BIAA
&WT15H INSTITUTE AT ANKARA
lr Sen
Carchemish sa kisad puratti Author(s): Irene J\ Winter Reviewed work(s):
Source: Anatolian Studies, Vol. 33, Special Number in Honour of the Seventy-Fifth Birthday of Dr. Richard Barnett (1983), pp. 177-197 Published by: British Institute at Ankara Stable URL: http://www.jstor.org/stabTe73642708 Accessed: 10/11/2011 05:23
Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at http://www.jstor.org/page/info/about/policies/terms.jsp
JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of content in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms of scholarship. For more information about JSTOR, please contact support@jstor.org.
	
9	
STOR	
British Institute at Ankara is collaborating with JSTOR to digitize, preserve and extend access to Anatolian Studies.
http://www.jstor.org
CARCHEMISH $A KISAD PURATTI*
By Irene J. Winter University of Pennsylvania
Perhaps no other site in the region of northern Syria and south-eastern Anatolia played as important a role in the history of the early first millennium B.C. as Carchemish, "on the banks of the Euphrates." It is one of the best-documented sites of the period, due to a combination of Neo-Assyrian references and the excavated material of the site itself, including inscriptions, reliefs and large-scale architectural projects initiated by the rulers of Carchemish. All of these documents attest to its immense wealth and power.
The site was first explored in the 1870's on behalf of the British Museum, once George Smith had determined that the modern town of Djerabis must be ancient Carchemish; and was subsequently excavated and published under the Museum's auspices.1 Several encyclopedic compendia published in recent years have summarized in cogent syntheses the information known about Carchemish.2 Nevertheless, I would like to include this present review of the material as a tribute to Richard D. Barnett - whose own work has been closely associated with the site in particular and with North Syria in general — in order to add a few points regarding the nature of Carchemish and the role played by the state in the history and art-history of the times.
I propose to investigate the following issues: first, the role of Carchemish as a centre of art production and of craftsmen, relevant not only to its own works, but to those from other sites; second, the role of Carchemish as an economic and trading centre, especially with regard to the acquisition of metals; and third, the role of Carchemish in relation to the western expansion of the Neo-Assyrian empire and the determination of Assyrian military strategies from the 9th through the late 8th centuries B.C.
The modern Turkish/Syrian border, established along the line of the Ottoman railroad to Baghdad, runs between the twin villages of Kargamis, to the north and Jerablus to the south, bisecting the ancient site of Carchemish. Its hinterland, the plain of Jerablus, extends some 20 miles to the south, with small side valleys off to the west that rise sharply into the hills of the Kurd Dagh.3 The area is extremely fertile, if spatially limited; watered not only by the Euphrates, but also
*My sincere thanks to J. D. Hawkins, R. C. Hunt and H. Tadmor, for comments upon a reading of the first draft of the present study. Photo Credits: Pis. XLIV, XLVI, XLVII, XLVIII, XUX(a,b,d), Courtesy of the Trustees, The British Museum; Pi. XLV(a), Courtesy of M. Pierre Amiet, Conservateur-en-chef, Departement des antiquites orientales, Musee du Louvre; PI. XLV (b), courtesy of J. D. Hawkins, Esq.
'D. G. Hogarth, Carchemish, Part I: Introductory, London, 1914; C. L. Woolley, Carchemish, Part IT. The Town Defenses, London, 1921; C. L. Woolley and R. D. Barnett, Carchemish Part III: The Excavations in the Inner Town, and The Hittite Inscriptions, London, 1952 (henceforth abbreviated Care. I, II, III.)
2 Ranging, e.g., from "Carchemish", in C. F. Pfeiffer, ed., The Biblical World: A dictionary of Biblical archaeology, Grand Rapids, MI, 1966, pp. 165-9 to "Karkamis," by J. D. Hawkins, in RLA V: 5-6, Berlin, 1980, pp. 426-46.
