The Reed Motif in the Thera Wall Paintings and its Association with Aegean Pictorial Art

The same is largely true of the Aegean wall paintings; many aquatic plants look like reeds but very few can be identified positively, mainly because the plant's tuft is never depicted. From this point of view, the only true reeds in Aegean art are depicted on the reed bed wall painting from Xeste 3. The strong iconographical similarities show that the fine Late Minoan IB 'reed vases' echo or imitate the marshland depicted on similar monumental wall paintings, which were used as prototypes. The precise dating of the wall painting of the reed bed from Akrotiri in the Late Minoan IA period proves the direct influence of monumental painting on pottery and offers a hitherto unknown iconographic model of Aegean art.

The stimulus for the theme of my paper, the reed motif in the pictorial art of Thera and the Aegean, is the wall painting from Xeste 3 at Akrotiri which depicts a marshy landscape with reed thickets, in which ducks and dragonflies fly.⁽¹⁾ Although the composition has been restored to a considerable extent, the original position of many pieces is at the moment unknown or conjectural. Most of the fragments of this wall painting were found in 1972 and 1973, scattered in Room 3b of the upper storey of Xeste 3, while a section of the wall painting had survived in situ, on the west wall of this room (Marinatos 1976, 24; Doumas 1992, 128, 131). However, the relevant excavation reports confuse the fragments of this wall painting with those from another representing at least three women carrying bunches of flowers (Doumas 1992, 131, figs. 131-134). It is now certain that these two themes do not belong to the same composition.⁽²⁾

The wall painting seems to be some 4 m. long, yet the west wall of Room 3b measures 3.95 m., indicating that the subject extended to the adjoining wall. This view is supported by two pieces of evidence:

     1. A fragment of the red border bands which forms a concave right angle, thus showing that the wall painting continued either onto the south mudbrick wall of the entrance or, more likely, onto the north wall of the room, west of the wall painting of the 'Great Goddess'.⁽³⁾

     2. A fragment forming a convex right angle and preserving a smooth right edge from contact with a wooden beam, perhaps from a door or a window frame.

These corner fragments add weight to the supposition that a window existed on the north wall, near the north-western corner, either corresponding to that on the ground floor, or symmetrical with the one to the east of the 'Great Goddess' (Palyvou 1990, 44, fig. 6).

Three sizeable areas of the wall painting have been restored so far:

     1. Section A, 1.60 m. high and 1.22 m. wide. It is the best restored part of the wall painting. It preserves both the left and the right edge and it probably corresponds to the western part of the north wall, next to the 'Great Goddess'. It depicts three equidistant yellow tufted reeds with dense younger reeds among them and two red dragonflies flying to the left (Figs. 1-2).

     2. Section B, 1.05 m. long and 1.70 m. high, belonging to the right edge of the wall painting which covered the west wall of Room 3b. It depicts the marsh of tall reeds, a duck flying to the right with its head turned back, and behind it a dragonfly on the tuft of a reed. Reeds are less dense around and behind the duck (Figs. 3-4).

     3. Section C, 2.70 m. long and 1.70 m. high, preserving a small part of the left edge of the wall painting, probably belongs to the lower part of the scene.⁽⁴⁾ Its biggest part depicts the marsh, and the larger reeds sprouting from the mud fit with the upper parts of those pictured in section A (Fig. 5).

The restoration of the duck and the dragonflies make sections A and B the sole parts of the wall painting in which we are certain of the position of the other pictorial elements apart from the reeds. The exact position of all the other individual fragments depicting ducks or dragonflies is at the moment unknown.

The upper part of the mural comprises a decorative zone of horizontal bands: a wide orange one, two black ones with a fine grey medial stripe, and two thinner red ones. As usual, impressed string lines were used as guides and generally kept to. The average height of this decorative border is 0.45 m..

The main pictorial zone of the wall painting is filled with dense reeds growing out of the marsh water, appearing against a white background (Fig. 6). The only reed completed so far is 1.21 m. high. The surface of the marsh is denoted by a broad red wavy band that runs along the length of the wall painting, forming smaller hooklets converging and joining together. The marsh is painted in thick, deep yellow ochre. Densely splashed dots of red paint give a vivid picture of the mud, where smaller reeds sprout. The maximum height of the surviving section of the marsh is 0.72 m., with the bottom pact still missing. The original height of the wall painting would not have exceeded 2.50 m..⁽⁵⁾

Two kinds of reeds are depicted, yellow ones and grey-blue ones, a combination of a warm and a cold colour familiar in the Theran wall paintings (Televantou 1994, 302). The yellow reeds were painted first and the grey-blue later, which sequence gives depth and perspective to the composition (Fig. 7). Only a few of the yellow reeds have tufts that seem to be fairly evenly distributed in the composition. Red dragonflies, of which fourteen have been counted to date, are shown on some tufts or leaves. A dragonfly of section A barely touches the edge of a thin leaf, as if balancing its weight against the leaf's endurance (Fig. 8).

The image of the reed is precise. The painter has profound knowledge of the plant and conveys details such as the way in which the leaf sheet envelops the shoot, the elegant curvature of the shoot and the variety in thickness of each tuft. The yellow and grey leaves are entangled as densely as in a land or water reed bed.

In the mural art of Akrotiri all plants are depicted growing out of perspective from the edge of their habitat: the crocuses and lilies out of the rocks, the sea daffodils out of the ground. This convention is followed here, with the reeds growing all together from the bank, except that here they are rendered with excellent perspective, as the thinner grey ones are painted on top of the larger yellow ones and together they wave, bend and entangle in the wind. The result, as far as density, movement and colour are concerned, far surpasses other plant motifs and can only be compared with the lilies of the 'Spring fresco' (Pl. 15). The realism in the painting of the tufts of the reeds verges on photographic reproduction.

