Form and Meaning in Figurative Painting

These communicative devices are then analysed in relation to the scene of male figures within the Xeste 3 programme. A programmatic parallel between the young males and young females is identified in the form of formal and idiomatic analogies suggesting parallel significance in the rituals depicted, in both cases connected with puberty and initiation rites.

The paper demonstrates how form and meaning are inextricably linked, both in terms of individual form and on the programmatic level within the architectural space.

A two-dimensional representation of the rounded, moving figure of a human being is, as we are all aware, an illusion. Yet within that illusion of reality lies a wide range of possible expression, from the ultimate in verisimilitude to the limits of minimal form. Function and context, audience expectation, cultural and artistic experience, as well as individual skill, combine to determine the level of representational necessity. The extremes of verisimilitude find expression in Roman portraiture in the ancient world, and in photographic realism in the modern. In alternative media, such as painted clay modelling over skulls, or holography, the mind could conceivably be manipulated into believing simulacra were reality, but in figurative painting the image and the model remain separate, understood as one only in the imagination.

Stressing the salient features of form draws attention to particular traits of type or of individual. This principle has long been understood by cartoonists - in caricature or the grotesque we recognise the individual. It is also true of works of art which seek to emphasise aspects of human nature, such as gender, fertility or prowess. The 'Venus of Willendorf' or some of Picasso's figures are quintessential examples, emphasising female sexuality through form and leaving the head, whose cognitive functions are here rendered irrelevant, as a small extension of the body without defining features.

It is remarkable how minimal form can be and yet still be understood as a representation (rather than purely as a symbol or sign). In some such cases, cultural understanding is a prerequisite, as in the virtually hidden figurative content of kelims, aniconic idols, and some pictograms. In others, both form and general meaning are clear, but the specific symbolic message necessitates cultural understanding. A good example is the phallus (linga) in Hindu art, which stands for the generative force of the god Shiva (Blurton 1992, 78-84). In yet other cases, minimal form generates a response of recognition sufficient to categorise yet insufficient to specify; examples abound: figurines (Hamilton et al. 1996) from numerous cultures (Cycladic included), Greek Geometric vase painting, some rock art, and certain modern artists such as Arp, Klee, Dubuffet and Henry Moore.

The body may be represented as an expression of an ideal (Greek and Roman art, the Italian Renaissance, Indian temple sculpture), or distorted and transformed for the purposes of expression (Chinese Shang bronzes, North-West Coast Native American split-representation, Cubism, Surrealism). Skin coloration, clothing or its absence, jewellery, hair and body decoration, can all be used to manipulate the basic form and enrich meaning. Stasis and movement, scale, position in the pictorial field, orientation, posture and gesture are all elements in communicating type and situation in representations of the human figure.

What is central to a study of such elements and compositional conventions is the issue of relativity of perception and the question of 'natural' versus symbolic expression. Take, for example, the idea of self and other. Paintings and sculptures of Europeans from the fifteenth to nineteenth centuries are distinctly different from one another when executed by African, Indian, Japanese or Native American artists. The 'Europeanness' may be instantly recognisable, but so too is the ethnic origin of the art. When we encounter an Egyptian painting of an Aegean envoy, does the figure look more Aegean or more Egyptian?⁽¹⁾

The question of natural versus symbolic arises most forcefully in the case of gesture (Bremmer and Roodenburg 1991; Morris et al. 1979; Morris 1994). Certain movements of the arms and hands are clearly functional and when transferred to art become expressive gestures. But they are not necessarily as 'natural' as one might think, and there is little universality in the details. Beckoning, greetings and farewells, which might seem straightforwardly functional, are actually amongst the most varied gestures worldwide, while, as northern Europeans and Greeks well know, fundamental misunderstanding can arise from differences of meaning in simple head nods and shakes or gestures such as 'thumbs up' (Axtell 1991, 20-33, 49-51, 59-61; Collett 1993, 17-19, 70, 199-203; Morris et al. 1979, 162-168, 185-196, 241-6). In translating these gestures into single images, artistic communication necessitates that the most clearly defined moment of the action be translated into form, frozen and emphasised (cf. Gombrich 1966). This applies equally to symbolic gestures, including those which relate to specific ritual actions, such as the henu gesture of Egyptian art and hieroglyphs, which signifies praise but actually represents a typical moment in a dance-like sequence of ritual actions (Ogdon 1979; Wilkinson 1992, 16-17; Betro 1996, 39). Many gestures in art are culture-specific in their meaning and are purely conventional, as for example the Buddhist middle finger to thumb perfect pose, and the Byzantine fourth finger to thumb teaching gesture. Most gestures in art have an expressive source combined with a conventional idiom and, for this reason, many appear to be cross-cultural. Yet such an assumption can be dangerously misguided, for gestures, in art as in life, can be ambiguous. A raised hand or hands, palm(s) outward, can express adoration (Egyptian), prayer and supplication (Near Eastern and Early Christian), benefaction (Byzantine) or greeting.⁽²⁾ What these have in common is an intention of well-being towards the other. Where they differ is in the hierarchic order and consequent message.

Cultural significance extends beyond the particulars of gesture to, in many cases, range of expression. Pain (other than that experienced in war), sorrow, fear, pathos, shame and ecstasy all find expression only in a limited range of the world's art. For many cultures they have not been seen as relevant to the art experience, though no one can doubt that they were a part of life. Just as the range of human gesture is culture-specific, so too is the range of human emotion as expressed in art. The reasons for this lie in the differing functions of visual representation.

It is issues such as these that I wish to consider in this paper on form and meaning in the figurative paintings of Thera. So much has been written on the iconographic significance of these wonderful, miraculously preserved paintings. Yet, all such iconographic interpretation is predicated on an understanding of the communicative range of the art and the relationships between forms and their meanmgs.

THERAN FIGURATIVE PAINTING⁽³⁾

Painted figures, like their counterparts in life, communicate through a number of individual, social and cultural traits which combine to tell the viewer something about type (age, gender, social position) as well as the particulars of the contextual situation through action and communication (meanings of glance, gesture, pose in relation to other figures or objects). In this paper, I shall consider the communicative devices used by the Theran artists in painting the human figure within those two broad categories, ending by looking at one particular group of paintings - the males from Xeste 3 - within those terms.

ACTION AND COMMUNICATION

Action and communication between figures within the pictorial space provide the framework for iconographic meaning and visual narrative. Stasis and mobility, posture and gesture, rhythms and repetitions in the motions of bodies all combine to produce a language of the figure, visual and without speech, yet effective in its communicative power.

