Syrian Glyptic and the Theran Wall Paintings

Syrian cylinder seals of the eighteenth and seventeenth centuries BC have been found at sites in the Aegean and are the principal surviving vehicle for the spread of iconographical motifs in what was evidently a lively maritime trade involving a variety of different commodities. Perishable or recycled materials such as textiles, wooden objects and metalwork must also have played an important part in the transmission of motifs but the evidence no longer survives. That motifs could be transferred from one medium to another is attested by the appearance of similar designs on both Kamares ware and seals.

The main surviving Near Eastern vehicle for the spread of iconographical motifs is the cylinder seal which was probably developed in southern Iraq or south-western Iran soon after 3500 BC, was adopted shortly before 3000 BC in Syria and is attested there throughout the third millennium. The quality and originality of Syrian seals during the Early Dynastic period subsequently declined; although fine Syrian seals are known from impressions from Ebla, surviving seals of the late third and early second millennium BC are mostly relatively crude and confined to stones such as chlorite which could easily be cut with hand-held tools (Collon 1987, 13-25, 38-55). However, when about 1920 BC Assyrian merchants established trading colonies in central Anatolia, Syria was drawn into a trade network of increasing complexity, and by the end of the nineteenth century the eastern part of the country had been absorbed into the Upper Mesopotamian kingdom of Shamshi-Adad I of Assyria (ca. 1809-1776).⁽¹⁾

The lingua franca of the Near East during the first half of the second millennium BC, as later during the Amarna age in the fourteenth century BC, was Akkadian, written in cuneiform on clay tablets. Together with this form of writing the countries surrounding Mesopotamia adopted or continued to use the cylinder seal. Tablets have been found at a large number of sites, the most important being Mari on the middle Euphrates with huge archives of cuneiform tablets. Numerous seals and sealings have also been found and they document the beginnings of a distinctive elite Syrian style on hard stones, especially haematite which was the main material for seals during the first half of the second millennium throughout the Near East. For the next two centuries, the independent and semi-independent kingdoms of Syria had highly literate administrations and their officials and rulers owned some of the finest and most distinctive seals of the period; they were the technical equals of the best Babylonian products but avoided the stereotyped and repetitive scenes of much Babylonian glyptic. By sea Syria was open to influences from Egypt, and Egyptian motifs appear on many Syrian seals, and are especially evident in the products of the Green Jasper workshop which may have been located in the harbour town of Byblos (Collon 1986). To the east, Syria was linked by a long-established trade route along the foothills of the Zagros with the Diyala kingdom of Eshnunna and the city of Susa in south-western Iran. There is also glyptic evidence to suggest that it had contact with Margiana east of the Caspian, on the route to a possible source of tin in northern Afghanistan (Collon 1987, 44). To the south, Syria was linked to Babylonia by a major trade route along the Euphrates and had especially close ties with the city of Sippar near Baghdad. It is probable that Syrian glyptic was much influenced by that of Sippar.

'Aegean' motifs have long been recognised in Syrian glyptic and Henri Seyrig published an important article surveying the material (Seyrig 1956). However, before the publication of the seal impressions from the Alalakh archives, Syrian seals lacked a chronological framework and were dated far too late (Frankfort 1939, 260-273). It is now apparent, however, that Syria was using some 'Aegean' motifs long before they are attested in the Aegean and that the influence worked the other way. This, as Beatrice Teissier has demonstrated, is also true of some 'Egyptian' motifs (Teissier 1996).

The Alalakh archives were excavated by Sir Leonard Woolley at Tell Atchana on the Syrian-Turkish border (Woolley 1955). They were found in two successive palaces but it is only the earlier archive which is relevant here. The Level VII palace was the seat of governors who were cousins of the kings of Iamhad, the rulers of nearby Aleppo; the latter seem to have used Alalakh as a summer residence. Certainly the tablets are part of a royal archive and were encased in sealed clay envelopes many of which survive. The archive can be dated between ca. 1720 BC and the destruction of Alalakh around 1650 BC (middle chronology), and the associated seal impressions therefore fall within this time-span and can often be attributed to a specific reign on the basis of their inscriptions (Collon 1975). These impressions have provided the chronological framework which had previously been lacking and excavations at other sites have served to confirm this framework.