The tufts are not all painted in the same way. Some are thick and dense, others thinner but longer. The bushy tuft is painted with deft, long strokes of light yellow emerging from the deeper painted end of the reed stem (Fig. 9). On one occasion, the tuft may not have been painted with a brush but with the impression of a real tuft dipped in yellow paint and imprinted directly onto the plaster (Fig. 10). The result is impressive. The tufts are soft and translucent, swaying lightly, or almost upright, as if their movement depends not only on the wind but also on their weight. The white background to the scene - hitherto perceived as a stylistic convention of the 'Cycladic school' rather than the projected image of natural space - evocatively conveys the limpid sky of the islands (Hood 1987, 104; Sapouna-Sakellaraki 1981, 483; Davis 1990).

At present, fragments of six or seven ducks have been identified in the composition. The male duck already published (Marinatos 1975, 120, pl. 144b; 1976, 34, col. pl. L, pl. 42a; Doumas 1992, 131, fig. 135; Morgan 1988, 63) is of the common species Anas platyrhynchos, known as the mallard (see Petrou 1995, 84, 215; Harte this volume) (Pl. 16; Fig. 11). At the same height in relation to the reeds flies the female duck at the right edge of section B of the wall painting, proceeding right but with her head turned back (Fig. 4). Also female are two smaller ducks flying to the left, while a fifth one flies to the right, next to a dragonfly. There are smaller fragments of the neck and plumage of a sixth and possibly a seventh duck.

The ducks are painted in the four basic colours of the wall painting, in various combinations: yellow, red, black and blue. Three nestling chicks waiting to be fed are depicted in black outline only. These appear on a fragment not yet restored, so its position on the wall painting is not yet known. The reed tufts, the marshland and the dragonflies are painted without outline. The leaves are painted with long firm strokes of diluted colour arising from the shoot of the reed, resulting in a light, transparent effect. The stroke seems to be free and easy, but the possibility of the use of stereotypes (e.g. a french curve; cf. Birtacha and Zacharioudakis, here, vol. I) for the depiction of the leaves cannot be ruled out. In fact, during the preparation of the drawing of the wall painting, the latter possibility came to seem the more likely for the curvature of the leaves.⁽⁶⁾

There is no doubt that the painter of the marshy landscape is the so-called 'Painter of the saffron gatherers' (Hollinshead 1989; Immerwahr 1990, 237; Televantou 1992a, 59-61; 1992b, 152), who executed the 'Spring fresco' and all the known wall paintings of Xeste 3 except the ladies carrying bunches of flowers. The dragonflies of the marshy landscape are identical with those depicted on the frieze of the swallows feeding their young (Doumas 1992, 135, fig. 97). The tufts of the reeds have been painted in the same detailed manner as the crocus clusters. Furthermore, this painter not only uses the same colours in his wall paintings but also works in the same shades: thin grey-blue, thick ochre yellow, thick deep red.

In the composition of the reed bed it is evident that the 'Painter of the saffron gatherers' had observed the reeds, ducks and dragonflies well, just as he had the lilies and swallows (Immerwahr 1990, 237), for in design, motion, colour and anatomical details the depiction of plants and animals is characterised by the same realism, fidelity to nature and "love for the joy of life and capturing the moment" (Televantou 1992a, 61). In spite of the artist's ability, however, the three white nestlings depicted in the marsh are neither ducklings nor goslings, for in nature the former are grey and the latter yellow. The same inconsistency occurs with the young swallows in their nests from Xeste 3 which are depicted yellow instead of grey-black.

Green does not exist in the Thera wall paintings, thus grey-blue and ochre yellow paint was used for the reed thickets, as for most of the plants (Doumas 1992, 19; cf. Perdikatsis et al., here, vol. I). The neck of the mallard was painted black, but the male red dragonflies (Crocothemis erythraea) and some other details of the ducks are painted with colours that match these species in nature (Petrou 1995, 97). It is obvious that the painter preferred fidelity to colour contrast in the depiction of a plant or an animal to fidelity to colour tone. Though conventional, the colours chosen for this wall painting render the colours of the reed bed very vividly, especially those of late autumn when the old plants are dark yellow and the new ones green-grey.

On entering Room 3 through the open polythyra, the spectator beheld the wall painting with the 'Great Goddess' in front of him, the marshy landscape on his left and the 'Saffron Gatherers' on his right (Televantou 1992a, pls. 18-19). The almost human scale of these wall paintings gave the spectator the illusion of a real world. The frontal perspective in which the reed thickets, the marsh and the muddy bottom of the mire are painted is of considerable interest. The whole composition is in reality the image projected to the spectator watching from the opposite shore, approaching the shore or standing in the water. Reeds of 1.5 to 2 m. high stood in front of his eyes with ducks and dragonflies flying a little higher up, while below he observed the undulating shore and the reed shoots springing out of the mud. This type of perspective is totally different from that of the riparian landscape and the whole of the West House miniature frieze, where palm trees and other plants grow from the river bank in bird's eye view en face, while the rivers are depicted from directly above (Televantou 1994, 196, 333, αναδ. πίν. 3, αναδ. σχέδ. 2; for the perspective in Theran wall paintings, cf. Laffineur 1990).

The rule of the tripartite division of Theran wall paintings into horizontal zones is also confirmed in those from Room 3 of Xeste 3 (Iliakis 1978, 618; Linardou 1992, 71; Doumas 1992, 20). Their common feature is the deep coloration of the upper decorative and the lower ground zone, which contrast markedly with the white background of the middle zone. Nevertheless, on the scenes of the lateral walls, that is the 'Saffron Gatherers' and the marshy reed bed, the ground is extremely undulating in contrast to the flat ground of the 'Great Goddess' of the central mural.

The interpretation of these wall paintings as an iconographical complex is beyond the purpose of this paper. However, I would like to mention some new evidence that derives from the restoration of the riparian scene. The fowling scene with birds caught in nets, described by S. Marinatos (1974, 17), does not seem to exist in the wall painting. The ducks fly free over and among the reeds, and one of them is probably feeding the three chicks in the nest.