3 See panoramic view of the plain, taken from the citadel of Carchemish, in Care. II, frontispiece.
178
ANATOLIAN STUDIES
by the river's only significant western tributary, the Sajur.4
The mound of Carchemish commands the entire plain, consisting of a high citadel and an extended lower town (PI. XLIV, a, b). According to British surveys, there are more than 30 additional mounds in the valley.5 Several mounds have been shown to contain first millennium material, although none rivals Carchemish in size or strategic position. These would probably correspond to the "cities of Carchemish" mentioned in Assyrian sources - the towns and villages of the surrounding territory under the dominion of the capital.6
The location of Carchemish at a major east-west crossing of the Euphrates, as well as its position on the Euphrates, assured the site's importance as a principal thoroughfare in the ancient world. Easy access to the west was possible either via Gaziantep/Killiz to the northwest, or via Aleppo to the southwest. Communications between Carchemish and Aleppo are possible by several routes, the most preferable taking a southerly road out of the plain, passing through the towns of Membidj (Assyrian Nappigu/Nanpigi, classical Hierapolis) and El Bab. To the northwest, one would most likely follow the Sajur to near its source at Gaziantep, located on a large plain between hills to east and west, and thence up to Maras at the head of the Kara Su valley. From that point, one has access across the Amanus or due north into the Taurus and onto the Anatolian plateau.7
This combination of location on the river, radial routes to and from Carchemish and fertile hinterland placed the city/state in the most favoured of positions, at the hub of virtually all essential activities and movements in the period. Thus, Carchemish meets all of the requirements of modern economic geography in the location of a "central place": a high coefficient of importance in relation to the surrounding territory, control of or dominance over auxiliary towns and villages, necessary arable land and pasturage to support a concentrated population, and routes of access to major resources as well as to other central
4 This smaller river rises near Gaziantep and enters the plain at the northwest, to join the Euphrates further south - just opposite Tell Ahmar, ancient Til Barsib (cf. sketch map in Care. II, Fig. 5).
^Hogarth, D. G., "Carchemish and its Neighbourhood," LAAA 7/(1909), 165-89.
6D. D. Luckenbill, Ancient Records of Assyria and Babylonia, Vol. I, Chicago, 1926, §651. (See also, recent speculation that the territory controlled by Carchemish may even have extended to Tünp, south of Gaziantep on a tributary of the Sajur: J. D. Hawkins and A. Morpurgo-Davies, "Buying and Selling in Hieroglyphic Luwian," in Serta Indogermanica: Festschrift für Günter Neumann, J. Tischler, ed., Innsbruck, 1982. p. 92.)
7For topographic details, I have referred to the US War and Navy Department Agency, Army Map Service Sheets, nos. 117470-475, Washington 1945 (Syria) and to the map of Prof. F. Sabri Duran, published by Kannat Kitabevi, Istanbul, 1951 (Turkey). For the Aleppo route, cf. R. Dussaud, Topographie historique de la Syrie antique et medievale, Paris 1927, and J. Edde, Geographie de la Syrie et du Liban, Beirut 1931. For the Killiz-Gaziantep route, see H. H. von der Osten, Explorations in Hittite Asia Minor, 1929 [O.I.C. 8] Chicago 1930, and A. Archi, P. E. Pecorella & M. Salvini, Gaziantep e la sua regione, Rome 1971. According to personal communication from M. V. Seton-William in 1972, there is a large mound at Killiz that has never been excavated.
CARCHEMISH $A KI$AD PURA TTI
179
places, such that conditions for a viable economic life may be demonstrated.8
Because very few contemporary documents explicitly refer to the economic and political life of Carchemish, the role played by the state in the first half of the first millennium B.C. must be reconstructed from archaeological evidence, previously-attested historical references, mention in Assyrian sources and modern inference. Of these, Neo-Assyrian accounts of interaction between the two states, from the reign of Assurnasirpal II (883-859 B.C.) to the conquest of Carchemish by Sargon II (in 717 B.C.), provide the most direct information.9 The relatively large number of 9th—8th century texts from Carchemish, by contrast, written in hieroglyphic Luwian and inscribed on stelae and architectural orthostats, were mainly for display purposes: providing titles and attributes of patron-rulers, but neither annalistic nor event-full.10
Of the archaeological remains at Carchemish, several building phases can be distinguished, associated with particular rulers, into which the sculpture, reliefs and inscribed slabs can be placed.11 The best reconstruction for the 10th—9th century reliefs — incorporating textual and epigraphic evidence as well as stylistic arguments — places the material of the Water Gate earliest in the sequence; followed by the Long Wall of Sculpture attributed to Suhis II; and then, the King's Gate associated with his son, Katuwas. The reliefs of the uninscribed Herald's Wall have been variously dated; I would prefer to see them contemporary with or just before the Long Wall of Suhis; Genge would bring them down to Katuwas.12 The Royal Buttress and several miscellaneous pieces are then to be associated with Yariris and Kamanis, regent and son respectively of one Astiruwas, founder of a new dynasty prominent in the first half of the 8th cen-
8Primary studies: W. Christaller, Central Places in Southern Germany, Englewood Cliffs, N.J., 1966 (originally published in German, in Jena, 1933) and A. Lösch, The Economics of Location, New Haven, 1954 (originally Jena, 1944). More recently, see P. Haggett, Locational Analysis in Human Geography, London, 1965, B. J. Berry, Geography of Market Centres and Retail Distribution, Englewood Cliffs, N.J. 1967, and R. J. Chorley and P. Haggett, Network Analysis in Geography, New York, 1970. For applications to earlier, historical situations, see G. W. Skinner, "Marketing and Social Structure in Rural China," JAS 24 (1964) 195-228 and 363-99, as well as G. A. Johnson, "A Test of the utility of Central Place Theory in archaeology," in Man, Settlement and Urbanism, eds. P. J. Ucko, R. Tringham and G. W. Dimbleby, London, 1972, pp. 769-85.