       Conventions of the human figure

To understand how this language is working when the figure is in action, it is first necessary to know the basic conventions of the human figure within the cultural idiom. This enables the viewer to be aware of how mobility changes convention for the purposes of communication. Though the static, hierarchic posture of gods, kings, priests and nobles in Egyptian art4 is less adhered to in the Aegean, it is still the basis for depictions of the human figure and does occur in its full conventional form, most notably in procession scenes. From Thera, the 'Priestess' and the best preserved of the 'Fisherman' panels from the West House provide solitary examples (Pl. 5; Doumas 1992, figs. 24 and 19).

As in Egypt, the figure stands with head in profile but with eye in frontal view, affording the most characteristic and recognisable features when drawn in two dimensions. Frontal shoulders give way around the hips to a return to profile view for the legs. This is shown in the figure of the 'Fisherman', whose nakedness reveals the twist of the body (Pl. 5; Doumas 1992, fig. 19). In this stance, the two feet are generally (pre-Amarna) both viewed from the outside, showing only the big toe, while hands may be shown with left and right distinguished, or simply as two the same (Schäfer 1974, 294-299; Immerwahr forthcoming). The far leg strides forward. In the standing position, both heels are firmly on the ground, but in running, the back heel is raised. In Egyptian art there is a space between the heel of the front foot and the toe of the back, but in the 'Priestess' and the 'Fisherman' there is none. It is a static, in this case, presentation pose. Yet in the 'Fisherman', the flowing line of the sway back, and the tilt of the head forward, release the figure from the conventional stillness.

The feet part when the figure is more actively engaged: either with feet still on the ground, as in the 'Boxing Boy' on the left (Doumas 1992, figs. 79-80), the House of the Ladies (Doumas 1992, figs. 6-7, 11-12), soldiers marching and people walking in the miniatures (Doumas 1992, fig. 28); or with one or two heels raised, as in the Xeste 3 girl with veil, crocus gatherer climbing a rock, girl tipping out her basket, girl balancing a pot on her head in the miniatures (Pl. 11; Doumas 1992, figs. 107, 120, 123, 28).

In Egyptian art the static pose is reserved for figures of status (also shown on a larger scale) their offering bearers, and priests (Fig. 5). Those whose bodily preservation in the afterlife is of less concern (the workers, defeated enemies), or who are involved in extreme physical movement (dancers, acrobats, wrestlers), are released into mobility of limbs and body. This is particularly notable in the shoulder region. But, while in Egyptian art the nearest shoulder is often pushed forward to enable the figure to hold something with two hands or to work at something in front of the body, only in the New Kingdom does this involve an entire twist of the body into profile. In the Old and Middle Kingdom, the far shoulder is shown in frontal view, while the near shoulder either lies against the body or overlaps with the far shoulder as though they were parallel to one another. In New Kingdom painting, a less conceptual and more perceptual approach is taken, and we see significant shifts of viewpoint, including unfamiliar views such as the three-quarter back view, though still with profile head (Tomb of Rekhmire, TT 100: Mekhitarian 1954, 51).

Theran art is full of experimentation of viewpoint (Betancourt 1977). We see it in the animal depictions as much as the humans, most notably in the varied movements of the swallows in the 'Spring Fresco'. What is interesting is that such experimentation is even applied to an otherwise static figure in a presentation pose. The second 'Fisherman' (Pl. 7; Doumas 1992, fig. 18) stands with legs only slightly apart, one foot overlapping the other, his entire body in profile, with near shoulder brought forward and the far shoulder hidden behind the profile line of his chest. This is a bold departure from convention and stands, I believe, as a conscious contrast to the mixed-view profile of the other 'Fisherman' from the same room. It is the same combination as we see in the two 'Boxing Boys' (Doumas 1992, fig. 79), though there with a kinetic function: the frontal shoulders in the righthand figure give the dangling arm a sense of recovery before the blow, while the forward push of the near shoulder in the left hand figure delivers the punch.

In the Xeste 3 paintings, hierarchy is maintained through the strict frontal shoulder view of the centrally positioned seated goddess (Pl. 12; Doumas 1992, figs. 122, 125), in contrast to the forward movement in action of the two other seated figures in central positions - the wounded girl and the man with jug (Doumas 1992, figs. 105 and 114). Frontality of the body is also maintained in the woman offering a necklace (Doumas 1992, fig. 101), where the breast marks the return to profile, and the two mature ladies on the opposite wall (Pl. 4; Doumas 1992, figs. 131-134). In the boy holding a cloth (Doumas 1992, fig. 113) and the lefthand 'Crocus Gatherer' (Doumas 1992, fig. 118), the frontal shoulder view has a function of mobility, enabling the figures to turn their heads backward, a more natural solution than that achieved in the girl with veil who turns despite her forward posture (Doumas 1992, fig. 107). All the other figures, male and female, bring the near shoulder forward in action, with varying degrees of naturalism. The girl with veil approaches the Egyptian solution of overlapping shoulders and looks as though she might be folded in half (Doumas 1992, figs. 107-108), while the seated man with jug attempts a three-quarter view but with near shoulder painfully contorted forward, despite the well observed line of the back (Fig. 2; Doumas 1992, fig. 114).

The Theran figures retain their formal independence by avoiding the solution of overlapping that Egyptian (and later, Greek Geometric) artists used when depicting numbers of figures or figures walking beside one another. That device is used for animals (sheep and goats and dolphins in the miniatures (Pl. 3; Doumas 1992, figs. 28, 36, 41-42), and antelopes (Doumas 1992, fig. 83)) but, with the exception of certain figures on the ships in the miniatures, not for the humans. This provides a notable anomaly in the 'Boxing Boys' (Doumas 1992, fig. 79) who, despite their extremely active pursuit, overlap only their arms and otherwise stand, albeit one knee flexed, in the static pose. This could be said to be a trait of boxing, as opposed to wrestling, but I think it is also used symbolically, as a mark of hierarchy in the figures, emphasising what the jewellery on the left hand boy already says - that this is no childish game but a ritualised sport (Morgan 1985, 16-18; 1995a; Marinatos 1984, 106-112).

       Gesture, posture and expression

The above has shown that variations on the conventional idiom can be used to define hierarchy, relationships between figures, and intention of action. The same is true of gesture and posture, which, along with facial expression, also have the power to evoke emotional states and physical sensations.

Expressive gestures are those used in life as in art, in which parts of the body - usually head, arm, hand - move in a specific pattern of functional emphasis, expressive of response, desire or emotion. They can be seen in the miniatures, with the man who turns with outstretched arm in the 'Meeting on the Hill' and the man on the hill of the Arrival Town who similarly beckons, and the small male figure at the base of the town who taps his neighbour on the shoulder (Pl. 3; Doumas 1992, figs. 27, 38, 47). Both gestures function unequivocally in drawing the attention of the one communicated with towards the communicator.

The wounded girl of Xeste 3 touches her bleeding foot with one hand and claps her palm against her forehead with the other (Doumas 1992, fig. 105). This could denote pain. But the facial expression does not confirm this, and nor (for a wounded foot rather than head) is this a natural or functional gesture.