It has already been pointed out by Doumas (1992, 31; cf. Marinatos 1984, 68, fig. 49) that the 'Goddess' is in the midst of chthonic, terrestrial, aerial and imaginary creatures. If so, the marshy landscape and its fauna represent the missing natural element: water. Thus, the 'Great Goddess of Nature' is presented and adored in a symbolic milieu of serenity, fertility and natural life.⁽⁷⁾

As N. Marinatos (1984, 71; 1993, 203) has demonstrated, the wall paintings from Rooms 2, 3 and 4 of Xeste 3 "form a coherent pictorial programme of independent but related themes which revolve around nature and womanhood", but her claim that "the predominant symbolism in the frescoes...alludes to the spring" (Marinatos 1984, 65, fig. 44; 1985, 219) should be reappraised in the light of the new evidence. The tufts and the colour of the reeds, as well as the red dragonflies, allude directly to the hot Mediterranean autumn, just like the flowers of the saffron crocuses. Ducklings mainly hatch in spring, but there is a second brood in autumn. The colourful plumage of the male duck appears in late spring or early summer. Thus the wall paintings of Room 3, and in particular the marshy landscape of the west wall, echo an environment spanning more than one season, from spring to autumn. Clearly, the painter's chief concern was to render all plants and creatures of the marsh at the peak of their fertility and activity and not necessarily a single season. Freedom, imitative symbolism⁽⁸⁾ or whatever, the result depicts marsh life in its prime, in its sublime moment.

Reeds grow all over continental and insular Greece, close to rivers, lakes, streams and ditches, often forming dense thickets (Baumann 1982, 77 no. 133)⁽⁹⁾ (Fig. 12). The feathery tuft appears from August onwards on both common species:

     1. Arundo donax, 3-7 m. high with a wide and thick tuft up to 0.70 m. long and symmetrical/opposed green leaves (Cook 1974, 368, fig. 164).

     2. Phragmites communis, grows to a height of 1.50 to 3 m.. Its leaves are longer and thinner. Its tuft is thin and triangular (Cook 1974, 423, fig. 202).

The species depicted in the marshy reed bed seems to be Arundo donax (Möbius 1933, 30; Morgan 1988, 20-21, pls. 20-21; Huxley and Taylor 1977, 157, fig. 415; cf. Sarpaki this volume). However, the stems of the reeds are thin and tall, but not straight and rigid like those of real reeds in the swamps. The way in which they bend in the wind strongly resembles reeds on land, as does the length of their lanceolate leaves.

THE REED MOTIF IN AEGEAN PICTORIAL ART

The reed motif is not easily identified in Aegean pictorial art (Morgan 1988, 20). On sealstones, because of their limited surface, the reed appears as a supplementary motif in nature, and its miniature rendering makes recognition uncertain (Ruuskanen 1992, pl. 14:E2a.1, pl. 15:E3.15-18, pl. 16:E2b.10, pl. 18:E9.1). On pottery the 'reed' is usually a simple but vague floral device and rarely can the actual plant be securely identified. The same is largely true of the Aegean wall paintings: many aquatic plants look like reeds but very few can be identified positively, mainly because the plant's tuft is never depicted. From this point of view, the reeds in the Xeste 3 wall painting are the only true reeds in Aegean art.

       A. WALL PAINTING

Scholars have shown that the apparent naturalism of many plants depicted in Minoan wall paintings is deceptive, since very often features of different species of plants are combined (Hood 1987, 63; Dickinson 1994, 165). The painted flowers and plants display only a "general similarity to the real ones, they are stylized and disintegrating" and seem to obey the "Minoan propensity to create decorative artistic hybrids" (Cameron 1984, 129 n.8, 134; cf. Warren here, vol. I, fig. 19). Thus, following this canon, the reeds in Aegean art can be identified only approximately. In the few cases where they appear, they are mostly depicted as low bushy plants without tufts. These plants seem to reproduce the reed-grass motif of the pottery, depicting simple grass rather than weeds.

       1. Akrotiri

Doumas (1992, 22, 27) has claimed that on Thera the art of wall painting is characterised by the tendency to represent the plants as true to life as possible. However, this does not seem to be true of the miniature wall paintings, where it is difficult to identify many of the plants depicted (Morgan 1988, 20). Reeds do not appear clearly on the river banks of the subtropical landscape of the miniature frieze, but simple red and blue slanting plants with long symmetrical leaves are depicted (Doumas 1992, figs. 30-34; Televantou 1994, 250; Pl. 2, 3-3.40 m.). Some bushy plants have dotted or fringed leaves, a type which appears on Theran pottery as well (Marthari 1992, 279, fig. 17.23-25).⁽¹⁰⁾ Reeds or other water plants appear beside the quay and around the river that surrounds Town IV of the same mural (Doumas 1992, figs. 35-36, 44-45; Pl. 3). In her attempted restoration of the wall painting of the blue monkeys (Room Beta 6), N. Marinatos (1984, 116, fig. 83) conjectures a small stream with rushes growing nearby, by incorporating the fragments of rushes published by S. Marinatos (1969, pl. 5.2; Doumas 1992, 110, figs. 85-86).

The broad red undulating band that runs the length of the wall painting of the reed bed and renders the surface of the marsh is unique in Theran art.⁽¹¹⁾ The wavy band with the characteristic 'hooklets' indicating water (seashore, river bank, marsh) is a convention deeply rooted in Aegean pictorial art (Morgan 1988, 14, 34; cf. Sakellariou 1985, 297, fig. 4) and encountered in the art of Egypt and Mesopotamia as well (Doumas 1985, 25-30, 35-36). The conventional depiction of the seashore with black 'hooklets' (Doumas 1985, 29-30) appears in the scene of the 'shipwrecked', in the river of the first town and on the quay of the second town of the miniature frieze (Doumas 1992, figs. 26, 29; Pl. 1, 0.70-2 m.).⁽¹²⁾ In the river of the east wall, the combination of the black wavy bands of the shore with the wide yellow band defining the shore line is of great interest. Red curved patterns on the inside of this band possibly depict the muddy shallows of the riverbed (Doumas 1992, figs. 30-34; Pl. 2, 3-3.40 m.). By contrast, the river running around the first town does not have the same yellow band.