9 See complete list of references in Hawkins, RLA, V/6, pp. 441-2.
10 For the inscriptions from Carchemish, see J. D. Hawkins, "Building Inscriptions of Carchemish," Anat. Stud. XXII (1972) 87-114; id., "Some Historical Problems of the Hieroglyphic Luwian Inscriptions," Anat. Stud. XXIX (1979) 153-167; id., "Kubaba at Karkamis and Elsewhere," Anat. Stud. XXXI (1981) 147-175, and summaries in id., RLA V/6, pp. 442-6.
11 In addition to the primary publication of this material, see important review article of Care. III: H. G. Güterbock, "Carchemish", JNES XIII (1954) 102-14; also recent detailed studies of the sculpture and reliefs by Orthmann and Genge: W. Orthmann, Untersuchungen zur späthethitischen Kunst [Saarbrucker Beiträge zur Altertumskunde, Bd. 8] Bonn, 1971; H. Genge, Nordsyrisch-sudanatolische Reliefs: Eine archäologisch-historische Untersuchung, Datierung und Bestimmung, Copenhagen, 1979; henceforward Orthmann, USK, and Genge, NSR. See also summary of monuments and their dating in Hawkins, RLA V, pp. 439-41, and reviews of Orthmann by Winter, JNES 34 (1975) 137-42 and Hawkins, ZA 63 (1974) 309 10.
12Genge, NSR, Ch. IV:1, p. 57. It is curiously disturbing that the name of such a wealthy king as Sangara, attested in Assyrian tribute lists from at least ca. 870 when first mentioned
180
ANATOLIAN STUDIES
tury; and some later inscriptions and reliefs have been suggested by Hawkins to belong to the reign of Pisiris, king at the time of the conquest of Carchemish by Sargon of Assyria.13
The extent of the building programme is particularly impressive. Within a walled Inner Town covering some 2Vi square miles, only a small fraction of the total area has been excavated.14 This consists of the "hilani", a large, mainly unexcavated building to the east of the elaborate King's Gate against which the Royal Buttress was subsequently added; the Temple of the Storm God, whose exterior west wall forms the Long Wall of sculpture attributed to Suhis II, that leads directly to the Great Staircase mounting the citadel; and the approach along which the Herald's Wall reliefs face, leading down to the Water Gate and the river.15 Despite the fact that the work was conceived and executed over several reigns, the end product was an imposing inner town, surrounding and leading to the citadel, that was decorated with pious and powerful scenes and inscribed texts appropriate to a prosperous and powerful regional centre.
With this background established, we may now proceed to consideration of the first of three issues to be discussed in the present article: the role of Carchemish as a centre of artistic production and possibly even of craftsmen supplied to other North Syrian states.