In the miniatures, two women in the Arrival Town raise one hand (Pl. 3, 2.60-3.60 m.; Doumas 1992, fig. 38). Though originally interpreted as waving towards the incoming ships, I pointed out elsewhere that the gesture (particularly of the woman in front) was directed at the horns of consecration in front of her (Morgan 1985, 5-19; 1988, 118). Here a culturally specific gesture can be recognised only through its contextual associations, since it occurs repeatedly in relation to the same cult object. The gesture is similar, but not identical to the 'adoration' gesture of Egyptian art (which, however, uses both hands (cf. above, note 2)) and clearly has a similar meaning, though with a narrower referent.

Another highly specific gesture is that used by two of the men in the 'Meeting on the Hill' and one near the stern of ship 1 in the miniatures (Doumas 1992, figs. 27 and 40). Both elbows are outstretched, with one hand raised, resting on the other arm. At first interpreted as 'giving the rhythm' in relation to the figure on the ship (Marinatos 1974), it became clear that repetition of the gesture served the function of a visual analogy between the two scenes (whether in narrative terms - the same figure shown twice - or to show a similarity between the two events). Its interpretation as a ritual gesture rests on iconographic context as well as comparisons with certain figurines (Morgan 1988, 117-118). In two-dimensional art, however, it is not a familiar conventional gesture, and it is likely that its recognition by Bronze Age viewers would depend on knowledge of ritual activity.

In contrast, the gesture of outstretched arm with stick, held by one of the participants in the 'Meeting on the Hill' (Pl. 1, 0.70-2 m.; Doumas 1992, figs. 26-27 lower left), recurs on a number of sealstones. This could be an ambivalent gesture. Usually it denotes the power or command of the holder, whether of divinity or authority ('Mother of the Mountains' seal type, Chania 'Master Impression', 'Chieftain's Cup', the 'Priest King' relief, seal from Naxos: Morgan 1988, 118; Hallager 1985; Koehl 1986; Niemeier 1987).⁽⁵⁾ In the Theran painting there is no indication that the figure is in any way special, other than the gesture itself. I have suggested placing the figure between the pastoral scene and the 'Meeting', so that the man walks towards the latter (Morgan 1988, 157, pl. 1). In this case, it is conceivable that the priestly figure whose head is turned back and whose arm is outstretched is beckoning for the man holding the symbol of power to come forward. The clothing of the man with the stick, however, is in no way distinguished, which fails to support this suggestion.

In contradistinction to ancient Egyptian gestures, which are not only constantly repeated but also confirmed in meaning by their use as hieroglyphic determinatives, many Aegean gestures recur relatively infrequently, so that supportive cross-referencing is sometimes insufficient to establish unequivocal meaning.

Offering gestures, however, by combining symbolic convention with expressive movement, are usually clear in intention. In Egyptian art there are several variations, all of which rely on the movement of hands from the giver to the recipient, and most of which depend for their form on the shape of the object held.⁽⁶⁾ The same is true of Aegean art. The 'Priestess' (Doumas 1992, fig. 25), holding a brazier with one hand above, one below, takes the most conventional of the Egyptian gestures. It is a form derived from two hands holding an object, the far hand depicted above for clarity. In the case of the 'Priestess', the hand above holds a plant over the brazier and is therefore functional in position, but the resultant gesture is strictly conventional. It is repeated, more loosely, in the arm positions of the three boys bringing objects towards the seated man in Xeste 3 (Doumas 1992, figs. 112-113, 115): one hand is above the other, whether they are carrying a vessel or a cloth, even though, in the case of the latter, the position is adapted for a more natural effect. The woman with the necklace, in contrast, uses one hand, dangling the beads in front of her (Pl. 19; Doumas 1992, figs. 101, 104). She, like the boy with the cloth and the woman with the pot on her head in the miniatures, sways her body in balance as she moves.

This rhythmic awareness of the body is characteristic of Theran figurative art. The people are animated in their bodies and even, to a certain extent, in their faces. Most of the younger large-scale figures have parted lips, an indication of liveliness and anticipation of speech. The two older women in Xeste 3, in contrast, keep their mouths closed, as does the more static figure of the 'Priestess', and the 'Young Boxers' in their ritualised sport.

Intention of action is expressed in the forward movements of limbs and body of a number of figures: robing, offering, bringing, picking, pouring, balancing, beckoning. Rhythms and repetitions of the emotions of bodies set up communications between figures: the eye contact and raised arms of the "Boxing Boys', whose eyebrows echo the tilt of their heads (Doumas 1992, fig. 81); the three naked men in the sea in the north wall miniature, falling as in a cascade (Doumas 1992, fig. 29); the rhythm of the warriors marching in step (Doumas 1992, fig. 28); the turned heads of the boy with the cloth and the girl wim the veil in Xeste 3 (Doumas 1992, figs. 113 and 107, discussed below). All these communicate through body language, and, despite the lack of dance among the activities, the motions of many of the figures have a sequence and rhythm that is itself dance-like in effect.

STATUS 

Status can be revealed pictorially in a number of ways. Some of the factors may be corollaries to social indicators in life: clothing, hairstyles, jewellery, which are frequently indicators of wealth as well as positions in the social hierarchy. (Though, of course, all the large scale figures virtually by definition must portray figures of social and/or religious significance.) Other factors are purely pictorial devices: relative scale, position in space, orientation, focal position.

       Clothing, jewellery and hair

Much has already been written on clothing, jewellery and hairstyles,⁽⁷⁾ and they are the subject of a paper by Laffineur (this volume), so I will keep my observations here brief. The elaborate patterned textiles and flounced skirts of the large scale women and girls in Xeste 3 and The House of the Ladies are clear markers of significance (whether of event or personage). Within the miniatures, simpler, repeated clothing distinguishes between town and country. Priestly garments (male and female) are long and straight, tied at the neck ('Meeting on the Hill') or wound around the body and tossed over the shoulder ('Priestess'). The origins of the types lie in the Near East, and there are contemporary parallels in the Aegean in the Kean miniatures and on sealstones (Morgan 1988, 93-96; 1990). The revealed breast on women, being primarily a signifier of age rather than status, is discussed below. Nakedness, reserved for males, has two, divergent, meanings, both common to the contemporary Egyptian and Near Eastern tradition. In a sprawling figure confined to the lower zone, the significance is that of a defeated enemy (north wall miniatures). On a young boy, it signifies pre-pubescence. The latter is discussed below.

