the head and the forehead, and from the chin, and by removing more material from the lower part of the back of the head, it is possible to obtain a form that is no longer peg-like. In comparison with the “bandit”-type more volume was removed, but the quality is just as good, if not better. The size of the portrait becomes smaller; the so-called Eutropius measures 29.5 cm. only, while the example I have used for the "bandit"-portrait, Licinius from Ephesos, measures 32.0 cm. The 3^rd Century head used to illustrate the first phase measures 34. 2 cm.

Fig. 31) Hypothetical illustration of the re-carving after the ‘Eutropius’-type method. The silhouette of a 3^rd-century philosopher portrait forms the original head. The back of the head, the beard and the upper part of the face were removed. The angle of the neck was altered. The ears could remain in their original position.

The angle of the neck in the “Eutropius”-portrait gives an individual leaning more forwards, but observed from the front this is without significance. The gaze is no longer turned upwards. As the chin is where there originally was a beard, the upper part of the ears are more or less untouched, while the lower part of the ears are adjusted to the new portrait.

The Eutropius-portrait might originally have been a Zenon. The length and the form of the head attributed to Zenon in Ny Carlsberg Glyptotek in Copenhagen, see fig. 32, can be compared to that of “Eutropius”. The sketch in fig. 33 indicates the relationship between the portraits of Zenon and “Eutropius”.

Fig. 32) Left: a portrait-herm attributed to Zenon in Ny Carlberg Glyptotek, Copenhagen, front and left profile. Photo from Poulsen 1954 cat. no. 72. Right: ‘Eutropius’, right profile.

Fig. 33) Hypothetical sketch of the portrait of Zenon in fig. 32 and ‘Eutropius’ at the top of each other.

The "Eutropius"-type is about 150 years later than the "bandit"-portraits, and it might be too daring to claim that there is a technical development from the "bandit"-portraits to the "Eutropius"-portraits. On the other hand, such a process cannot be excluded. We know for instance that The council of Ephesos already at the time of Marcus Aurelius and Lucius Verus in the 2^nd Century AD wanted to re-work bronze-statues of past emperors[42], something which we have already seen earlier only in cases of damnatio memoriae. The council was not allowed to do that, but this information indicates the re-carving traditions in Ephesos. Ephesos was a sculpture center, and methods for re-carving might very well have been developed there.

There is no proof that the technique behind the "bandit"-type was developed in Ephesos; such portraits are, as we have seen, found at different places, but this method might have been developed further in Ephesos, into the "Eutropius"-type. Lines drawn from the wrinkles above the eyes and lines from the nostrils to the corners of the mouth in both examples form a cross by the root of the nose, fig. 34. The cross can be called the center of the face. In the “Eutropius”-portrait this center is moved upwards; two thirds above the chin.

Fig. 34) Hypothetical sketch illustrating the centre in the face at the ‘bandit’-portraits, left, and the ‘Eutropius’-portraits, right.

During this work it has been useful to study unfinished Roman portraits, of which I have found ten pieces, showing different stages in the process. Fig. 35 illustrates an unfinished Trajan from the Capitoline Museum[43]. The short distance between the ears and the back of the head makes it likely that this is a portrait of Trajan about to be re-carved into Constantine.

Fig. 35) Unfinished portrait of Trajan (AD 98-117). Photos from Fittschen/Zanker 1985, 42 cat. no. 43 pl. 48

A close-up at the chins and noses in portraits of Trajan[44], the unfinished Trajan, and Constantine[45] in fig. 36 strengthens this hypothesis. The Trajan-portrait is to the left, the unfinished Capitoline portrait is in the middle, and the Constantine-portrait is to the right. There is a certain resemblance. Furthermore, the eyes of the unfinished portrait seem to have been enlarged, fig. 37.

Fig. 36) From left to right: 1) Portrait of Trajan (AD 98-117), decennalie-typus. Photos from Fittschen/Zanker 1985, 42-43 cat. no. 42, pl. 45-47; 2) same as in fig. 35; 3) The Constantine colossus, same as in fig. 8.

Could it be that what we recognize as the typical late Antique portrait-style is a consequence of re-carving, like we know is the case in one of the most famous examples, the colossus of Constantine, fig. 7?

I have shown how Gallienus introduced re-use of sculptures from popular emperors, and that the "Gallienian classicism" was a result of re-carving. It is not impossible that the same was the case with the "Theodosian classicism". The classicising Theodosian portraits might very well be influenced by the fact that several of the portraits used for re-carving were Julio-Claudian.

What I have not mentioned, is the re-carving of female portraits in particular, which has been claimed by at least one scholar to be an updating of the portraits as the women changed their hairstyles[46]. Considering the amount of re-carved portraits in general, it seems likely that the re-carved female portraits also represent new individuals in the second phase. The reason why the faces were normally not changed, only the hairstyles, is that the female portraits were idealizing, without individual facial features. The portrait in fig. 38 is from early Gallienian times, now in the Capitoline Museum[47]. It is in fact the single re-carved private female portrait from the 3^rd Century that I have found. The double Scheitzelzopf shows that it was re-carved short after the original execution.