Scholars who have studied the art of this period have generally conceded the unity of a "Late Hittite" style pervading the sculptural reliefs of various sites in northern Syria and southeastern Anatolia. One need only refer to the close thematic and stylistic parallels which may be drawn between the orthostats from Carchemish and those from Tell Halaf or Zincirli on the 9th century B.C.16 Nevertheless, despite these relationships, no one would suggest that the reliefs were all done by the same hands or workshops. Proportions of figures, control
by Assurnasirpal II, to 848 B.C., when last mentioned by Shalmaneser III, should not appear on any of the monuments preserved from Carchemish itself (cf. Hawkins, RLA V/6, pp. 443-4). In order to accommodate the Assyrian evidence, the dynasty of kings represented by buildings and inscription: Suhis I, Astuwatamanzas, Suhis II and Katuwas, have all been placed prior to Sangara, beginning in the middle of the 10th century (Hawkins, "Assyrians and Hit-tites," Iraq XXXVI (1974) 70-2). As for Sangara, there are several possibilities: (1) the king did not in fact build or inscribe any buildings, the major construction having been completed prior to his reign and/or the demands of Assyrian tribute having depleted his resources; (2) evidence of his building has not yet been discovered; or (3) he ruled and built under a different name.
13Hawkins, Iraq, XXXVI, p. 73.
14Cf. plan, Care, II, PI. 3.
15Cf. plan, Care, III, PI. 41a. There was evidently also a temple to the goddess Kubaba on the citadel, although few remains were found (cf. Giiterbock, JNES XIII, p. 109).
16 Thematic parallels exist in considerable number: bowman vs. horned animal, chariot riding over fallen enemy, camel rider, animal combats, chimaera and the "Humbaba" motif at Tell Halaf and Carchemish; lion-genius holding reversed animal, winged griffin genius, chimaera, storm god, seated woman in high polos, bowman vs. horned animal, chariot riding over fallen enemy at Zincirli and at Carchemish [cf. Orthmann, USK, Pis. 12a, 9a, 8e, lie, llg and 10a (Halaf), compared with Pis. 33d, 24a, 28c, 33g, 27b, 28a (Care); and Pis. 60a, 59b, 61c, 58d and 59c (Zinc.) with Pis. 33a, 26d, 27b, 23e, 29f, 24a, 33d (Care.)]. The closest comparisons in style and details are to be made in facial physiognomy, dress and ornaments such as crests on horses' heads [ibid., Pis. 8d, 28a, 57c; 11a, 29c, 57d; 1 lb, 24a, 57a].
CARCHEMISH $A KlSAD PURA TTI
181
of composition, use of slabs and variation in details are sufficiently distinctive as to argue rather for a common cultural environment in which they were all separately produced.
The situation is quite different, however, when one compares the early sculptural remains from Til Barsib with those of Carchemish. Til Barsib (modern Tell Ahmar) lies on the east bank of the Euphrates, less than 13 miles downstream and visible from the Carchemish citadel. It, too, commanded an important river crossing; so much so that it became the seat of the Aramaean state of Bit Adini, ultimately taken by Shalmaneser III of Assyria in 856. Two large stelae found at the site are not only pre-Assyrian in date, but pre-Aramaean their inscriptions in hieroglyphic Luwian referring to a Hamiyatas, in one case as ruler, in the other as the predecessor of the current ruler.17 The closeness of the storm-gods depicted on each stele to early works from Carchemish can only be seen from the perspective of very close interaction between the two centres.
The god of Stele A (PI. XLVa, = Hawkins, Tell Ahmar 2) not only wears the same garment and headgear as the Carchemish gods of the Herald's Wall and the Long Wall (cf. PI. XLVIa,6); also, outlines, proportions, smiting gesture, weapons and details of beard and hair-curl are virtually identical, as are the way in which hands - palm, thumb and grasping fingers — are rendered.18 To this group can also be added the slightly later Stele B (PI. XLVb, = Hawkins, Tell Ahmar 1), although it is heavily restored. The guilloche base-line finds its counterpart on the Long Wall, and the bull on which the god stands can be compared directly to the bull-basin of Katuwas from the Temple of the Storm God at Carchemish as well as a relief from the Herald's Wall (cf. PI. XLVIc, PL XLVIIIa),19 where the same proportions, conventions for body-divisions and general form are used. A similar identity may be demonstrated for the fragments of relief showing soldiers holding the decapitated heads of enemies from Til Barsib with the foot soldiers represented on the Long Wall — including the same way of holding the diagonal spear shaft that falls between the soldiers' bodies and the decapitated heads held in outstretched hands.20
These parallels suggest not just a sharing of motifs common to neighbouring states, but an actual sharing of craftsmen. Indeed, Barnett had noted, with respect to Stele B and the reliefs of Katuwas, that the same individual could well have done both.21 The implications of this are far-reaching in terms of the relationship between political power and artistic production. One need not even argue that the finished works were executed at Carchemish and then exported. For, on the basis of the extraordinary open-air site of Yesemek, where sculpture was being carved at the quarry prior to distribution, this could have been the
17 The most current editions of the Til Barsib stelae are to be found in J. D. Hawkins, "The Hieroglyphic Luwian Inscriptions of Syria," to be published in Annates Archeologiques de Syrie [Stele A of Hamiyatas =Hawkins, Tell Ahmar 2] and in id., "The 'Autobiography of Ariyahina's Son': An edition of the hieroglyphic Luwian stelae Tell Ahmar 1 and Aleppo 2," Anat. Stud. 30 (1980) 141-56 [Stele B, naming Hamayatas as a previous ruler, =Hawkins, Tell Ahmarl]. This material is further discussed in the present volume by the same author, to whom I am indebted for permission to make reference to his forthcoming ms. For Stele A, see also, Poetto, "Una revisione dell'iscrizione Luvio-geroglifica di Til-Barsip II," Oriens Ant., 17 (1978), 279-85; for general discussion, Ussishkin, "Was Bit-Adini a Neo-Hittite or Aramaean State?" Oricntalia, NS 40 (1971) 431-7.
