Version classiqueVersion mobile

Les arts de la couleur en Grèce ancienne… et ailleurs

Philippe Jockey

Arts polychromes et dorés, synthèses et études de cas

The Acropolis Kore 682 from Athens. An Approach to the Reconstruction of a Greek Late Archaic Sculpture of a Girl

La korè 682 de l’Acropole d’Athènes. Essai de reconstitution d’une statue de jeune fille grecque de la fin de l’époque archaïque

Ελληνική πολυχρωμία, υστεροαρχαϊκή αττική γλυπτική, κόρη με διαγώνιο χιτώνα

Bernhard Schmaltz


Pour se faire une idée de la polychromie ancienne, les éléments les plus significatifs dont nous disposions sont les restes de la sculpture colorée mis au jour sur l’Acropole d’Athènes dans les années 1880. À l’époque, existaient encore souvent des couleurs brillantes, à présent largement perdues. D’après des aquarelles contemporaines et un moulage daté de 1910 et coloré peu après à partir des aquarelles et des traces de polychromie autrefois encore visibles, la korè la mieux préservée était Acro 682. Ce moulage, conservé au musée national d’Archéologie à Athènes, a été nettoyé en 2005 et utilisé pour un moulage ultérieur, qui nous informe amplement sur la décoration. Nous avons utilisé ce nouveau moulage pour compléter les parties manquantes comme les boucles cassées, les bras, les mains et les plis. En examinant tous les vestiges de la décoration sur l’original, sur le moulage et sur les aquarelles il est possible de reconstituer un exemple d’un type significatif de la sculpture tardive archaïque.

Texte intégral

1Looking back at the past, we may consider the problems that we Europeans had and still have when thinking about painted ancient sculpture. When, in the 15th century, ancient sculpture was rediscovered, generally in the form of rather broken and weathered statues, people were nevertheless fascinated by the marble material, the high standards of workmanship in carving, and the vivid representations of the human body, so often presented naked. So, as no colour was preserved, nobody entertained the possibility of coloured sculpture at all, and all admired the shining white marble surface, which in its uniformity shows so clearly the modelling of bodies and clothes.

  • 1 I. Kader, “Täuschende Spielereien”, in V. Brinkmann, R. Wünsche, Bunte Götter : die Farbigkeit anti (...)

2By the beginning of the 19th century we learned that ancient architecture was, to a large extent, decorated in vivid colours – we may remember Haller von Hallerstein’s 1811 drawing of the ornaments of the temple at Aegina with written reference to the preserved colours, or, for example, G. Semper’s reconstruction of the Parthenon from 1836 – but it was hard to believe that the delicately modelled surface of sculptures was painted and coated by a second skin of its own aesthetic value. Only in the course of the excavations on the Athenian Acropolis, which started in 1885, when some dozen broken marble statues with ample remains of brilliant colours were unearthed, did people come to believe that ancient marble sculpture was indeed painted. But very soon the archaeologists on the Acropolis realized that the newly excavated marbles lost their brilliant painted decoration within a few hours, owing to the effects of sunshine and wind. So they asked artists such as L. É. Gilliéron to produce watercolours of the most impressive pieces in order to preserve at least a faint idea of the Greek painted marbles.1 About twenty years later, when most of the colours had vanished, permission was given to produce casts of eleven of these statues and, shortly afterwards, five of these casts were painted, probably by Gilliéron’s son. His aim, however, was not to try a full reconstruction of what had existed two and a half millennia before, but to reconstruct the post-excavation state : the colours as they had been when the broken statues were unearthed. Looking back now, we clearly realize that at the very same time European art had discovered the aesthetic charm of the fragment, the fascination of the torso – we may recall artists such as Maillol and Rodin – and for this reason the archaeologists declined to restore the figure and the painted decoration in order to get an idea of what had been set up on the Acropolis in Archaic times.

  • 2 V. Brinkmann, Die Polychromie der archaischen und frühklassischen Skulptur (2003), on which see the (...)

3A mere half century later, new interest in antique polychromy arose, which initially focused on the material of the colours itself and the technique used to paint the marble. Following this line, V. Brinkmann studied the remnants of colours by means of new photographic techniques, revealed and recorded all these remains, and created some reconstructions as well, which are well known from his exhibition “Bunte Götter”.2 But he was primarily concerned with the character of the design and the variation of the different colours, and paid very little attention to one aspect, the interdependence of shape, design and colour. It is this point in which I am interested, and for this purpose a reconstruction that is as neat as possible is essential.