Jewellery is worn by all the large scale women, though the two mature women in Xeste 3 are without earrings. The goddess is particularly distinguished, with necklaces of ducks and dragonflies as well as beads, hair ornaments as well as earrings and bracelets (Pl. 13; Doumas 1992, figs. 125-126), and one of the women in the scene below carries a necklace, as though bringing an offering (Pl. 19; Doumas 1992, figs. 101, 104) (cf. the 'Mykenaia': Kritseli-Providi 1982, pl. Γ). The acrobat by a palm from A1 wears an earring (Doumas 1992, fig. 148); the better preserved 'Fisherman' wears a narrow gold necklace tied like string (Doumas 1992, fig. 23); and the lefthand 'Boxing Boy' is distinguished from his opponent by necklace, armlet and anklet of beads, and wears a gold hoop earring like the acrobat (Doumas 1992, figs. 79-81). None of the Xeste 3 men wears jewellery.

Hair is adorned in the case of the goddess (Pl. 13; Doumas 1992, fig. 126) with beads, in that of the wounded central girl (Doumas 1992, fig. 106) with plants (one a jewel, one perhaps live), and perhaps also in the case of the woman offering a necklace (Doumas 1992, fig. 102). Otherwise, ribbons hold hair in place in its different stages of growth, as discussed below. Hair, like nakedness, is more a signifier of youth than of relative status. Two exceptions are the frond-like streamers from the head of the A1 acrobat (Doumas 1992, fig. 148) (identified as an acrobat by N. Marinatos (forthcoming a)) - directly paralleled on two acrobats on a sealstone from Knossos (Bietak 1995, fig. 4; Morgan 1995b, fig. 5), and the hairlock worn as a mark of distinction by the helmsman on each of the large ships and by the man with the clearly defined gesture standing next to the helmsman on ship 1 (Doumas 1992, figs. 39-40) (Televantou 1994, 209-213). It is interesting to note that Egyptian sailors two centuries later are depicted with the same lock as a mark of rank or profession (Morgan 1988, 102-103).

       Position, orientation and focus in pictorial space

Relative size, such a clearly defined hierarchic indicator in Egyptian and Near Eastern art, is rarely employed in the Aegean. Indeed, on the north wall of the West House miniatures (Pl. 1, 0.70-2 m.; Doumas 1992, fig. 26) the defeated, drowning figures in the sea are larger than the priestly figures in the 'Meeting on the Hill', giving an uncanny (though surely accidental) impression of linear perspective. A certain hierarchy may pertain in the Arrival Town (Pl. 3, 2.60-3.60 m.; Doumas 1992, fig. 38), in that the people (especially the women) in the town are larger than the row of male figures at the base of the town walls. Yet, it is my impression that this relative size has more to do with clarity of narrative than scale of importance.

Only one unequivocal instance of hierarchic scale is evident, in the seated figure of the goddess from Xeste 3 (Pl. 12; Doumas 1992, fig. 122). Her head and body are significantly larger than those of the figures around her. The central seated male figure from Room 3b is also larger than the standing males (Doumas 1992, fig. 110), and this too is likely to be an indicator of relative importance. These two are also distinguished by their position within the pictorial space (Figs. 1-2). Both are raised above the other figures, seated yet with their heads considerably higher than those of the standing figures around them. This is in contradistinction to the seated figure in the other main composition, the wounded girl (Doumas 1992, fig. 105), who is both lower in position than the standing figures flanking her and no larger than the older of her two companions, the woman holding the necklace (Doumas 1992, fig. 101). Scale and position therefore jointly distinguish the goddess and the seated man.

Other figures who appear to be distinguished by position are those under the awnings and in cabins on the ships, as well as the women in front of horns of consecration in the miniatures (Pl. 3; Doumas 1992, figs. 35-38). These, however, are positions with life corollaries rather than purely pictorial devices. The zoning of drowning figures below marching warriors (Pl. 1, 0.70-2 m.; Doumas 1992, fig. 26) is an example of a situation from life (the physically defeated fall to the ground) appropriated by convention to become a pictorial device. It was universally used in the ancient world as a sign of dominance over the enemy.

In Xeste 3, the goddess and the man, seated at a higher level, also reflect a life situation - that of rulers raised on their thrones - yet this merely provides an analogy for a symbolic device rather than a one-to-one correlation between life and art. Neither is seated on a throne, neither shows evidence of being royal. Nanno Marinatos has shown that the position of the goddess reflects her divine status, the exotic animals flanking her acting as intermediaries - one mounting the stepped dais, the other on it - between the divine and the earthly realms (Marinatos 1984, 61-62). Such symbolism is not apparent in the figure of the man. He sits alone, probably on a rock (though here the painting is not preserved). Yet his position imparts special significance to his person.

Both figures are the focus of attention within the architectural/pictorial space. The device used to draw attention to focal points within the paintings is that of orientation. The eye of the spectator follows the direction of movement of the figures, and when two converge on to a single central figure this becomes the focal point. On a large scale we see this device in the 'Ship Procession' (Pl. 3; Doumas 1992, fig. 35): the two towns flanking the central focus of the ships, the attention of the townspeople drawing the spectator in towards the central action.

The large scale figures, almost exclusively, converge either on to each other or on to a central personage or object. These two formats may be combined, as in Xeste 3. Where there is a flow of emphasis in one direction it is usually from left to right.⁽⁸⁾

Examples of left to right flow are the miniature warriors on the north wall of the West House, and the mature women in Room 3b, upper level, in Xeste 3 (Pl. 4; Doumas 1992, figs. 28 and 131-134). An exception is the 'Priestess', who walks from right to left (Doumas 1992, fig. 24). This exception gives rise to two possible interpretations, depending on where the panel was originally situated. If, as S. Marinatos (1972, 26) originally thought, she stood on the wall separating the two rooms, then she would be standing, with her incense burner and offering gesture, in direct relation to the large scale cabins and in the real position of an offering before the stern cabin of a ship. If, on the other hand, as Doumas (1992, 47, 56) and Televantou (1994, 173-174) argue, the figure stood on the east door jamb between Rooms 4 and 5, then the reversal of the norm would have an architectural context - the 'Priestess' would be walking from Room 4 into Room 5 (towards the ships).

The acrobat from A1 (Doumas 1992, fig. 148) could also be said to be anomalous, in that his head faces to the left. But, although too little survives of the composition to be sure, I wonder whether he is part of the same programme as the monkeys moving from left to right towards the shrine (Doumas 1992, fig. 147), the palm tree having been to the right of the shrine and the acrobat acting as a figure converging on to the central cult area.