18Orthmann, USK, PI. 53c (Til Barsib) with Pis. 23e, 26b (Care).
19 Ibid., PI. 53d (Til Barsib) with Pis. 24a, b. etc. and 25e (Care).
20 Ibid., PI. 54a-c (Til Barsib) with PI. 25a, b, d (Care).
21 Barnett, Care. HI. p. 263; and cf. Gcnge, NSR, p. 57, for the same possibility.
182
ANATOLIAN STUDIES
norm throughout the region.22 Nevertheless, Carchemish is clearly the dominant installation, with a more coherently-conceived aesthetic and a more ambitious building programme — from which must have come the conception and the impetus for the sculpture. It would suggest, therefore, that before the conquest by Shalmaneser III and before the probably quite recent takeover by Aramaeans, the rulers of Til Barsib had turned to the older and more established state of Carchemish for craftsmen.23
The idea of Carchemish as a centre of artistic production and stimulus in the 9th century B.C. suggests itself again in relation to the statue of a king standing upon a double-lion base associated with Palace J at Zincirli.24 In general, garment, belt and sword of the royal figure can be related to the figure of Katuwas that introduces that king's major inscription slab.25 But more important, the head of the Zincirli statue is virtually identical to a head in the round which is all that has been published of a statue of the reign of Katuwas that once stood an one of two double-lion bases from Carchemish.26 I have elsewhere argued for differences in the bases from the two sites being attributed to the fact that the Carchemish bases are earlier than the statues they support.27 The Zincirli statue and base would then be contemporary with the Katuwas statue; the two heads so close that one might even argue for borrowed craftsmen. In any event, unlike the more generalized correspondences in sculpture cited above between reliefs from Carchemish and those of Zincirli, in this case, the work from Carchemish appears to have provided the model for the isolated example from Zincirli an emulation that seems appropriate given the greater size and wealth of Carchemish when compared with that of Sam'al, and the latter state's recurring tendency to align itself as a client of larger states in subsequent periods.28
Thus, both the Til Barsib stelae and the Zincirli statue speak to the artistic role exercised by Carchemish as a cultural centre of the North Syrian and Southeast Anatolian states in the 10th-9th century B.C.   And in addition, in the
22 U. B. Alkim, Yesemek Tasocagi ve Heykel Atelyesinde Yapilan Kazi ve Arastirmalar, Ankara, Türk Tarih Kurumu Basimevi, 1974. With this in mind, one is tempted to encourage archaeological survey along the Euphrates basin for comparable quarries in our region, particularly as Sennacherib recorded exploiting a quarry for stone jars and sculpture at "Kapridargila, on the border of Til Barsib" (Luckenbill, ARAB, II, § 390).
231 would not be at all surprised if the basalt "Kubaba" stele from Birecik, just upstream of Carchemish, also belonged to this group, as once again, the dress, hairstyle, proportions, polos, hand positions, and mirror can be directly compared to female figures from Carchemish, such as the goddess from the Long Wall, and the seated and processing women from the King's Gate (cf. Orthmann, USK, Pis. 5c and Pis. 23b, 29f and g.