Since any evidence for influence of Syrian glyptic on Thera wall paintings has to be shown clearly to antedate these paintings, the evidence which follows will be based on well-dated groups, such as the Alalakh seal impressions, and on seals which can be related to them stylistically. The following features of Syrian glyptic are relevant to the Thera wall paintings:

FLYING GALLOP

Among the Alalakh Level VII seal impressions was the fragment of an envelope bearing the partial impressions of two cylinder seals. One (Fig. 1a right), belonging to a scribe, was typically Old Babylonian in style but for the addition of an Egyptian ankh between the figures of the king with a mace and suppliant goddess. The other (Figs. 1a left, 1b-c), according to a cuneiform bi-script below it, belonged to a servant of the Hurrian goddess Hepat. It shows a bearded figure standing to the left of a stylised tree above which was a winged sun-disc - a motif derived from Egypt; only the left part of these motifs and the upper part of the figure survive. The figure wears a headdress which is a variation of the high-brimmed headdress of Babylonian kings, but his garment has the thickly rolled borders (perhaps fur-trimmed) of the Syrian elite and he is probably a king. Behind him is a separate scene with a bull charging towards the right and supporting two athletes who are doing handstands on its back on either side of a round-handled ankh. Above is an ibex facing left in flying gallop with a griffin(?) and lion on either side. The bull leaping scene, which antedates Cretan examples, has been discussed elsewhere in the context of the Tell el-Dab'a wall paintings (Collon 1994), but since bull  leaping scenes are not so far attested on Thera, it is not relevant here. The flying gallop is, however frequently depicted on the Thera wall paintings and its evolution in Syria can be demonstrated.

When I first published the Alalakh bull leaping scene I related it to an unprovenanced seal which was once in the Erlenmeyer collection in Basel (Fig. 2) and suggested that the two seals were "probably by the same hand" (Collon 1975, 61 n.4). I have since revised my opinion and believe the more slender proportions of the athletes on the Alalakh impression to be evidence of its later date: it was probably made in the same Aleppo workshop (Collon 1982a) around 1700 or in the early years of the seventeenth century (the date of the bulk of the archive), whereas the Erlenmeyer seal, with its more solid proportions, would have been a product of the workshop in its heyday during the reign of King Abban of Alalakh (whose own seal was, I believe, made in the workshop) around 1720 BC. The Erlenmeyer seal has the identical scene of two athletes doing handstands on the back of a bull but with the addition of a third athlete (a catcher or bull leaper landing) in front of the bull. Above the bull is another in flying gallop towards the left. Further seals with the flying gallop (Figs. 3-5) are probably also early products of the Aleppo workshop and Fig. 6 comes from a related workshop.

A seal from the Bibliotheque Nationale collection (Fig. 7), of about the same date and also attributable to the Aleppo workshop, shows a figure, known from Mesopotamian iconography as the 'nude bearded hero with curls', fighting with a bull: he holds the bull upside down, grasps it by one hindleg and its tail, and places his foot on the bull's neck. The motif is well attested on earlier Mesopotamian glyptic (e.g. Amiet 1980, no. 1471), and is frequently found on Syrian seals; this is a particularly powerful rendering of the scene. If, however, it is turned through ninety degrees, it will be seen that the bull is very close in posture to the ibex in flying gallop on the Alalakh seal impression (Fig. 8). The same motif could be executed vertically or horizontally, although in the latter case the front legs were generally extended, as on Figs. 2-5 and as on one of the Thera wall paintings (Doumas 1992, figs. 30-33).

ATHLETES

Several of the Syrian seals referred to above (Figs. 1-4, 7) show a slim young athlete; he is quite different from the traditional Near Eastern nude bearded hero who is a heavyweight wrestler (Fig. 7). The athlete wears a thick headband which may have streamers although it seems more likely that it is the athlete's hair which hangs down behind: in a single plait when he is bull leaping (Figs. 1-2) and in several strands otherwise. He is naked except for a single belt and, perhaps, a rounded split 'breech' kilt similar to that worn by the athletes on the Tell el-Dab'a wall paintings and by Cretan athletes (Bietak 1994; Buchholz and Karageorghis 1973, nos. 1162, 1166, 1225-1226, 1230). Unlike his Cretan counterparts, however, his feet are bare and in this he resembles the boxers from Thera (Doumas 1992, figs. 79-80).

Similar athletes to those just described are shown fighting on Syrian seals belonging to different stylistic groups but probably also to be dated to the late eighteenth or seventeenth centuries BC (e.g. Collon 1987, no. 705). The closest parallel to the Thera boxers in terms of posture is on an actual cylinder seal from Level VII at Alalakh (Fig. 9). It is extremely coarsely engraved and shows kilted boxers which are quite different from the lithe athletes previously discussed, but they resemble the Theran boxers in that they seem to be aiming, or even grasping, at each other's faces. There is no evidence, however, for the single boxing glove which the Theran boxers wear.