Convergence is exemplified by the two lines of men in the 'Meeting on the Hill' (Doumas 1992, fig. 27). On a one to one basis, it is seen in the 'Boxing Boys' (Doumas 1992, figs. 79, 81), whose eye contact and crossed arms provide an eternally balanced central focus bouncing back and forth between the two. In these cases, the meeting of the figures is itself the focal point. Similarly,the two 'Fishermen' face towards one another, so that their point of convergence becomes the corner of their joining walls, the probable placement position of their ritual offerings (Marinatos 1984, 37-38, plan, fig. 31). The two surviving women in the House of the Ladies, if walking forward, would meet on the narrow east wall. Yet here, the left-right idiom of directional hierarchy is evident in the woman holding the cloth who brings it to a (mainly lost) woman on her right (Doumas 1992, fig. 7). Seen in this way, the woman on the opposite wall (Doumas 1992, fig. 6) becomes part of a central convergence onto the focal point of the recipient of the garment. Equally, we should remember that anyone entering the room did so from that corner,⁽⁹⁾ and could become an actual priestly presence within the iconographic as well as the architectural space.

The most obvious examples of convergence towards a central figure are the three parallel scenes in Xeste 3 (Figs. 1-3): the goddess (Pls. 12, 11, 13, 10; Doumas 1992, figs. 122-130), the wounded girl (Doumas 1992, fig. 100), the man with water jug (Doumas 1992, figs. 109-111). All three are seated, defining their centrality in contrast to the standing figures that flank them. In each, there are sub-scenes of contrary directional flow, but the main focus is not in doubt. With the goddess scene (Fig. 1, above) on the upper level of the adyton, the sub-scene consists of the one to one convergence of the two crocus gathering girls on the adjacent (right) wall (Pl. 8; Doumas 1992, fig. 116). Their eye contact connects them, but this is achieved through the turning back of the lefthand figure's head: Both otherwise move towards the left to the scene of the goddess. In the scene on the lower level of the adyton (Fig. 1, below), the sub-scene, again on the righthand wall, becomes a second focal point. The focus is here a cult object, in the form of horns of consecration on a shrine. The girl with a veil has body and limbs facing to the left, but turns her head back to the right to look at the horns (Doumas 1992, fig. 107). She is therefore the formal corollary of the backward turning crocus gatherer (Doumas 1992, fig. 118). Yet her bodily position, facing to the left, brings her into the main scene of convergence on to the wounded seated girl in the centre of that wall. She therefore functions as a link between the two scenes and simultaneously acts as part of a left-right directional flow towards the focal point of cult, and part of the convergence on to the focal point of the narrative in the form of the wounded girl. In the case of the male scene in Room 3b, the exact placement of all the figures is less certain, but the overall disposition is clear. Two boys stood on the south wall and the adult man sat on the southern end of the west wall. The fourth figure, a boy, stood either further along the west wall, as Doumas originally envisaged, or on a partition wall projecting from the northern end of the pier-and-door partitions (Marinatos 1993, figs. 210 (plan) and 215 with partition; Doumas 1992, 130 for the northern partition, although there is none in the plan ibid. fig. 92).⁽¹¹⁾ In Fig. 2 I have drawn the former. The yellow ochre divide, visible on the fragments showing both the adult man and the fourth figure, could then have corresponded with a partition (as in Palyvou here, vol. I, fig. 13). Either way, the three boys - two from the left, one from the right - converge on to the central scene of the seated man (Doumas 1992, figs. 109-111). The sub-scene of directional flow in this case takes place in the boy with cloth (Fig. 4; Doumas 1992, fig. 113), who walks towards the seated man but turns his head back in the direction of (though not looking at) the young boy behind. Given that this is on the ground floor level, I take this to be an echo of the girl with veil, on the same level in the adyton (Fig. 1).⁽¹¹⁾ This would explain the figure's position on the left, looking back. It matches the girl's position on the right, looking back, and provides a symmetrical rhythm of convergence and divergence. This analysis draws attention to the formal cohesion of the Xeste 3 programme, a point which will be discussed below in relation to the meaning of the scene with the males.

THE SOCIAL INDIVIDUAL

       Youth, age and gender

One of the remarkable aspects of the Theran figurative paintings is the concentration on youth and the careful distinctions between different stages of maturity. This has been commented on in detail by other authors (Davis 1986; Doumas 1987; Withee 1992; Younger 1992, 288; Marinatos 1993, 202-203; cf. Koehl 1986), and is the subject of two of the papers in this Symposium (Davis this volume, appendix; Doumas this volume).

Nakedness, as elsewhere in the ancient world, is an indicator of extreme youth in boys. However, only the 'Boxing Boys' (Doumas 1992, fig. 79) (who are not entirely naked since they wear signifiers of their sport around their waists) and the smallest of the Xeste 3 boys (Fig. 4; Doumas 1992, fig. 112) have the bodily features of the pre-pubyscent child: protruding belly, narrow shoulders, shorter legs and (in the 'Boxers') large head (cf. Weingarten 1995, 257-258). The 'Fishermen' and the other Xeste 3 boys have young but fully grown bodies.

Amongst the females, who are never shown naked, size, definition and exposure of the breast are factors indicating relative age. The large breast with sharply defined nipple of the woman holding a garment in The House of the Ladies (Doumas 1992, fig. 10), is perhaps an indicator of lactation. Amongst the Xeste 3 women, only those with mature hair (as defined below) reveal developed breasts: the woman holding a necklace (Doumas 1992, fig. 101), the goddess (Doumas 1992, fig. 125), and (of greater size) one of the mature women from Room 3b (upper storey) (Doumas 1992, figs. 131-132). The frontal view torso of one of the 'Crocus Gatherers' (Pl. 20; Doumas 1992, fig. 119), and the breast of the 'Crocus Gatherer' with basket on her shoulder (Pl. 10; Doumas 1992, fig. 130) show a young, small breast, with nipples indicated by a red dot, hence without prominence. In both cases, their hair is intermediate between childhood and adulthood. Of the very young girls, those whose breast appears to be visible ('Crocus Gatherer' facing left, girl emptying basket (Doumas 1992, figs. 121 and 124)) have flat chests and no indication of nipples.

It is the hair that most clearly defines youth and age in the Theran paintings (Davis 1986; Doumas 1987; cf. Leach 1958). In the very young, the head is shaved and represented blue to show the stubble, whilst individual locks are left as markers of their youth. This hairstyle, unusual though it may seem to modern eyes, was practical and widespread. It would have helped in keeping the scalp clean and free of parasites. It also acts as a signifier, marking the individual in terms of her or his early stage in life. It was common practice in Egypt, where, from the Old Kingdom on, the single 'lock of youth' was displayed on the right side of the head, later also on the back (Müller 1980; Janssen and Janssen 1990, 37; Davis 1986, 401-402 nn. 13-15); and it is seen in Egyptian representations of Nubian and Syrian children as well. The practice is not restricted to the ancient world; it was, for example, also common practice among children in China and Japan (e.g. Smith and Weng 1973, 181; Keyes 1989, 8, fig. 12). Some adults in Egyptian art also sport the style, signifying their ethnic origins in Egyptian an - Libyans, Hittites, Shasu-Bedouin (Morgan 1988, 90 and n.16, 96, 102) - but the locks are usually in combination with short hair rather than shaved heads. In the Aegean, evidence for locks comes from Cretan ivory and clay figurines with holes in their heads for locks, and from bronze locks for presumably wooden figures (Davis 1986, 404 and nn. 31-32; Morgan 1988, 198, n.95). There were no other painted representations of the style in Aegean art until the discovery by Manfred Bietak and his team of the paintings from Tell el Dab'a (Bietak 1992; 1994a; 1994b; 1995; 1996a; 1996b; Bietak and Marinatos 1995; Marinatos 1996; Morgan 1995b; 1998; Shaw 1995; Warren 1995). There, one of the bull leapers, also distinguished by his paler skin (discussed below), has the long locks of bull leapers on a blue scalp, showing that he, too, is to be understood as young (Fig. 6).