       2. The Cyclades

In the Cyclades the reed motif appears outside Thera only once, in the LC I/LM IA miniature paintings from the north-east bastion of Ayia Irini on Kea. The wall painting represents a marshy landscape in late summer. On the quay of a town by the sea a woman passes over a bridge, beside a river and a marsh whose big reeds sway in the wind (Morgan 1990, 254, fig. 1; 1995, 243). Two different floral designs are depicted, one consisting of isolated intertwining bramble and myrtle plants, the other of "grasses and bushes painted in muddy colours which give the impression of a marshy landscape near the sea... Clumps of reeds, grasses, bushes and small trees are painted in muddy colours of blue and brown over a tan background. Some fragments seem to represent underwater grasses"  (Abramovitz 1980, 71, 74, pls. 8, 9a-e, 10a).

Morgan (1990) has discussed in detail the strong iconographic and stylistic similarities of this miniature frieze with the one from the West House at Akrotiri. Among these, we mention only that the water plants depicted are painted in blue and light brown and that, in both cases, they are not easily identifiable.

       3. Crete

Some Minoan floral wall paintings are cited here, not only because they include reeds and other plants but also because they are parts of riparian scenes iconographically very close to the reed bed of Xeste 3.

       1. On a fragment of an MM IIIB wall painting from the South East House at Knossos there are "spikelets of reeds or grasses" (Evans 1921, 426, 539, fig. 390; Cameron and Hood 1967, 21, pl. D fig. 1).⁽¹³⁾ The flowered plant depicted is obviously not a reed; more likely it is a fennel (Ferula communis) (Baumann 1982, no. 96).

      2. On a fragment of an MM III wall painting from the South House at Knossos "a small reed-like plant rises out of what may be taken for a blue pool with undulating red and yellow banks"; another plant grows out of a yellow ground (Evans 1928, 378, fig. 21a,b).

       3. The LM IA wall painting from Room P of the Unexplored Mansion at Knossos has been reconstituted by Cameron (1984, 128, 129 n.8, 131, 134, 146, pls. 43d-e, 44d, 46:12-16, 47, 48; Boulotis 1995) and recently by Chapin (1997, figs. 1-2) as a natural landscape full of plants and flowers which cannot be identified with certainty. This wall painting has many iconographic similarities to the Theran reed bed, not only in the style and the subject itself (for example, the overlapping plants), but also in secondary elements such as the thick decorative bands of the upper zone and the undulating ground zone.

       4. According to Cameron's reconstitution (Cameron 1968, 26, fig. 13), the LM IA frieze from the House of the Frescoes at Knossos (Evans 1928, 431-467, pls. X-XI, figs. 264, 275; 1935, 891) depicts a dense rocky landscape full of plants such as ivy, papyrus, shrubs and others, where birds fly among rivers and waterfalls and monkeys climb on the rocks (Hood 1987, 62, figs. 31-33, 50A; Chapin 1997, 19 figs.; Bietak, here, vol. I, fig. 17). Morgan (1988, 20, fig. 22) thought that reeds with their tufts are depicted on this mural, but I think that these plants look like osiers(?) (see the 'Blue Bird' fragment of this fresco (Hood 1987, 61, fig. 32; Boulotis 1995, fig. 1)).⁽¹⁴⁾

The same iconographic type of hybrid 'papyrus-reed' with the "flowering papyrus-like tufts" is depicted on the LM II argonaut frieze (Evans 1935, 891, figs. 870-871) and on the murals from the Throne Room at Knossos (Evans 1935, 910, fig. 884, col. pls. XXXII-XXXIII, XXXVc; Cameron 1987, 321-325, fig. 7).⁽¹⁵⁾ Bushy plants in dense or sparser clusters convey the natural surroundings in the frieze of the partridges from the Caravanserai at Knossos, possibly of LM IB date (Evans 1928, 117, fig. 55; Hood 1987, 68, figs. 40-41).

       5. In the Royal Villa of Ayia Triada, the three sides of Room 14 seem to have been decorated with an LM IA mural that covered the entire walls from floor to ceiling (Evans 1921, 538, fig. 391; 1928, 354 fig. 201; Pernier and Banti 1947, 31; Hood 1987, 63, 303 n.38; Cameron 1987, 321 fig. 10 (the drawing is printed right to left)). The 'Great Goddess' was depicted on the central wall, to the left a young girl picking flowers and to the right a rich landscape full of plants and animals. Although much has been written on the similarities between the iconographic programme of this room and that of Room 3 of Xeste 3, Morgan (1990, 264, 262) remarks that "a juxtaposition between human cult activity and the natural world seems to be typically Cycladic".

       6. A fragment of an LM IA miniature frieze from Katsambas depicts two birds flying over an undulating (rocky?) landscape from which sprout reed-like plants (Alexiou 1955, 318, fig. 2; Shaw 1978). The theme of ducks flying among rocks recalls directly similar miniaturist representations, mainly on sealstones, and gives good evidence for the identification of a widespread iconographic prototype in Minoan art, which is reproduced on a large scale in the reed bed wall painting from Xeste 3.

       4. Egypt

Riparian scenes of hunting or fishing are depicted in the wall paintings of numerous tombs of nobles, dating to the New Kingdom (Eighteenth to Twentieth Dynasties, 1570-1080 BC), but such scenes occur in Old Kingdom tombs as well (Lhote 1954, pl. 6). Ducks fly in large flocks among papyrus thickets and sometimes among reeds, trying to escape the hunters who have surrounded them.

It is worth noting that when reeds appear they are painted in a similar way to the Theran ones: their leaves are long and symmetrical, some of them have long, thick tufts and overlap the papyrus plants as they bend in the wind.⁽¹⁶⁾ The best examples of wall paintings in this category decorated the tombs of Nakht (Lhote 1954, pls. 52-53) and Horemheb (Lhote 1954, pl. 58). Sometimes the birds' nests and eggs are depicted among the ducks and the thick marshy vegetation, as well as butterflies, grasshoppers or locusts, dragonflies and other insects (tomb of Nakht: Lhote 1954, pls. 52-53; tomb of Menna: Lhote 1954, pl. 54; tomb of Horemheb: Lhote 1954, pl. 58; tomb of Khonsu: Lhote 1954, pl. 59).