Fig. 38) Female portrait from the early Gallienian times, AD ca. 253-260. Re-carved short after the production of the original portrait, due to the double Scheitzelzopf. Photo from Fittschen/Zanker 1983, 113 cat. no. 170, pl. 197 fig. 199.

Another aspect, that I will leave for another occasion, is the re-carving of portraits in Early-Christian times. We know that the Christians discussed whether it was accepted to bath in termae where pagan sculpture were exhibited, because of the form held by the sculptures[48]. We also know from Paulinus of Nola that it was discussed how Christians should regard portraits[49]. The increased number of re-carvings in the 5^th Century AD compared to the declined productions of new ones, could indicate that the belief in the portraits form disappeared, but still under Justinian I it was possible to seek protection in imperial portraits by the ad statuam confugere - principle[50]. And as proved by André Grabar, the significance of the imperial portrait increased in late Antiquity, guarantying the true value on wages and documents[51]. The individual facial features of the emperor were less important. The inscriptions under the portraits were exchanged as new emperors followed, recalling of the Greeks and their exchange of inscriptions at the bases of the idealizing statues.

[1] Aalto Bauer 1990; Alchermes 1994; Brandenburg 1987; Brenk 1987; Brilliant 1982; Deichmann 1975; Delachenal 1995; Esch 1960; Fabrisius Hansen 2001; Gandolfo 1988; Kinney 1997; Krautheimer 1961; Lindros-Wohl 2001; L’Orange/Von Gerkan 1939; L’Orange for example 1961, 1963, 1972; Manderscheid 1981; Marvin 1983; Mc Cann 1972; Meier 2001; Merrimann 1985; Pensabene 1992; Settis 1986; Ward-Perkins 1964. For re-use of sculpture see Bergmann/Zanker 1981; Blanck 1969; Curran 1994; Dwyer 1982; Elsner 1998; Ensoli 1997; Goette/Hitzl 1987; Harrison 1990; Hannestad 1994; Jucker 1986; Pandermalis 1972; Pekàry 1985; Pollini 1984; Varner 1983, ed. 2000. See for example also: Amedich 1991; Andrea 1980; Bergmann 1977, 1999; Berczelly 1996; Brinkerhoff 1970; de Kersauson 1986, 1996; Fittschen/Zanker 1983, 1985; Inan/Rosenbaum 1966, 1979; Johansen 1995; Koch 1993; Mango 1963; Sande 1991; Sarardi-Mendelovici 1991.

[2] Bergmann/Zanker 1981, 356 fig. 29 a-d; Varner 1993, 120 pl. 145-146.

[3] Magi 1945; Mc Cann 1972, 251-255; Kleiner 1992, 191-192 fig. 158-159.

[4] Jucker 1981, 318-319.

[5] Varner ed. 2000, 220-223, cat. 58.

[6] Cantino Watagin 1984, 232 fig. 3-6.

[7] Evers 1991, 805 fig. 5, 7.

[8] Conforto 2001, 22-58. See also Barcelo 1991-1992, 123-131; 1992, 105-114.

[9] Varner 1993.

[10] Bergmann 2000,273, 583-584; Sande 1992, 599-607.

[11] Fittschen/Zanker I, III 1981, 1983; De Kersauson 1986, 1996; Inan/Rosenbaum 1966, 1979, Johansen 1983.

[12] Blanck 1969.

[13] Prusac 2002, 155-162.

[14] Sande 1991, 76; Kitzinger 1954, 91 nr. 20.

[15] Pekàry 1985, 132-136.

[16] Cambi 2000, title.

[17] Heracl. frg. 5 Diels.

[18] c. Cels. 8. 41, c. Cels. 8. 38 f. See also J. Geffcken ARW 19, 1916-19, 286 ff.

[19] Suet. Tib. 58. See also Grabar 1968, 64; Sande 1993, 77; SHA Caracalla 5. 7-8.

[20] 1) Portrait of a young female from the beginning of the 3^rd Century re-carved from Antinous. In Ny Carlsberg Glyptotek, Copenhagen, inv. no. 3286. Johansen III 1995, 200-201, cat. 88; Bergman 1981, 189; Fittschen/Zanker I 1985 additional note to cat. no. 55; Blanck 1969, 45 no. A18 Taf. 14-15. 2) Portrait of a Gallienian priest re-carved from a Trajanic female. Formerly in Brescia Museum, present place unknown. Matheson in Varner, ed. 2000, 77 fig. 10a-d; Blanck 1969, 53 no. A30 Taf. 20-21.3) Portrait of a young early Severian female re-carved from a Hadrianic male. Palazzo dei Conservatori, inv. no. 2746. Fittschen/Zanker III 1983, 103-104 no. 152 Taf. 180. 4) Constantinian male portrait re-carved from a female portrait, possibly Julio-Claudian. Side Museum inv. no. 116. Prusac 2002, 219 no. 384; Inan/Rosenbaum 1966, 200-201 no. 276 pl. 153 fig. 1-2.