It is in the Xeste 3 paintings that relative youth is most clearly expressed. The righthand 'Crocus Gatherer' (Doumas 1992, figs. 120-121) and the girl with the veil (Doumas 1992, fig. 107) are the youngest of the girls, with blue heads punctuated by a back and, in the latter, sidelock, plus a short forelock in the shape of a curl. On the 'Crocus Gatherer', black dots on the scalp clearly indicate the beginnings of regrowth - the stubble. The other 'Crocus Gatherer' (on the left, looking back: Pl. 20; Doumas 1992, figs. 118-119), the girl with basket on her shoulder (Pl. 10; Doumas 1992, figs. 129-130), and the girl tipping out her basket before the goddess (Pl. 11; Doumas 1992, figs. 123-124), are next in age, with a mass of short curls indicating regrowth as well as forelock and, on the first two, lock of youth at the back. All three have a shaved area in front of the ear, shown in blue with black dots, like the youngest two; and all but the girl with the veil (perhaps the youngest of all) have a band at the top of the forehead, which would keep hair out of the eyes as it started to grow. The curls on the 'Crocus Gatherer' and girl with basket are incised into the paint (Pls. 20, 10; Doumas 1992, figs. 119 and 130). Next in maturity comes the goddess, with long hair loose down her back and a band, but still with shaved area by the ear (pl. 13; Doumas 1992, fig. 126). (It is unlikely that this feature can be explained as part of the band, as in all cases there is intervening hair). Closely following the goddess, come the wounded woman and woman holding a necklace (Doumas 1992, figs. 106 and 102), both with long hair loose but lightly bound in ribbons, with natural hair growing in front of the ear. The oldest of the women are the two from Room 3b, upper level, who wear their long hair tied in snoods (Doumas 1992, figs. 132 and 134). The observations above concerning breast size fit this sequence exactly. There are therefore three main ages, though here the subtle differentiations allow one to distinguish five. In paintings outside Xeste 3, the women from the House of the Ladies (Doumas 1992, figs. 9-10) belong in the middle range (childbearing maturity), while the 'Priestess' (West House; Doumas 1992, fig. 25) would appear to belong to the youngest. The latter, however, is further distinguished by the snake-like lock on her head and the red ear and lips, so that the significance may in this case be ritual rather than belonging to a specific stage of maturity.⁽¹²⁾

Boys are equally well defined, though with a certain degree of ambiguity. The youngest is surely the small yellow-skinned boy from Xeste 3 with his single back lock and short forelock (Fig. 4; Doumas 1992, fig. 112), followed by the 'Boxing Boys' (Doumas 1992, figs. 79, 81), with their immature bodies and shaved heads with long back and sidelocks and shorter top locks. The Xeste 3 boy with large bowl (Doumas 1992, fig. 115) has shaved head with curls growing back on the upper half of the head, the slightly more mature stage though not yet fully grown up. Perhaps the curly-headed acrobat (Doumas 1992, fig. 148) represents the fully but just grown out stage. The somewhat ambiguous figures are the two 'Fishermen' (West House; Pls. 7, 5; Doumas 1992, figs. 18-19, 22-23) and the boy with the cloth (Xeste 3; Doumas 1992, fig. 113; Fig. 4). Both are naked, like the other youths, and both have the shaved head and sparse short locks of pre-pubescence, but their youthful yet fully grown shape and tall frames suggest that they are somewhat older. The seated man (Doumas 1992, fig. 114) is mature, with fully grown hair, a garment wrapped around his waist, and a sagging belly. The belly, however, has been overemphasised in the literature and published drawings, and may as well have to do with his seated but stooped pose. That he, too, is relatively young, is suggested by a hair lock projecting from a topknot over the rest of his black hair, defined by incised wavy lines.

The presence of a double chin appears to be less a definer of age than a stylistic trait, since it occurs on young women as well as older women in the House of the Ladies and Xeste 3. This trait reappears in Mycenaean paintings (Cameron 1978, 591; Televantou 1988, 163 and n.105).

The other indicator of relative age is colour of eye markings (Davis 1986). However, this is not consistently applied, and it may also have to do with the traits of individual artists. It is mainly applied to female figures. The iris is always dark, painted either grey-black (House of the Ladies, 'Fishermen', Xeste 3 Room 3b older women) or brown with a small black pupil ('Boxing Boys', the majority of the Xeste 3 figures, Al acrobat). This is no doubt a genetic/ethnic feature. Interestingly, the girl with basket on her shoulder from Xeste 3 has reddish hair with blue eye (Pl. 10; Doumas 1992, fig. 130). In the white of some, but not all, of the eyes are lines either in blue or (more rarely) red. Blue is used for the youngest and newly mature girls (including the goddess) and the youngest boy in Xeste 3. Red is used for the women of the House of the Ladies and the oldest man (Doumas 1992, fig. 114) and the young man with bowl (Doumas 1992, fig. 115) in Xeste 3.⁽¹³⁾ The two mature women in Xeste 3 (Pl.4; Doumas 1992, figs. 131-134) have no lines. However, they also have no internal ear markings or (in one case) red lips, such as the other women have. Either they are unfinished or by a different artist. Since there are no markings in the eyes of the 'Boxing Boys', the 'Fishermen', or the 'Acrobat', it is a reasonable assumption that only the artists of Xeste 3 and the House of the Ladies used this convention.

Females receive additional colour to parts of the face as indicators of gender: red lips and cheek (House of the Ladies), red lips and ear ('Priestess', West House). Both male and female can have painted fingernails to provide a contrast for the colour of the skin: pink for white-skinned females, white tor red-ochre males (Xeste 3 and 'Fishermen').

Gender, until recently, has been taken to be easily distinguishable in Aegean painting: white for female, red ochre for male. New discoveries and debates, however, have added complexity to the issue,⁽¹⁴⁾ and though there is no ambiguity in any of the Theran figures, one, at least, has become cenual to a consideration of colour conventions. Of the four male figures in Room 3b of Xeste 3 (Fig. 2), the smallest is painted in yellow ochre, several tones lighter than the larger figures (Fig. 4 left; Doumas 1992, fig. 112). Since he is clearly both a (naked) male and the youngest of the children, it is a safe assumption that this is a variation on the convention for indicating immaturity. Children play outdoors, and it is unlikely that boys were simply less exposed to sunlight. More probably, the boy before manhood was considered as being between the feminine and the masculine and hence is shown in an intermediate colour.