In these scenes the representation of the swamp and its plants, as well as the human faces, is always static. The plants are presented upright and motionless, the birds fly in unnatural positions over them, and the lack of scale creates a highly conventional rendering of landscape. Only at the end of the Eighteenth Dynasty are the marsh plants and birds released from these canons, becoming as realistic as the Aegean ones. Thus, the marsh plants (probably reeds or rushes) on a floor of the palace of Amenophis in Tell el Amarna (ca. 1350 BC) are painted in a vivid manner (Lhote 1954, pl. 13; Hayes 1959, 291, fig. 179).

Although most of these riparian scenes are dated a little or much later than the Theran wall painting, it is evident that there is a close relationship between them, and the existence of some common iconographic conventions as a result of interaction and mutual influence between people and artists of the eastern Mediterranean cannot be ruled out.

       B. POTTERY

       1. Minoan pottery

The Minoan floral motif, with one to five pairs of lanceolate leaves sprouting from a tiny or longer shoot and rarely symmetrical, is conventionally called 'grass' or 'reed'. This motif already exists on the pottery of the MM IIA period (Betancourt 1985, 93, fig. 69), but from MM IIIA onwards, down to LM IA, it becomes one of the commonest patterns on vases of different shapes and sizes (Evans 1921, 579, fig. 424; Betancourt 1978, 382, fig. 1; 1985, 109, figs. 84:L,P, 98:E; Dickinson 1994, 116, fig. 5.10:3).

The reed, in light paint on dark, appears as the central motif on Kamares ware vases dating from all phases of this pottery style (Walberg 1976, fig. 46 nos. 13, 19; Marthari 1992, 48, 272, pl. 65b).

The LM IA reeds usually appear slanting and rarely upright, depending on the height of the vase, and growing from a ground line (Evans 1930, 549, 627 n.2, fig. 349; Walberg 1976, 50, fig. 36:24-26, 29-37; Dickinson 1994, 112, fig. 5.8:4; Macdonald 1996, 20, pls. 3b, 5a, 6c). However, on a stirrup-jar from Gournia (Betancourt 1985, pl. 18:1), the widely spaced reeds arise from a dotted band that depicts the water of a marshy landscape rather than the earth, as has been suggested (Morgan 1988, 37, pl. 22).

The reed motif does not appear on the contemporary LH I pottery of the mainland, although bichrome floral motifs are depicted on a panelled cup exported to Akrotiri (Marinatos 1972, 31 n.2, pl. 62c; Davis 1978, 216; Marthari 1980, 194, 201; 1992, 53, pl. 81b; Demakopoulou 1988, 155 no. 107).

       a. Minoan pottery imported to Thera. Reeds are rarely depicted on LM IA vases imported to Akrotiri.⁽¹⁷⁾ Nevertheless, the two characteristic modes of rendering them can be seen on two different shapes. Reeds painted on a little cup have wide leaves and stem rising from the band running around the base of the vase (Marthari 1992, 46, pl. 35b: National Archaeological Museum no. 60). Those covering the body of a jug with cutaway spout, which is one of the most impressive Minoan vases found at Akrotiri (Marthari 1992, 46, pl. 46a: National Archaeological Museum no. 89), are thinner and taller (Fig. 13). The shoot and the leaves are executed with long, thin brush strokes, which show part of the shoot and continue as a lanceolate leaf. Some of the reeds continue under the first ground line of the lower part of the vase. On this jug some of the qualities of wall painting are observed: a tripartite division of the surface, the development of the main motif in the middle zone, speed in execution, freedom in design and, finally, the depiction of a real reed bed.

       2. Theran pottery

     a. Light on dark/dark on light decoration. The reed motif decorates most of the shapes of the local pottery at Akrotiri. Marthari (1992, 272, figs. 15-17) has described and discussed them in detail. As on Minoan pottery, the reed motif on Theran vases appears in two versions, which are actually two different motifs: the smaller and simpler 'grass-reed', of decorative value only, and the pictorially elaborated plant which is more easily identifiable as a reed.

In the 'grass-reed' motif, the identification of the plant depicted is almost impossible. Theran vase painters used this conventional motif extensively, because of its simple form and quick reproduction (Marthari 1992, 278-279). It occurs equally in both techniques of Theran pottery, dark on light and light on dark, on conical cups (Marthari 1992, 73, 75, pls. 95b, 100b), trefoil-mouthed jugs (Marthari 1992, 132; pls. 237b, 238), eyed jugs (Marthari 1992, 133, pls. 244, 248a), hole-mouthed jugs (Marthari 1992, 137, pl. 262), bridge-spouted jars (Marthari 1992, 141, pl. 285a), hole-mouthed jars (Marthari 1992, 139, 140, pls. 267b-268), bridge-spouted skyphoi (Marthari 1992, 151, pl. 303a), piriform jars (Marthari 1992, 163, pl. 335a), rhyta (Koehl 1990, 350, figs. 1, 8; Marthari 1992, 96, 98, 101, 105, fig. 17 no. 23, pls. 155-156, 160b, 162a, 167), conical pithoi (Marthari 1992, 174, pls. 361-362a), cylindrical pithoi (Marthari 1992, 177, pl. 368b) and bath tubs (Marthari 1992, 178, pls. 372, 374b). Semi-globular cups (Marthari 1992, 80, 81, pls. 113b, 118) and piriform cups (Marthari 1992, 83, 85, pls. 123b-127), two of the most popular shapes of Theran pottery, are commonly decorated with stylised white or black reeds. The reed is repeated at regular intervals round the vase, and is rendered in a few simple strokes which are mostly conventional and rarely very successful in effect (Marthari 1992, pl. 118b) (Fig. 14). Finally, two shapes - the jug with cutaway spout (Marthari 1992, 131, pls. 230-232, 234) and the kalathos (Marthari 1992, 196, pls. 398-400) - are almost exclusively decorated with reeds.