Fig. 1 — Kore Acro.682 (around 1910.)

(Reproduction of the photograph Alinari 24644.)

  • 3 AD I (1889), pl. 39.

4Among the eleven casts produced in 1911 is that of the kore Acro.682. This figure was nearly complete (the feet were added by H. Schrader in 1907 : fig. 1) and in 1886 so much of the painted decoration had remained that the statue was praised as the best preserved kore of all, and a watercolour of it was published in a distinguished publication as early as 1891 (fig. 2).3 Since then it has often been cited as giving a good impression of Archaic polychromy. For this figure Gilliéron’s son produced a kind of copy. Looking to the original and to the watercolours of his father, he transferred the decoration, the motifs, the design and the colours to the cast, and so created a kind of three-dimensional copy of this broken statue of a girl dressed in a typical diagonal mantle.

5This coloured cast was stored in the National Museum at Athens, cleaned in 2005 and used for a further mould and a further cast ; this new cast proved to be most informative. The unpainted white surface shows a lot of the incised ornaments of the garment borders including the battlement, the small stars of pointed leaves and the dots along the incised meander ; furthermore, some colours, especially blue, have oxidized in the form of thin crusts, which are reproduced by the cast as a very low relief (fig. 3). All these details, which are hard to recognise on the painted surface of the coloured cast, originate from the original statue of 1911. By means of these details it was possible to reconstruct directly the system of the decoration and its exact location.

  • 4 Karakasi 2001, pll. 144ff.
  • 5 Ibid., pll. 254ff., “Antenor-K.”, p. 264.

6It was reasonable to use this cast for a reconstruction. To achieve this, it was necessary to look for parallels and to consider the lines of the drapery, the layers of the plaits, and the swelling of the surface. For example, to supply the left hand of the kore, we had to consider two different postures shown by other statues. Some korai have the left hand clasped into a fist, and the thumb lies at the side of the forefinger, seizing at the same time the tip of the garment, so that the back of the hand is turning away from the figure’s side.4 Other korai have the back of the hand turned to the front, and three extended digits together seize the tip of the drapery, the forefinger and the middle finger from the front and the thumb from the back.5 On the kore Acro.682, the fracture near the wrist clearly shows that the back of the hand was turned to the front, so that the attitude was rather similar to that of the “Antenor-kore”. As in the lower part of the figure, the large border with the magnificent meander turns into a broad curve to the side without becoming narrower, and the outer contour of the vertical border cannot be reconciled with the seizing fingers, but has to stand rather outside the figure. The result is that the contour of the garment on the top of the vertical border must have turned to the fingers in a way that seems rather strange to our understanding of natural textiles.

Fig. 2 — Watercolour of part of kore Acro.682.

(L. É. Gilliéron ; after Antike Denkmäler I [1889], pl. 39.)

Fig. 3 — Detail of the cast of Acro.682 in Kiel, Antikensammlung. Upper part of the central border below the waist.

(Photo : B. Schmaltz.)

  • 6 Karakasi 2001, pl. 146.
  • 7 D. Ohly, Die Aeginete II / III (2001), pll. 201ff.

7Aside from some other details of the reconstruction, one characteristic should be discussed first, because of its importance for the shape of the statue. To her right side, the himation hangs freely in long, zigzag-lined folds, the shorter one ending near the right knee, the other below the middle of the calf. Large parts of these folds are restored. But looking to the back of the figure it is possible to follow the zigzag for some way from one layer of the garment to the other,6 and from the step of the lowest and longest zigzag the length of the following three zigzags and of the total fold can be roughly calculated. A comparative example is offered by a similar marble kore from Aegina,7 which has even longer folds. It is obvious that such rich drapery decorated with a colourful border would have intensified the splendour of the figure and modulated the formal concept of the statue.

8But now let us proceed to the most interesting step of the reconstruction, the painting of the decoration. As mentioned above, immediately after the excavation of the kore much of the decoration was preserved, although to a large extent it has vanished since then. But, as we have seen, we could restore most of the patterns by following the traces on the cast itself. Only the lowest border of the garment above the feet and the diagonal border crossing the breast are painted without using the evidence of the preserved statue itself, but are rather borrowed from similar statues or fragments. So the reconstruction of the statue itself may at least somewhat resemble the original, and this applies to the system of decoration too.