The precedent for this variation on a convention can be found in Egyptian painting. There, as we know, male skin is generally red ochre, female yellow ochre. Exceptions occur both for formal reasons (as when overlapping figures need to be distinguished and men are painted alternately dark/light/dark) and for symbolic reasons. Pale skin could signify the difference between Egyptians and foreigners - easterners (Syrians, Hittites, Bedouin) and westerners (Libyans) - southerners, of course, being darker. Interestingly, Aegean men are the only foreigners painted the same colour as Egyptian men. I take this to be less representational (given that they are northerners) than symbolic of affinity. Aegeans are the only foreigners who do not normally appear as enemies of the Egyptians.⁽¹⁵⁾ Skin colour can, however, also distinguish age amongst Egyptians. Old men occasionally have pale skin (Wilkinson 1994, 125), as do young boys. Fig. 5 shows a detail from a Twelfth Dynasty wall painting from the tomb of Antefoker at Thebes (Davies 1920, pl. XXIIIB). As in Xeste 3, the young boy is distinguished not only in size, but by the colour of his skin.

Recognition of this distinction provides an alternative reading of the bull leapers from Knossos. That the white figures, like their dark counterparts, wear loincloths and have male musculature has been pointed out by Nanno Marinatos, who argues that interpreting them as young males, rather than as women, resolves this dichotomy (cf. above, note 14). The pale-skinned bull leaper from Tell el Dab'a (Fig. 6), contemporary with the Theran painting and hence earlier than the Knossos leapers, confirms this assumption. Both red ochre and yellow ochre figures take part in the Tell el Dab'a bull leaping scene, as well as one white figure. Bietak (1994b, 199-200) considers that the coloration signifies distinctions between skills, rather than gender, which is consistent with the view of the lighter figures being adolescent, a view also held by Marinatos (forthcoming b). The question of male initiation has been discussed by Koehl (1986) and Säflund (1987) in relation to the Ayia Triada 'Chieftain's Cup' and 'Sports Rhyton', as well as in connection with bull sports by Nanno Marinatos (1993, 212-220). The issue of colour conventions (though only visible in the paintings) is, I believe, central to such discussion and bears heavily on the Xeste 3 paintings. Distinctions in colour are, I propose, indicative of pre-and post-initiatory status.

       Idiom, convention and the Xeste 3 males -binterpretation

This suggestion now needs to be considered within the context of the Xeste 3 paintings as a whole (Figs. 1-3). The central man (Doumas 1992, figs. 110, 114), whose belly, clothing and fully grown hair with lock show him to be mature, is flanked by two youths and the boy, all of whom are naked and have the shaved heads with locks of youth. The young man holding the cloth and turning his head back (Fig. 4, right; Doumas 1992, fig. 113) is slightly taller than the one holding the bowl (Doumas 1992, fig. 115), and hence, perhaps, slightly older. The young boy (Fig. 4, left; Doumas 1992, fig. 112), considerably smaller and with a firm rounded belly of youth as well as an entirely shaved head with single lock, is significantly paler.

Nanno Marinatos has convincingly argued for an interpretation of the women and girls of Xeste 3 in the adjacent ground floor adyton rooms in terms of female initiation rites (Marinatos 1984, 73-84; 1993, 203-211; cf. Karageorghis 1990). Doumas (1986; 1987; cf. Televantou 1988, 149; 1994, 185, 371-374) cogently proposes that the young males of Room 3b are also initiates, whereas Marinatos suggests that the males are adjuncts to the female scenes, preparing ritual clothing for the females. Yet, as Doumas stresses, there is clearly an architectural parallel in the two sections of Room 3,⁽¹⁶⁾ both of which are enclosed by pier-and-door partitions. This means also that the room in which the male scene takes place, though extremely close, is physically separate from those with the female action, while its relative importance is clear from its size (Fig. 3). Most importantly, the figures converge on the seated male, rather than moving towards the females. Rather than an adjunct to the female initiatory rites, I maintain that the male scene acts as a corollary - an independent but parallel performance with a related meaning. My reasoning for this lies in the observations outlined in this paper on me human figure and idiomatic meaning.

A programmatic parallel is drawn between the scenes in terms of the disposition of figures, their actions and their communicative details (Figs. 1 and 2). In the upper adyton scene, the central figure of the goddess is seated and raised. She faces to the left. In the lower adyton scene, the wounded woman is seated (though not raised) and faces to the right. In Room 3b (lower level), the central man is seared and slightly raised (his head almost touching the upper border bands), and faces to the right. In all three cases, figures of the same gender converge from left and right towards the central figure. These figures are distinguished from one another according to age, displaying a range from pre-puberty to young adult. All bring or carry an object, either a container (basket, bowl) or something for the body (cloth, necklace). The girls' baskets contain crocuses as offerings, which other girls have collected, crocuses being the component of the dye saffron used for cloth. The boys bring liquid containers⁽¹⁷⁾ and a cloth, while the central man holds a large water jug -all components which, though they may have other functions, are used in the process of dyeing cloth. The young girl approaching from the right towards the wounded girl seated on a rock wears and holds out a cloth like a veil, but turns her head backwards, looking over her shoulder. The young boy approaching from the left towards the central man seated on a rock(?) holds out a cloth and turns his head backwards, looking over his shoulder. The heads of those two figures, mostly shaved with locks on top and at the side, represent the same stage of maturity. Such formal and idiomatic analogies would suggest to me parallel significance in the rituals depicted. Both have a connection with cloth and call to mind the iconographic cycle of ritual robing and associations between the goddess and saffron (discussed in Morgan 1990; and see Warren 1988, 20-23; Niemeier 1986, 78-84; Marinatos 1993, 141-145).⁽¹⁸⁾ Male/female counterparts and corollaries are ser up: seated figure, bringing cloth with head turned, bringing containers, varied stages of maturiIy. It would seem that, rather than being adjuncts to the female scene, the males are engaged in rites of their own, and, given the careful attention to stages of youth and maturity, that these too are connected with puberty and initiation.

CONCLUSIONS

The aim of this paper has been to show how form and meaning are inextricably linked. This is so on the level of individual form (the details and disposition of individual figures) and on the programmatic level in the directional flow of figures within the architectural space. The cumulative effect of analysing formal meaning leads through into iconographic interpretation of scenes within the context of their programme.