Marthari has noticed that, in comparison to the light on dark, the dark on light reed is commonly combined with other floral motifs, such as the crocus (Marthari 1992, 132, pls. 237b, 238), the palm (Marthari 1992, 273, pl. 296b) and the myrtle (Marthari 1992, 273, pl. 162). The reed also appears among "plants of the Theran agricultural production", that is grapes, vetches (Marthari 1992, 273, pl. 343), wheat (Marthari 1992, 152, pls. 307-309a) and olive trees (Marthari 1992, 133, pls. 244, 248a).

The 'grass-reed' is a highly stylised motif, and for this reason it does not appear on shapes with potential for more fully developed pictorial themes, such as the polychrome kymbe (Mastrapas 1991, pls. XIV, XLIII:3-4) and the ewer (Immerwahr 1990, figs. 2-3).

     b. Bichrome/polychrome decoration. The reed motif appears among the patterns of the 'black and red' style pottery of Akrotiri, dating to the end of the MC period (Marthari 1992, 34, pl. 30:3-4).⁽¹⁸⁾ On LC I pottery, reeds are restricted to some jugs with bridge spouts (piriform teapots). The shape is not a common one at Akrotiri, for fewer than ten specimens have been found, and precise or close parallels have not been found elsewhere in the Cyclades (Marthari 1992, 148). Their decoration is usually bichrome (red and black) and rarely polychrome (with the addition of white). Often, red and black reeds alternate in the middle zone of the vase. They are depicted sparsely, either with simple curved strokes, as on most of the other shapes (National Archaeological Museum 1470) (Fig. 15), or with equidistant groups of dense shoots (Thera Archaeological Museum 2157; Marthari this volume, fig. 10) or dense, bushy overlapping plants (Thera Archaeological Museum 4944). The reeds sprout from wide or thin bands that cover the lower parts of the vases. On none of these jugs is the tuft of the reed depicted.

The piriform teapot Thera Archaeological Museum 4944 is decorated with an all over pattern of overlapping red reeds (Marthari 1992, 146, pl. 295b; this volume, fig. 9) (Fig. 16). Only the large dots around the shoulder are painted in black. The reeds rise from multiple wide red bands which possibly render the surface of the water, and their stems seem to continue to the bottom. The leaves often do not join the stems, and the strokes are free, starting from the tips of the leaves and ending where they touch the stem. They are opposed or alternating in relation to the stem. Some reeds, to the left of the spout, are depicted as if bending in the wind.

The piriform teapot Thera Archaeological Museum 3720 has fine pictorial decoration consisting of palm trees and reeds in yellow and black (Marthari 1992, 146, pl. 296b; this volume, fig. 8). The thinner black reeds are painted above the large yellow ones, exactly as on the wall painting under discussion here. Despite the restricted space, the colours and design endow the reeds with movement and perspective, the black ones giving the impression of younger shoots. It is worth noting here that overlapping motifs are a characteristic of wall painting which does not commonly appear on Theran vase painting (Marthari 1992, 224, 395, pl. 144-145).

Hands of individual vase painters are difficult to distinguish among the mass produced Theran pottery. The 'grass-reed' motif does not provide enough stylistic evidence for such identifications, and the elaborate bichrome reed is not so frequently depicted as to offer a sufficient number of representations to enable us to make the kinds of definitions ⁽¹⁹⁾ that can be attempted for the crocus or other motifs (Marthari 1992, 106, pls. 170b-171).

In the Theran vases presented above (the only ones with bichrome reed decoration), the combination of colours and the rendering of the dense overlapping reeds bring them very close to the wall painting of the marshy landscape from Xeste 3. Furthermore, the co-existence of palms with reeds alludes directly to the riparian landscape of the miniature frieze from the West House (Televantou 1994, 250; Höckmann 1978, 608), where bichrome reeds are also depicted. It is evident from these vases that the painting on vases of natural scenes was not inspired by the natural habitat, but consciously or unconsciously reproduced widely diffused stereotypes which were probably influenced by monumental painting.

It is hard to say if these wall paintings were the only prototypes for this kind of decoration, just as if is hard to determine to what extent the evolution of the reed as a pottery motif was autonomous. Marthari (1992, 277) has pointed out that the Theran piriform teapot derived from the LM IA jug with cntaway spout decorated with vertical reeds; but in spite of this influence, the reeds on Minoan jugs are sparsely pictured, whereas on Theran vases they are dense and colourful.

       3. Melian pottery

As on Thera, the reeds on Melian pottery (Atkinson et al. 1904, 169, pl. XIX:9; Zervos 1957, 238, fig. 320; Marthari 1992, 279) are painted in both dark on light and light on dark style, the latter in imitation of late Kamares ware. The reed is a common motif of the MC III-LC I bichrome pottery of Phylakopi, depicted on its own or combined with floral and other motifs (Atkinson et al. 1904, 127 fig. 97, 140 figs. 112-113; Zervos 1957, 238, fig. 321; Marthari 1992, 279). It decorates various shapes, such as the conical cup (Atkinson et al. 1904, pls. XVII:30, XVIII:6, XIX:9-10; Dawkins and Droop 1910-11, pl. VIII:40; Barber 1974, 33, fig. 6:40; Marthari 1992, 279), the hole-mouthed jar (Atkinson et al. 1904, 132 type 3; Marthari 1992, 140), the kalathos (Atkinson et al. 1904, 169, pl. XIX:I0) and the conical pithos (Barber 1974, 39, pl. 4e; Marthari 1992, 92) among others. Red and black reeds alternate on a bridge-spouted bowl, just as on the Theran piriform teapots (Dawkins and Droop 1910-11, pl. VIII:153; Barber 1974, 38 no. 153; Marthari 1992, 279).

Marthari (1992, 383) has demonstrated that Melian LC I vase painting is not as rich in naturalistic motifs as Theran pottery. However, it is from Melos that two of the most interesting specimens of bichrome pictorial pottery come. Both were found at Phylakopi and come from bath tubs or large vases.