GRIFFINS

The Syrian bird-griffins antedate any from Crete or Mycenae. They are a manifestation of the predilection of the Near East for combinations of the most powerful creatures known, best exemplified by the winged human-headed lions and bulls of Assyria in the first millennium BC. The earliest of these composite creatures is the Mesopotamian lion-headed eagle which is first attested in the late fourth millennium BC; by the second millennium, Syria had developed the griffin, a creature with a rounded bird's head generally with a long curl, a hooked beak, wings and a leonine body. It does not usually have the bent wings of the Thera bird-griffins (Doumas 1992, figs. 122, 128) but this feature does occasionally appear (Fig. 6); nor does the small scale of the seals allow for the depiction of the distinctive spirals and triangles found on both Thera examples. The curl on the top of the heads of most Syrian griffins falls in separate strands along the necks of the Thera griffins. Although generally shown seated on their haunches, the Syrian griffins are also often depicted standing on their hindlegs in a posture not unlike the flying gallop of one of the Thera examples (Figs. 6 and 10); unfortunately the other Theran griffin is fragmentary but it too seems to be standing on its hindlegs in much the same relationship to the seated figure as the griffin is to the kneeling figure on an actual seal excavated at Tell Chagar Bazar (Fig. 10) - a product of the North Syrian workshop (Collon 1985). A closer parallel is provided by a seal in Vienna (Collon 1987, no. 315) but this is the product of a later period and as such shows the influence of Aegean wall paintings on Syrian glyptic.

GUILLOCHES

The guilloche is a typical feature of Syrian seals (Collon 1975, 193-194) and is also found on the Mari 'Investiture' wall painting of ca. 1775 BC (Fig. 14). It probably represented water and, by extension, fertility generally, as convincingly argued by Henri Seyrig (1960, 240 n.1 and pl. IX:12; see also Collon 1975, 193-194). It is used as a vertical (Fig. 11) or horizontal divider, as a frame above and/or below a scene (Fig. 6) or around it (Collon 1975, nos. 106, 108-109). Generally it consists of three or four strands and it is often built round central dots or circles. Both S and Z coils are found (i.e. coiling towards the right or the left). Complex triple or quadruple many-stranded plaits also frequently appear (Fig. 6). The guilloche was particularly favoured on seals of the Aleppo workshop Collon 1982a) (e.g. Fig. 7) and Green Jasper workshop (Fig. 11) (Collon 1986). The latter, as mentioned above, may have been located at Byblos, and certainly supplied an international clientele: its products have been found westwards on Cyprus (Klavdia and Kition), Crete (Poros) and, in a Punic context at Carthage. The piece illustrated on Fig. 11 can be dated to ca. 1800 thanks to the name of its owner, a ruler of Buzuran, who is mentioned in the Mari archives. The texts indicate that Buzuran was close to Mari, and the seal must have been acquired as an exotic piece to enhance its owner's prestige.

Probably for technical reasons, the guilloche does not appear on Syrian seals before about 1800 BC but related motifs, often in the form of entwined snakes, are already found on some Uruk cylinder seals of the late fourth millennium (Collon 1987, no. 9), on Early Dynastic cylinder seals of the middle of the third millennium BC (Amiet 1980, nos. 1068, 1079, 1099, 1100, 1247-1253, 1297, 1320, 1735), and are attested on Near Eastern Early Bronze Age artefacts (Parrot 1961a, 139-140, figs. 168B and 169A from Mari, probably of Iranian manufacture but made for the Syrian market; and ibid. 138 fig. 167, the plaque of Dudu from Lagash in Sumer). The use of cylinder seals, many with spiral designs, on storage jars and hearths in the Argolid during the Early Bronze Age (for example at Lerna, Tiryns and Zygouries), seems to have derived from Levantine practice (Collon 1987, 141-142). Syrian seals influenced the use of the guilloche, running spirals and complex plaits on Anatolian seals (Boehmer and Güterbock 1987, esp. 33ff.).

There is, however, a long tradition of spiral motifs in the Cyclades which can be traced back to the early third millennium BC and perhaps even earlier. The designs on the famous 'frying pans' (Buchholz and Karageorghis 1973, nos. 855-856), however, are overall patterns linked in a number of different directions. They are the antecedents of the patterning on the ceremonial axe from Mallia (Buchholz and Karageorghis 1973, no. 232) and of a seal design from a later level at Alalakh (Collon 1975, no. 235), but not of the sober borders of the Thera wall paintings (Doumas 1992, figs. 85-86, 90, 93-94 but cf. also figs. 43 (on a boat) and 138 (on clothing)). These latter are closer to Syrian seals (Fig. 6) and to the wall paintings of Mari on the middle Euphrates, notably on the wall painting referred to above depicting the investiture of King Zimri-Lim (Fig. 14), where they are used in a similar way (Muller 1994, 50 and fig. 1). Contacts between Mari and the west are well attested in the Mari correspondence, Zimri-Lim (reigned ca. 1775-1760 BC) was the son-in-law of Iarim-Lim I of Iamhad, at whose court he had been exiled before recovering his family's throne, and the Level VII archive at Alalakh was that of Iarim-Lim's descendants, from his grandson Abban onwards. However, these Theran examples are different in execution in that some of the strands are continuous (Doumas 1992, figs. 57 and 62, and cf. figs. 56 and 59). On the other hand, most of the single-strand guilloches are not continuous and are formed round a central circle like the Syrian ones.