  • 8 See P. Wolters, AD I (1889), text to pl. 39 ; W. Lermann, Altgriechische Plastik (1907), pp. 87ff.

9The colours reported at the time of excavation were red, blue and green, with the latter two generally not preserved in their original tone, but oxidized.8 As in the case of other korai from the Acropolis and elsewhere, the colours were mineral colours, i.e. cinnabar, malachite and azure or Egyptian blue. What is not known is the matrix that was used to mix the powdery colours to make them fit for brushing and to fix them to the marble surface. Essentially there are two possibilities, an organic matrix or an inorganic one. As organic material, egg, casein or gum arabic, for example, most frequently in some combination, could have been used. In a few cases, an inorganic material has been identified : a calcareous matrix such as lime-water, usually mixed with organic materials like casein. The organic material is water-soluble, as tests have shown, so that after six weeks of rain and sunshine (of course in the rainy northern part of Germany) the tested colour had obviously vanished. On the contrary, the inorganic matrix was so solid that after even fourteen weeks the colour was still in fairly good condition. Another effect of the lime-water was that it made the colour lighter, i.e. red became somewhat pink. To compensate for this effect we had to add more pigments, but this rendered the brushing of the material more difficult. It also has to be remembered that, for example, cinnabar was rather expensive in Antiquity.

10Of course, colour gradation is not only affected by the matrix, but also by the size of the pigment-crystals and the size of the grains ; the former depends on regional circumstances, the latter on the workshop that worked up the pigments. So, without serial analyses of samples from the korai of the Acropolis, it is impossible to get an idea of how the colours of our kore really looked.

  • 9 V. Brinkmann, Die Polychromie der archaischen und frühklassischen Skulptur (2003), fig. 107.
  • 10 T. Karagiorga-Stathakopoulou, AD 51/52 (1996/1997), A, pll. 2ff.

11Two more factors should be considered now, because they affect the tone of the colours and their stability. As our experiments have shown, the same matrix mixed with different colours reacted differently to sunshine and rain. Therefore we might suppose that, in order to achieve maximum stability for the overall decoration in Antiquity, different colours were mixed with different matrices – varying either in concentration only or in composition of the ingredients. But in fact we do not currently know either the components of the matrix or the mixture, and this fact may be illustrated by two heads. In the one case,9 cinnabar has been used for the lips, iris and hair, but here the same colour has vanished almost totally ; in the second case,10 the mineral hematite is wonderfully preserved on the hair, eye and eyebrows, but no colour at all has been preserved on the lips, although cinnabar, which might have been used, is said to be more resistant. There is no way of explaining the current state of these two heads.

12Finally, in the course of our experiments we learned that there is a significant difference between applying colour to a marble slab purchased in a shop and to a marble carved by a chisel, as in Antiquity. In the first case, the marble is sawn and its surface is rather sealed off, while in the other case each strike of the chisel damaged the crystals and caused very thin fissures, so that the colour cinnabar could be applied easily as one nearly homogeneous coating that adhered very well. For azure the result was not so good, but six weeks later the same colour could be brushed very easily, nearly as easily as cinnabar previously. Again, the result affects the colour gradation.

13To sum up : even if we do not know the exact tone of the original colours of the painted kore Acro.682, we have plenty of other evidence available, such as the motifs and composition of the ornaments, the basic colours, and their distribution. So it seems appropriate to try a reconstruction as a working hypothesis (fig. 4).

Fig. 4 — Reconstruction of kore Acro.682 (Kiel, Antikensammlung), front view.

(Photo : B. Schmaltz.)

Fig. 5 — Reconstruction of kore Acro.682 (Kiel, Antikensammlung), view upright to the feet.

(Photo : B. Schmaltz.)

Fig. 6 — Reconstruction of kore Acro.682 (Kiel, Antikensammlung), view facing the breast.

(Photo : B. Schmaltz.)

Fig. 7 — Reconstruction of kore Acro.682 (Kiel, Antikensammlung), rear.

(Photo : B. Schmaltz.)