The first sherd (National Archaeological Museum 11420; Atkinson et al. 1904, 141, fig. 114; Zervos 1957, 250, fig. 335; Mastrapas 1991, 40, 129, 167, pls. X:2, XLII:3; Morgan 1988, 64, fig. 48; Porter this volume, fig. 5) shows the front part of a duck with half-spread wings beside a blossmning papyrus or sea daffodil. ⁽²⁰⁾ The calyx and the anthers of the flower are painted red, while the stamens and the shoot of the plant, as well as part of the duck's plumage, are black. The similarity to the sea daffodils from the House of the Ladies at Akrotiri is remarkable (Doumas 1992, 34, figs. 1-4). Both have seven stamens and the same colouring on the same parts of the flower.

The second sherd (National Archaeological Museum 11412; Atkinson et al. 1904, 142, fig. 115; Zervos 1957, 247, fig. 331; Mastrapas 1991, 40, 129, 167, pls. X:1, XLII:3; Morgan 1988, 64, fig. 47) shows the upper part of a duck walking rather than flying among thick aquatic plants. Even without the tuft, the shoots and the leaves of at least two reeds are easily identified. The duck's head is black, the neck and eye are red. The reeds are also painted red.

The ducks on the two sherds show close iconographic parallels to those from the riverine landscape of the miniature frieze and the fragment from Sector A (Doumas 1992, 25, figs. 30-34; Televantou 1994, 234, figs. 52-53; Pl. 2), but are less like the larger scale ducks in the marshland scene of Xeste 3.

The correspondence between the subject of the two Melian sherds and the marshy landscape presented above is striking, and proves that there was developed in the Cyclades, concurrently with the monumental painting, an important genre of vase painting of narrative character, on a par with the wall paintings. The possibility that such vases were decorated by the same artists as created the wall paintings is slight, but cannot be eliminated (Mastrapas 1991, 129, 166; Immerwahr 1990). Landscapes of sub-tropical/Nilotic character - like those depicted on the two sherds from Phylakopi - recall the prototypes used in wall painting, pottery, sealstones and decorated artefacts (Sakellariou 1985; Mastrapas 1991, 168; Ruuskanen 1992, pl. 14:E2aA).

       4. The 1M IE 'reed painters'.

The LM IA-B floral style reaches its aesthetic and technological peak with the 'reed painters' of the LM IB special palatial tradition, whose span of activity did not exceed more than a generation (Popham 1990, 27). The 'reed painters' were identified by M. Popham (1967, 341-343, figs. 2, 8, pl. 79e-f) and P. Betancourt (1976, fig.1) who attributed to four hands about ten fine vases (cups, flower pots and jugs), richly decorated with dense reed thickets sprouting from the water of a marsh or river (Morgan 1984, 172-173, fig. 4a-c; 1988, 37, pls. 23, 47, 48) (Fig. 17). These vases - certainly of Minoan origin ⁽²¹⁾ -were found at Knossos, Phaistos and Zakros, and on Rhodes, Melos and Kea (Betancourt 1985, 140, 145, pls. 18:D, 21:A-C; Morgan 1988, 37, pls. 23, 48), while recent specimens come from the Menelaion of Sparta (Catling 1996, 72, fig. 2:10) and from Naxos (Hadjianastasiou 1989, 212, pl. 40c) (Fig. 18).

The decoration completely covers the vases, as these have no spouts or secondary handles, creating an iconographic result very close to that of the reed bed wall painting. The tufts of the reeds are absent, but even so there is little doubt of the identification of the plants as reeds.⁽²²⁾ The wavy band with bifid stalks, from which the reeds sprout, has been painted just like the red wavy band of the wall painting. Groups of little dots give an image of the muddy bed, which is emphasised with similar red dots in the wall painting.

There is not doubt that these 'reed vases' echo or imitate the marshland depicted on similar monumental wall paintings, which were used as prototypes. The precise dating of the wall painting of the reed bed from Akrotiri in the LM IA period proves the direct influence of monumental painting on pottery, and offers a hitherto unknown iconographical model.

Walberg (1986; cf. Davis 1990, 224) has maintained that Kamares ware had a great influence on wall painting, not only in technique but also in composition, while Boulotis (1993; 1995, 16) considers it possible that some of the skilled artists of this colourful palatial pottery went on to the art of wall painting. Judging the iconographic achievements of each local school as a whole, it seems that, in both Crete and Thera, wall painting reproduces the most characteristic style of decorative pottery. Thus, the colourful Kamares ware pottery is directly related to Minoan wall painting with its multi-coloured background, in the same way as Cycladic pottery with its white background is related to Cycladic wall painting with the same background (Davis 1990, 214, 223; Marthari 1987).

An important element noticed by Marthari (1987, 366, 368) is that in the polychrome Theran pottery and wall painting there is a tendency "to limit and repeat motifs" (this appears in all monumental compositions with natural themes - antelopes, lilies, papyri, monkeys and reeds), while in Minoan compositions the artists include a great variety and quantity of subjects in one composition. Nevertheless, the almost simultaneous spread of the art of wall painting within Crete and the south insular Aegean proposed by Boulotis (1995, 16) and Marthari (1992, 237) means that the necessary ground for the development and consolidation of these characteristics already existed. The obvious implication is that the bichrome and polychrome Cycladic pottery was that ground⁽²³⁾ where the chromatic and pictorial prototypes of wall painting were formed. The art of wall painting may have come to the islands as a new skill from Minoan Crete, but, once there, its rapid transformation should be credited to the achievements of Cycladic vase painting, which had a long tradition of naturalism in contrast to Minoan and Helladic pottery (Marthari 1992, 385).⁽²⁴⁾ For instance, as has been convincingly demonstrated (Mastrapas 1991, 149, 236), the bird motif had a long and strong tradition in Cycladic pictorial pottery, going back to the Early Bronze Age, while it had yet to appear in Crete and the Greek mainland.⁽²⁵⁾ It is believed (Immerwahr 1990, 241; Marthari 1992, 228) that the bird is a subject which passed into wall paintings from vase paintings, although shortly afterwards vase painters seem to have been influenced in their turn by the great monumental art of wall painting (Marthari 1992, 229, 237).⁽²⁶⁾