24 Ibid., PI. 62c, d and e. The statue base, with a small genius figure in relief kneeling between two lions, is very close in subject matter and conception, though different in style from two such bases found at Carchemish, ibid., Pis. 32d and e (= Care. II, B.26 and III, B.53a; see on this also, M. Mallowan, "Carchemish," Anat. Stud. XXII (1972) 83-4).
25 Ibid., PI. 35g (= Care. II, A. 13).
26 Ibid., Pi. 32a,b. (= Care. Ill, B.54a). See addendum, p. 197.
"Contrary to Orthmann, USK, p. 60; cf. my review of same, JNES, 34 (1975), p. 138. In fact, this would not be the only work stylistically contemporary with the Herald's wall and apparently re-used by Katuwas in the Royal Buttress - cf. Care. Ill, p. 195 and discussion by Güterbock, JNES, XIII, pp. 106-7.
28 E.g., vis-a-vis Assyria in the reigns of Kilamuwa (9th c.) and Panammuwa and Bar-Rakib (8th century) — cf. H. Donner — W. Röllig, Kanaanäische und aramäische Inschriften, Wiesbaden 1968, Nos. 24, 215 and 216-18, resp. On the Kilamuwa text, cf. also, F. M. Fales, "Kilamuwa and the foreign Kings: Propaganda vs. Power, "Welt des Orients X (1979) 6 22. Genge (NSR, Ch. II, pp. 48—50) would see this Zincirli ruler as a crude Aramaic copy of a Carchemish prototype.
CARCHEMISH ŠA KIŠAD PURA TTI
183
historically specific instance one sees an example of a more general situation: -one in which it is to be expected that dominant centres should exert such "influence" upon less powerful, less wealthy, less established places, while ambitious sub-primary places should wish to "emulate" what is being produced in the major centre.29
It is indeed unfortunate in this regard that so few artifactual remains were retrieved as the result of excavations at Carchemish which could shed light on smaller-scale artistic production at the site. Yet, one group of pieces found in the original excavations may perhaps be used as evidence for internal production of portable objects, comparable to the larger, fixed monuments. I refer to fragments of two steatite/chlorite pyxides found in the vicinity of the Water Gate, and a fragmentary pyxis lid reported as a surface find.30
Only a small portion of the bottoms of the two pyxides remain, but there is enough preserved to indicate that both contained scenes of animal and human figures. The first (PI. XLVlIa) shows the feet of a bull, with heavy hooves and a straight tail falling between the hind legs, that has an exact parallel in proportion and rendering on reliefs from the Herald's Wall (PI. XL Villa).31 Behind the bull is a couchant lion with marked shoulder outline, again with the stylization common to reliefs as well.32
J&h^fr W^Hsi., ^^A^t \t
Fig. 1   Roll-out drawing of large steatite pyxis, PI. XLVIIZ) & c (=BM 116122).
On the second pyxis, despite the meagre preservation, we are able to reconstruct most of the encircling scene (cf. PI. XLVIIZ? and c, and drawing, Fig. 1). A well-executed guilloche band forms the bottom edge. Above this, from left to right, we see parts of human feet pointing right, set on a stool or pedestal; the bases of two conical plants or altars, one of which bulges out just as the pyxis breaks off; one small palmette flower; the feet and hem of a long fringed garment belonging to a standing figure facing left (an attendant to the possibly seated figure facing right?); then a second pair of feet facing left and a vertical element that is possibly a plant stem. That whole group seems to form a unit, not unlike scenes on some ivory pyxides of the period.33 It is followed by a pair of striding male legs in a short skirt to the right, before whom is a small leaping lion followed by a larger lion facing left and probably opposed to the man. A small goat, whose legs appear to be dangling in mid-air, is perhaps being carried by another man facing right.   Beyond him is a lion-and-bull combat, in which the bull's chest is
For a discussion of the processes involved in such interaction, see Winter, "Perspective on the 'Local Style' of Hasanlu IVb: A study in receptivity," in Mountains and Lowlands: Essays in the Archaeology of Greater Mesopotamia, eds. L. D. Levine and T. C. Young, Jr., Malibu, 1977, pp. 371-86.
30 Care. II, PI. 28:2, 3 and 4.
31 Care. Ill, B.49b (= Orthmann, USK, PI. 27a).
32Care. I, B. 1 la and 1 la (= Orthmann, LASA', Pis. 26b and a).