  • 11 G. Dickins, Catalogue of the Acropolis Museum 1 (1912), pp. 40ff.
  • 12 D. Brooke, “Terracottas”, in S. Casson (ed.), Catalogue of the Acropolis Museum 2 (1921), p. 337.

14It is obvious, I think, that the painted decoration, the brightness of the colours painted on the white marble-ground, the variety of the ornaments, and the composition of the continuous borders and the scattered small motifs all intensify the splendour of the drapery and enhance the representation of the young lady, creating a kind of “epiphany”. At the same time, it is striking that from the foot up to the diadem two colours prevail on the white marble, red and blue, the latter with the addition of its “neighbouring colour” green. That colour chart predominates on the entire figure, even on the hair and the eyes : as is the case with about 80 % of all painted heads of korai from the Acropolis, both hair and iris are red,11 which to us seems to be rather unnatural. Interestingly, a multitude of small terracottas of the same type and from the same excavation all have black hair and black eyes, reinforcing the point that the choice of colour was not a matter of naturalism, but a convention of art :12 using only three colours contributed to the unity of the statue.

15Furthermore, the splendour of the garment is now so dominant that the human body beneath it is almost totally hidden. It can be imagined, of course, that the breast, chest, waist and hip are below the drapery, but the natural forms of all these parts are not visible. Only the legs from below the knees can be seen. This lack of interest in the human body concerns the concept of the figure too, because there is no central line reflecting the natural symmetry of the human body. Predominant, on the contrary, are the splendour of the garments and of the jewellery, of the colours and of the variety of ornaments, and the vivacity of the oblique or diagonal borders. With that in mind, we may remember what is said in the Homeric Hymn to Aphrodite about the goddess of beauty and love (hymn. Ven. 61ff.), around one hundred years prior to this statue’s creation. Aphrodite fell in love with Anchises and, to bewitch him, she went to Cyprus, where the Graces beautified her. They washed the goddess and applied ambrosial salve to her, and then beautiful garments and golden adornments. The shape and the appearance of the goddess of beauty, of her body itself, which might be of interest to us, are not referred to at all. The text seems to reflect a conception of female beauty quite different from ours, but rather near to that of the reconstructed statue from the Acropolis.

16But let us consider once more the conception of the reconstructed statue. Standing directly in front of the feet (fig. 5), we realize that the upper part of the body is turned a little to its left side ; we can estimate the turn by concentrating on the coloured curls and their symmetrical design ; and going further on, so that we look vertically at the breast (fig. 6), we become aware that the head itself counteracts to a slight extent this turn. So it is obvious that the entire figure is twisted, turned around its longitudinal axis, from the feet up to the shoulders, and the head responds with a partial reversion.

17When we take the middle view as the main view of the figure (fig. 4), facing the head directly, we notice that the upper part of the figure rotates a little to the side of the hand seizing the drapery, whereas the lower part rotates clearly to the other side. At the same time we observe that, just at that side, to which the part of the body is turned, the body reveals its distinct natural roundness. The chiton’s blue borders lead around the calf, and similarly the painted border at the left shoulder marks the roundness of the arm and of the shoulder. On the contrary, the other side is defined by a closing contour, and no line and no form indicate the three-dimensional character of the sculpture. By these differences in design on each side the torsion of the figure is intensified, which could perhaps be seen as a manifestation of vitality. This impression is intensified when we look at the figure a little from below, as it certainly stood on a rather high base (the position of the camera was fixed at the level of the sandals, which carefully treated presumably should have been to be seen).

18This view shows that it is obvious that the statue rises as an inseparable unity of body and garment, rising on both feet without any reference to gravity ; it rises from a very narrow basis and expands like a slim wedge to the square, somewhat masculine shoulders, framed in the lower part by the verticals of the paryphe and the long zigzag folds. At the same time the statue seems to lift the chest in a kind of breath, or as a reflection of energy and vitality – without any visible support from the physical axis of the body or a symmetrical composition of the garment.