MOTIF WITH VOLUTES

If a motif with volutes on a fragmentary wall painting from Mari (Fig. 12a)belongs to the earlier phase of the palace's decoration, it may antedate by as much as 450 years the Thera wall paintings with similar motifs decorating ikria (Doumas 1992, figs. 49-60, 62). However, its close parallel on a fragment of wall painting from Knossos (Fig. 12b) may indicate that it belongs to the later phase of wall paintings which ended in the sack of the palace by Hammurabi of Babylon in 1760 BC. There is an excellent summary of the comparative techniques and iconography of Near Eastern, Theran and Aegean wall paintings in two articles by the same author writing under different names (Pierre 1987; Muller 1994). A plant with volutes appears on Syrian cylinder seals of the late eighteenth century (Figs. 5-6) and on an impression from Alalakh (Collon 1975, no. 76); it may have had some impact on the decoration of Theran ikria. The strings of beads and volutes which also decorate the ikria may have inspired the work of jewellers; indeed, the volute motif is common in Mycenaean jewellery (Buchholz and Karageorghis 1971, 110 fig. 37, nos. 1320, 1328) but does not seem to be attested earlier.

SEASCAPES

Two seals demonstrate the Syrian predilection for depicting scenes with many boats interspersed with mythological motifs. One is probably to be dated to the middle of the third millennium BC and, though unprovenanced, is probably north Mesopotamian or Syrian (Amiet 1980, no. 8564 (drawing); Collon 1987, no. 715 (photograph)). The presence of Egyptian motifs on the other (Fig. 13) is a clear indication of its Syrian origins. Although the likelihood of any influence of such seals on the subject matter of a wall painting on the island of Thera (Doumas 1992, figs. 35-43) is remote in the extreme, it is possible that the seals reflect the existence of mainland wall paintings with nautical themes which could have antedated the Theran wall paintings and might have influenced them. An Assyrian relief of the reign of Sargon II (721-705 BC) in the Louvre showing Phoenician ships transporting cedar logs along the Levant coast in a sea full of fish and mythical creatures is perhaps a late echo of this tradition (Parrot 1961b, 40 fig. 48).

SEALS AND THE TRANSMISSION OF MOTIFS

It seems, therefore, that Syrian motifs may well have been transmitted west through the intermediary of seals. Numerous examples of Near Eastern cylinder seals have been found in various contexts of varying date throughout the Aegean and Greece as the volumes of the Corpus der minoischen und mykenischen Siegel (Berlin 1964-) testify. One particularly telling piece of evidence is the appearance on Kamares ware of a floral motif (Buchholz and Karageorghis 1973, no. 890 from Phaistos and col. pl. facing p. 9) which is also found on a Syrian seal impression from Level VII at Alalakh (Fig. 15) and on a seal from Carchemish (Fig. 16). Furthermore, a sherd of Kamares ware with this design was actually found at the Levantine site of Ugarit (Fig. 17; Buchholz and Karageorghis 1973, 69 fig. 27, no. 897) and other seals of this type occur in a possible sixteenth century context at Ugarit (Figs. 18-20). It could be argued that the Kamares pottery, traditionally dated between 1900 and 1700 BC, influenced the seals, but this would be missing the point. What this sherd demonstrates is the transfer of designs from one medium to another in what was evidently a lively maritime trade involving a variety of different commodities. Motifs were probably also transmitted on textiles and wooden objects, which have not survived, and on metalwork which has only occasionally survived. As stated in the opening sentence of this paper, cylinder seals are the main surviving vehicle for the spread of iconographical motifs.

(1). Absolute dates in the Near East at this period are based on synchronisms with Babylonian chronology. There are three possible sequences of dates for the first half of the second millennium BC, based on recorded sightings of the planet Venus during the reign of the Babylonian king Ammisaduga. His dates, according to the high, middle and low chronologies, are respectively 1702-1682, 1646-1626 and 1582-1562 BC. The dates used here are according to the Babylonian middle chronology which is that most generally used. However, there is nothing in the seal evidence which would preclude the adoption of a low chronology, indeed rather the reverse (a high chronology would be much more of a problem). The reader wishing to adapt the dates used in this paper to the low chronology should deduct 64 years from them.

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

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