33 R. D. BarnettM Catalogue of the Nimrud Ivories, London 1957, S.l, 2, 10, 19, 20, 28, 30, for example.
184
ANATOLIAN STUDIES
already on the ground, one foreleg folded underneath him, the other stretched out in front and partly overlapped by the attacking lion. Finally, another man in a short skirt faces right, and before him are the hind legs of a hoofed animal -probably a goat — and part of another palmette plant.
The care with which the guilloche band is executed, and the overall proportions of the continuous loops, find close parallel in the guilloche forming the lower band of the Long Wall reliefs and inscriptions (cf. PI. XLVIa).34 We cannot see enough of the figure at far left to compare with other seated figures, such as the wife of Suhis on the Long Wall, but she, too, has her feet on a footstool (see Care. Ill, Plate B.40&). The long garment with vertical fold and fringed hem worn by the attendant figures in that grouping are clearly matched in the garments worn by gods who stand on top of a couchant lion in the Great Staircase relief, as well as those worn by the musicians of Katuwas from the King's Gate (PI. XLVIIIc).35 Similarly, the kilt worn by the various figures engaged in animal combats — short fringed skirt with diagonal seam ■- is identical to that of a kneeling hero on a Herald's Wall relief (PI. XLVIIIfr), and appears also worn by the two genii on the double-lion statue bases discussed above.36 The placement of a small rearing lion between the striding man and a larger lion is very reminiscent of the way in which animals are set in between the kilted hero of the Herald's Wall cited just above, and the bull and lion he grasps in hand; while there are also ample parallels for the lion-and-bull combat in which the forequarters of the bull are collapsing.37 A similar stylization of lion's feet and double-outlined shoulder can also be found consistently on the reliefs.38
As for the pyxis lid, only roughly one-quarter of the original has been preserved, but one can see that depicted on the top in flat relief is the paw of a lion extending over the twisted neck of an animal, probably a bull, with double-outlined shoulder (PL XLIXa). The border of the lid is again a band of guilloche pattern, and the outer edge is decorated with a design of rosettes within metope panels. As with the pyxides, the decoration on the lid may be directly related to motifs and elements on the Carchemish reliefs. For example, an almost identical scene is represented on a relief from the Inner Court, contemporary in style with works from the Herald's Wall, where a lion mauls a bull whose head twists back toward the lion, and whose shoulder is outlined by a double line (PI. XLIX6).39 The guilloche has been mentioned above; and the upper rosette border can be compared in both proportion and execution to the similarly-empanelled arrangement of the pattern on the headdress of the goddess Kubaba from the Long Wall procession (PL XLIXc).40
In short, then, these parallels in style and motif between the pyxides and lid on the one hand and the 10th 9th century reliefs from Carchemish on the other, strongly suggest that they all belong to a local tradition of carving which included soft stone for small objects as well as harder stone for architectural reliefs.
Carc. I, PL A.la; II, PL A.12a; III, Pls. A.24a4 and B.37-43, 46.
Orthmann, USK, PI. 29d.
Ibid., Pls. 26a, 32d and e.
For example, ibid., PL 33g.
Ibid., PL 26a.
Carc. III, PL B.57b.
Ibid., PL B.39a.
CARCHEMISH $A KI$AD PURATTI
185
Let me state the assumptions underlying this statement quite explicitly: it is assumed that fixed monuments such as major sculpture and architectural relief are most likely to have been carved locally, by local craftsmen, to exemplify the taste of the local population to which the monuments would be directed, unless a viable alternative explanation can be argued on historical grounds (as for Til Barsib above). It is further assumed — even within a relatively unified cultural region such as North Syria/Southeast Anatolia — that subdivisions based upon local variants of the shared "regional style" will be both apparent and describable, despite the many areas of common content and mode of representation (as was the case with the reliefs of Tell Halaf and Zincirli when compared with those of Carchemish above). Therefore, when portable objects — such as the steatite pyxides and lid — are found at a particular site in association with a series of fixed monuments, can be demonstrated to partake of virtually identical properties when compared to those monuments, and can be equally distinguished from fixed monuments of other sites within the same cultural "region", then I would suggest it is a reasonable conclusion that the portable objects were made within the same locale as the larger works.41
There are two major consequences of these assumptions. First, they allow one to establish (hypothetical) "workshop" traditions when objects are found at the same site. And second, the paradigm may be extended to include objects found at other sites, if supporting evidence can be adduced to attribute them back to a particular locale on the basis of their comparison with fixed monuments.