19Archaic Greek sculpture often has been characterized with terms like symmetry, “blockiness”, axial structure and tectonic disposition, and even the categories of weight and support have been mentioned (usually exemplified by male statues). But to me none of these terms seems helpful with regard to our reconstructed figure. This is even more striking when we look at the back of the statue (fig. 7). Here it is surprising to see that the garment is not the dominant factor, but the natural human shape : the roundness of the legs is elaborately carved out from the heel up to the full buttocks, and these are framed from above and from the side by the bordered edge of the mantle and are accentuated in their plasticity by this plain surrounding, with only the ample coloured border of the back. Just above the hair ends, which flows down all the way from the crown of the head as a homogeneous, exuberant, but natural mass.

20The zigzag border of the mantle culminates a little above the right buttock, and from this point the broad central pleat starts, a little overlapped by the hair, emphasizing the vertical dynamic of the right leg and elongating it up to the slightly raised right shoulder, as if the right side of the figure is dominated by a continuous axis, from which all the fine lines of the chiton and the border of the diagonal mantle lead to one point of the left side, to the grasping left hand, a motif which signals activity ; this intensifies the torsion of the figure and points the viewer to the front, to the space around it. This uneven construction of the figure is obviously enhanced by the painted decoration, which in its distribution and its blue-red restriction is not so much a reflection of life as an allusion to vitality. If we consider again a male statue, a kouros, with its strict symmetry, once more we realize the different concepts of these two prototypes of Archaic Greek sculpture.

  • 13 See Democr. fr. 246 (ed. H. Diels, Die Fragmente der Vorsokratiker II [1956]). The term seems to be (...)

21On the other hand it is obvious that there is a link in the basic character of both these types ; perhaps we could describe this common characteristic in terms of the standing of the figures. Both the kore and the kouros stand equally balanced on both legs, and rise up in a kind of powerful vertical motion without any sign of gravity ; they seem to represent an image of independent, self-supporting standing, which might be in this very period the characteristic of the human being, of human self-confidence. There is a Greek term, which might point out this special quality : αὐτάρκεια, which seems to be used for the first time a little later by Democritus.13


22Since the preparation of the present publication, some complementary aspects of the restoration of the polychromy of the kore Acro.682 in a recent article, by the author : B. Schmaltz, “Zur Rekonstruktion der spätarchaischen Mädchenfigur Akro. 682. Zum Aussagewert einiger Gewandornamente”, in K. B. Zimmer (Hrsg.), Von der Reproduktion zur Rekonstruktion – Umgang mit Antike (n) II. Summerschool vom 16. – 19. Juni 2014 in Tübingen (2016).


Bibliographical abbreviation

Karakasi 2001 = K. Karakasi, Archaische Koren.


1 I. Kader, “Täuschende Spielereien”, in V. Brinkmann, R. Wünsche, Bunte Götter : die Farbigkeit antiker Skulptur. Ausstellung Basel 2005 (2005), pp. 264ff.

2 V. Brinkmann, Die Polychromie der archaischen und frühklassischen Skulptur (2003), on which see the review of B. Schmaltz, Göttingische gelehrte Anzeigen 257 (2005), pp. 9-33 ; V. Brinkmann, R. Wünsche, Bunte Götter. Ausstellung München 2003/2004 (2003).

3 AD I (1889), pl. 39.

4 Karakasi 2001, pll. 144ff.

5 Ibid., pll. 254ff., “Antenor-K.”, p. 264.

6 Karakasi 2001, pl. 146.

7 D. Ohly, Die Aeginete II / III (2001), pll. 201ff.

8 See P. Wolters, AD I (1889), text to pl. 39 ; W. Lermann, Altgriechische Plastik (1907), pp. 87ff.

9 V. Brinkmann, Die Polychromie der archaischen und frühklassischen Skulptur (2003), fig. 107.

10 T. Karagiorga-Stathakopoulou, AD 51/52 (1996/1997), A, pll. 2ff.

11 G. Dickins, Catalogue of the Acropolis Museum 1 (1912), pp. 40ff.

12 D. Brooke, “Terracottas”, in S. Casson (ed.), Catalogue of the Acropolis Museum 2 (1921), p. 337.

13 See Democr. fr. 246 (ed. H. Diels, Die Fragmente der Vorsokratiker II [1956]). The term seems to be used here more in the sense of modesty than in the sense of self-sufficiency.


Professor i.r. of Greek Archaeology, University of Kiel, Germany.

© École française d’Athènes, 2018

Conditions d’utilisation :

Rechercher dans OpenEdition Search

Vous allez être redirigé vers OpenEdition Search