[21] Varner 1993, 343.

[22] 1) In Palazzo Massimo delle Terme, inv. no. 329. Varner 1993, 410 pl. 419-420; Sapelli 1998, 24 fig. 5; Felletti Mai 1953, 138 no. 273 a-b. 2) In the Vatican Museum, Sala dei busti, inv. no. 632. Varner 1993, 410 pl. 421. 3) In Museo Nazionale, Naples, inv. no. 5993. Varner 1993, 412-413 pl. 423-425. 4) In The Nelson-Atkins museum if art, Kansas City, Missouri, inv. no. 45. 66. Varner, ed. 2000, 200-203, cat. 52.

[23] See for example Magie 1982, 27.

[24] Bergmann 1977, 6 for further references.

[25] Inan/Rosenbaum 1966, 1.

[26] For cosmets, see Lattanzi 1968.

[27] Bergmann 1980, 26-27.

[28] See for example Kinney 1997, 117-148.

[29] Kinney 1997, 134.

[30] Kleiner 1992, 255, fig. 222.

[31] Blanck 1969, 65-84.

[32] See for example Varner, ed. 2000, 86-87, cat. 1.

[33] Aurenhammer 1988, 123-131.

[34] L’Orange 1984 Taf. 8 c-d; photo from La Rocca, E. in Aurea Roma 2000, 23 cat. 20.

[35] 1) Cosmet. Photo from Lattanzi 1968, 63-64 Tav. 32 a-b. 2) “Vitellus”. Photo from Stavridis 1982, 457-458 Tav. 143-144. 3) Private male portrait. Photo from Bergmann 1977 pl. 47 fig. 3-4.

[36] 1) A private male portrait. Photo from Inan/Rosenbaum 1966, 212 cat. 294 pl. 121 no. 1-2. 2) A private male portrait. Photo from Inan/Rosenbaum 1966, 214-215 cat. 298 pl. 168 fig. 3-4.

[37] 1) “Licinius”. Photo from Inan/Rosenbaum 1966, 149-150 cat. 191 pl. 175 fig. 1-2; Photo from M. Bergmann in Aurea Roma 2000, 22-23 fig. 22; 2) Private male portrait. Photo from Inan/Rosenbaum 1979, 184 cat. 148 pl. 119 fig. 1-2.

[38] Photo from Inan/Rosenbaum 1979, 168 cat. 121, pl. 100 fig. 1-3.

[39] 1)Licinius, Vatican Museum, Museo Chiaramonti. Photo from L’Orange 1984 pl. 30 fig. a-b; 2) Licinius, Ostia Museum. Photo from L’Orange 1984 pl. 31 fig. a-b; 3) Private male portrait, Vatican Museum, Sala 10, inv. no. 635. Photo from Giuliano 1967, 71 cat. 86 fig. a-b, pl. 52. 4) Private male portrait, Archaeological Museum, Milano. Photo from Bergmann 1977 pl. 45 fig. 5. 5) Private male portrait, Museo Nazionale delle Terme, magazine. Photo from L’Orange 1933, 140 cat. 113 fig. 214-215. 6) Private male portrait, Museo Nazionale delle Terme. Photo from Bergmann 1977, 141 pl. 40 fig. 2.

[40] Private male portrait, Ny Carlsberg Glyptotek, Copenhagen, inv. no. 1667. Photo from Johansen III 1995, 184-185 cat. 81; L’Orange 1933, 114-115 cat. 25 fig. 66, 69.

[41] Claridge 1988, 135-162.

[42] Elsner 1998, 71.

[43] In The Capitoline Museum. Fittschen/Zanker I 1985, 42 cat. 43 pl. 48.

[44] The example presented here is from The Capitoline Museum. Of the Decennalium-typus. Photos from Fittschen/Zanker I 1985, 42-43 cat. 43.

[45] The Constantine-colossus, Palazzo dei Conservatori, inv. no. 1622. Photos from Fittschen/Zanker I 1985, 147 cat. 122 pl. 151-152; see slso Evers 1991, 805 fig. 5, 7.

[46] Herrmann 1991: 35-50.

[47] In The Capitoline Museum, Salone 38 inv. no. 339. Fittschen/Zanker III 1983, 113 cat. 170 pl. 197, 199.

[48] Augustin Brev. XLVI, XV-XVI.

[49] Paulinus of Nola Carmen 27. 511-515; Elsner 1995, 249.

[50] Sande 1993, 79; Kitzinger 1954, 122-123.

[51] Grabar 1968, 64.