Something similar probably took place in the case of the reed motif, which already appears in Middle Cycladic bichrome pottery. Overlapping and dense reed thickets in the LM IA/LC I period are depicted exclusively on Theran and Melian vases, since there are no close parallels in Minoan art at that time. Thus, the motif of the dense reed thickets, though partly dependent on the Minoan reed motif of the MM III/LM IA ceramic tradition, seems to have enjoyed an autonomous iconographic evolution in Cycladic art, on both pottery and wall painting. If so, the possibility that the Cycladic(?) 'reed thickets motif', as a popular pattern of Theran pictorial pottery, influenced the Minoan 'reed painters' of the palatial workshops, who reproduced and elaborated it in the LM IB period, cannot be ruled out. Nevertheless, the development of the elegant LM IB floral style pottery, with its dense reeds, is due not only to the long familiarity of the Aegean vase painters with the reed motif, but also to the influence of Aegean (or even international) monumental painting depicting marshy landscapes. Iconographic programmes painstakingly reconstructed, such as the one in Room 3 of Xeste 3, contribute to the definition and the identification of such prototypes in Aegean art.

(1).      For a complete presentation of the wall painting, see Vlachopoulos forthcoming.

(2).      According to S. Marinatos (1976, 36), the three female figures "belonged, together with the...thicket of reeds, to the Western part of Room 3 outside the polythyron. The thicket of reeds covered the Western wall. The three women decorated the Southern wall, and perhaps the South-West corner of that narrow space."

(3).      "Only the North-West corner is occupied by a thicket of reeds" (Marinatos 1976, 34).

(4).      The left part of this section was found in situ on the west wall.

(5).      The wall painting with the 'Great Goddess' is 2.30 m., and that of the 'Saffron Gatherers' 2.44 m. high.

(6).      Stereotypes were also probably used for drawing the wings of the ducks, as their dimensions are very similar.

(7).      For the ritual character of the natural settings and fertility landscapes in the wall paintings, see Marinatos 1984, 85; 1993, 193; cf. Angelopoulou this volume.

(8).      N. Marinatos (1984, 68) thinks that in the iconographical programme of Xeste 3 "naturalism has been sacrificed to symbolism". 

(9).      Although extensive marshes are not common in the Cyclades, there are still some reed beds in marshy landscapes, such as those between Komi and Kolymbythres on the island of Tenos.

(10).     This hybrid type of 'palm-branch leaf' can be traced back to the MM III period (Evans 1935, 911).

(11).     The wide red undulating band only exists on the kymbe, without always indicating the surface of the water, the seashore or the ground line of the landscape. On the kymbe shape, see Mastrapas 1992, pls. XIV, XLIII.33-4.

(12).     On new evidence for the restoration of the miniature wall painting of the east wall, see Televantou 1990, 322, fig. 13.

(13).    N. Marinatos (in the discussion following this paper) interprets the "coiled tail of what appears to be a small harvest mouse" (Evans 1921, 539) as the foot of a duck sitting on the tuft of a reed. I believe that this oblique hook entangled in the spikelets cannot be a duck's foot, since it is very thin, the webs are missing and its position is wrong for a flying or sitting duck, details inconsistent with the naturalistic hand responsible for this Knossian wall painting. In any case, ducks never sit on single reed plants, since neither the stem nor the feathery tuft can stand their weight.

(14).     For the osier, see Baumann 1982, 64 no. 98.

(15).     The 'papyrus-reeds' of the Throne Room are 1.02 m. high and their leaves are painted in blue or green and red colours.

(16).     The reed depicted in the Egyptian wall paintings of the Old Kingdom is Phragmites australis or communis, but during the New Kingdom Arundo donax came to Egypt (Germer 1985, 203, 205; Schoske, Kreissl and Germer 1992, 6, Abb. 3).

(17).     By contrast, other floral motifs are very common on the LM IA pottery imported to Akrotiri (Marthari 1992, 32). The reed motif also appears on sherds of imported Kamares ware vessels (Marthari 1992, 31, pl. 25a:2).

(18).     On the pottery changes at Akrotiri during the MC-LC transition period, see Papagiannopoulou 1990, 64, table III (reed).

(19).     For the identification of painters in Theran pottery, see Marthari 1992, 224, 395, pls. 116b, 143b, 144-145, 163-164 (painter of the kymbe National Archaeological Museum 3267).

(20).     For the identification of the plant as the sea daffodil, see Baumann 1982, 177 nos. 356-359; Doumas 1992, 34, figs. 1-4. The iconographic type of the duck among papyrus appears in jewellery (MM III-LM IA gold pendant: Effinger 1996, 50, Taf. 9b, 46a) and on sealstones (Evans 1930, 116, fig. 66; Morgan 1988, 64, fig. 46).

(21).     Although Minoan in origin, some of them are possibly of Helladic manufacture (Jones 1986, 456).

(22).     However, from the botanical point of view, the plants depicted on the LM IB reed vases are more likely to be wheat tufts than reed leaves. For the wheat tufts, see Karamanos 1987, 70, figs. 11.5a, 11.6a.

(23).     Marthari (1992, 203), however, appreciates that the Theran polychrome pottery probably derived from the influence of Kamares ware.

(24).     Papagiannopoulou (1990, 66) believes from her own study of the Middle Cycladic period that "the pictorial motifs (with the exception of the human figure) which were most familiar in the pottery of the later Cycladic period are precisely the motifs which already existed on the Middle Cycladic pottery, and it is the same motifs, but more refined, that we meet on the frescoes. Hence there is a very good local sequence in pictorial decoration at Akrotiri."

(25).     For the MC III bird jugs, see Mastrapas 1991, 128, 132, 143; and cf. Renfrew, here, vol. I, figs. 17, 19. For the duck motif in Late Bronze Age Aegean pictorial art, see Vanschoonwinkel 1990, 337-341.

(26).     Immerwahr (1990, 243) discusses the interrelation of the pottery representations and the frescoes, and admits that "while there is interaction between the two media, in the case of each motif the relation seems different".

----------------------------------------------------------

----------------------------------------------------------