(1).     The question is complicated by the fact that Egyptian painters sometimes hybridised Aegeans with Asiatics. For illustrations and discussion, see Vercoutter 1956; Wachsmann 1987. Up to date bibliography on the subject is provided by Rehak (1996, 36 n.4).

(2).     Raising both hands is a universal gesture in art with strikingly different contexts and meanings: Demisch 1984. On the Egyptian adoration gesture: Wilkinson 1992, 28-29 (A30); Betro 1996, 38. On worship and supplication in the Near East: Gruber 1980, ch. II. In royal or sacred imagery, raised arms can also indicate power: Schapiro 1973, ch. 2.

(3).     Recent studies by Christina Televantou have considered the human figure in Theran art with particular attention to stylistic traits identified as characteristic of individual artists; Televantou 1988; 1992; 1994. See also Doumas 1985. For the human figure in Aegean art, cf. Tamvaki 1989; and on issues of power, relative scale and the ideal male, Crowley 1995; Marinatos 1995a. 

(4).     For a close analysis of the Egyptian figurative conventions see Schäfer 1974. On proportions of the human figure, as set out on the static standing figure, see also Iversen 1975; Davis 1989; Robins 1994. A recent article by J. Weingarten (1995) discusses the possibility of an Aegean canon in relation to Egyptian proportions. On Theran and Egyptian conventions see Bietak here, vol. I.

(5).     These figures are further discussed by Davis (1995), Koehl (1995) and Marinatos (1995a). On the sceptre as a transference of religious authority from the divine to the human sphere, see Krattenmaker 1995 (Minoan), Palaima 1995 (Mycenaean) and Younger 1995, 156-162, II.A catalogue nos. 1-34.

(6).     The hieroglyphic forms for 'to offer' are all derived from the shape of incense burners with elongated arm-like handles and a cupped hand on the end: Gardiner 1957, 454 D37-39; Wilkinson 1992, 52-53; Betto 1996, 59. Therange of offering gestures in Egyptian art, however, is more varied and depends on the shape of the object offered.

(7).    On the Theran paintings: Televantou 1982 (costume); 1984 (jewellery); 1988 (human figure); 1994; Morgan 1988, 93-103. Recent studies which include the paintings: Barber 1991; Kontorli-Papadopoulou 1996, 86-97; Rehak 1996, 46-48; Tzachili 1997; Younger 1992. On hair, see below.

(8).     On directional flow in glyptic scenes, cf. Sourvinou-Inwood 1989, 249f.. Here the issue of whether direction is 'correct' on the original or the impression intrudes. The significant side of a deity or ruler is taken to be the figure's right, processional movement to flow from the right field to the left. We are, then, looking at the original image on a ring. Niemeier (1989) illustrates the impressions of rings as the significant viewpoint. This corresponds with the predominant left-right flow of Aegean wall paintings.

(9).     The entrance to this room on the upper storey, where the paintings were, was either on the narrow east wall from Room 2 (Marinatos 1984, fig. 65) or from the south wall at the eastern end (Doumas 1992, fig. 1).

(10).    Doumas (1987, 155) states that the seated adult occupied the south end of the west wall and assumes that the fourth figure was also on that wall.

(11).    Televantou (1994, 162) thinks that both figures are dancing, that is whirling, and were painted by the same artist (Televantou 1992, 152).

(12).    Neither Doumas (1992, 47) nor Televantou (1994, 175) think this is necessarily a priestess. On the red lips, see Sapouna-Sakellaraki 1981, 501; Marinatos 1987, 30-31.

(13).    The eye of the boy with cloth (Doumas 1992, fig. 113) is not preserved.

(14).    Debate has centred on the gender of white-skinned bull leapers: Indelicato (1988,44) argues that white coloration for male figures is "a new device to suggest temporal sequence through colour contrast". However, N. Marinatos (1989, 28-32; 1993, 219-220; 1994, 89-94; 1995b, 579 and n.17) cogently argues that the white-skinned bull leapers should, judging by their musculature and clothing, be male rather than female. Rehak (1996, 41) notes the sexual ambivalence of the bull leapers - red and white skin, long female locks, masculine musculature - and suggests they represent "gender-neutral παιδία". Such figures are known chiefly from Knossos, but also from Tiryns (initially thought by Evans (1901-1902, 74) to be a male youth), and the Ramp House at Mycenae (Shaw 1996, 175-176, there considered to be female). Compare the controversy surrounding the pale skin of the male torso of the so-called 'Priest King' stucco fragments (Cameron 1970, 163-166; Waterhouse 1974, 153-154). For the view that conventions for skin colour in ancient art were less rigid than have been assumed, see Morgan 1995b, 42-43.

(15).    Exceptions, notable for their rarity, are restricted to the reign of Amenophis III: two in tomb paintings showing throne bases with bound captives (TT 93 (Kenamun) and TT 120 (Anen)) (Vercoutter 1956, 226-227, 287, 293, pλs. XII:116, XXVI:190, XXXIV:230; Wachsmann 1987, 38-40, pls. XLVI-XLVIII, LA, col. ill. (Anen); Robins 1997, 137, pl. 155 central bound figure), and one on one of the statue bases of Amenophis III from Kom el-Hetan at Thebes, which is inscribed with Aegean name rings (Cline 1995, pl. 6; Wachsmann 1987, 95-99, pl. LXVIII). (It has been argued that these place names represent an itinerary of an official Egyptian visit to the Aegean: Hankey 1981, 45-46; cf. Cline 1987). In all three examples, the prisoners depicted demonstrate the lack of iconography for Aegean captives, since their beards, hairstyles and clothing identify them clearly as Asiatics.

(16).    Beneath the floor of a part of Room 3b was a large hole, affording access to a lower level comparable to the descent in the adyton (Doumas, personal communication).

(17).    The shape of the container held by the boy on the right (Fig. 2) is metallic, and its yellow colour suggests bronze or gold, in contrast to the terracotta-coloured water jug held by the mature man. (The vessel held by the young boy on the left is nor well preserved, but also looks metallic.) Clearly, these were not used for washing. Saffron, in its use as a dye, could be held in metal containers, which would be less likely to stain than clay. (I am grateful to Mr M. Springett of the Chemistry Laboratories at Cambridge University for his advice on this.)

(18).    Televantou (1994, 372) suggests that both scenes are concerned with the delivery of sacred clothes. Tbe cloth brought by the boy is, of course, not yellow, though the veil of the girl is, as is the robe of the West House 'Priestess'. E.J.W. Barber (1991, 338 n.13) suggests that yellow cloth may have been reserved for "a particular kind of women and their associates". I am reminded of the saffron robes of Buddhist monks. It is unclear what role the boys take in relation to cloth and saffron or, indeed, the goddess, but the paintings imply a connection between these and both male and female rites of passage. It should be noted that the dress in the robing scene of the House of the Ladies is - like the cloth held by the boy - multicoloured.

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

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