In the present case, we have attempted to posit the first consequence, given the stylistic and thematic parallels between the steatite objects and the reliefs from Carchemish. But there are more far-reaching implications if one pursues the second consequence as well. For, one can also argue for very close parallels between the Carchemish "group" and some of the ivory pyxides and lids in North Syrian style found at Nimrud and elsewhere.
I will discuss the specific details of these parallels in a forthcoming work on the ivories of North Syrian style.42 Implications of these parallels, however -given the assumptions outlined above - are that a significant group of some of the finest North Syrian pyxides from Nimrud can be argued to belong to a Carchemish orbit (cf. PI. XLIXc?).43
Such an argument would be consistent with the attested wealth and importance of Carchemish in this period, and may be supported further by relevant textual evidence. Lists of tribute received from Carchemish by Assurnasirpal II include finished works of ivory; but even more important, both Assurnasirpal and Shalmaneser III actually received tusks as tribute from Sangara, king of Carchemish.44 The tusks provide us with the evidence that the raw material necessary to stock a local industry was indeed present - presumably supplied either from the Habur basin or from the middle Euphrates, since that is where Assyrian kings record hunting for elephant themselves or receipt of elephants
41 For a special case of the reverse of this situation, see H. J. Kantor, review of Tell Halaf III, in JNES 15 (1957) 171-4, and further comments in I, J. Winter, North Syria in the Early First Millennium B.C., with special reference to Ivory Carving, unpubl. PhD. dissertation, Columbia University, 1973, pp. 350-4.
42North Syria and Ivorv Carving, ms. in progress.
43BameU.CNI, S.50. '
^Grayson, ARI, 2, § 584 (Assurnasirpal IT); L. W. King, The Bronze Reliefs from the Gates of Shalmaneser, King of Assyria, London, 1915, PI. XXXIII (Shalmaneser III).
186
ANATOLIAN STUDIES
as tribute.45 Furthermore, I have argued elsewhere that the North Syrian style as a whole can be reasonably sub-divided on the basis of consistent variations in style and subject matter, such that independent sub-groups can be matched to different local centres of production.46 The present suggestion for Carchemish would therefore fit well into the overall pattern of production already suggested for the region: one in which luxury goods, particularly of ivory, were likely not to have been confined to a single specialized centre, but rather to have been produced, with varying degrees of mastery, in most of the prominent cultural and political centres within the North Syrian sphere.47
Thus, despite the fact that actual finds of ivory from Carchemish were limited and likely to be post-Assyrian-conquest in date,48 it is possible — using the steatite pyxides as stylistic and typological analogues and the reliefs as anchors - to argue for the likelihood at least that Carchemish would have been a centre for the production of high-quality ivory goods during the 9th and possibly 8th centuries B.C., along with high-quality small stone goods. And this fits extremely well, both with textual sources that support the presence of ivory materials and tusks, and with the picture generated in relation to the sculpture of the site: that Carchemish would have been a major cultural as well as political centre.
This reconstruction is further supported by the theoretical constructs of modern Central Place Theory: that, while raw materials are collected from the periphery of central places, luxury goods are most economically manufactured at urban centres where consumers congregate or from which goods will be distributed as merchandise, rather than near the sources of raw material itself.49 One may even press this construct a bit further in the present case. For, if there is a tendency for more luxury goods to be produced at central places of higher order, and if a central place of a higher order is indicated by a greater accumulation of wealth,50 then the wealthier central places would be the most likely to be producing luxury goods. Since Carchemish was clearly one of the two, if not the wealthiest state in North Syria in the 9th century B.C., as will be seen below, then it is certainly theoretically very likely to have also been a significant centre of luxury production, as long as — and we have demonstrated this to be the case — the requisite raw materials were at hand.
We may now proceed to a discussion of the second point to be made in the present study: the significant role of Carchemish as an economic and trading centre in the early first millennium B.C. As noted above, the tribute lists included in inscriptions of Assurnasirpal II and Shalmaneser III can be used to provide evidence for ivory and ivory production located at Carchemish. In addition, a careful reading of these lists of tribute, the collection of which constituted a
4S Grayson, ,4/?/, 2, § 44 (Tiglath Pileser I) and § 681 (Assurnasirpal II).
461. J. Winter, "Carved Ivory Furniture Panels from Nimrud: A coherent sub-group of the North Syrian Style," MMJ 11 (1977) 25-54.
47 It should be noted that similar regional production in wood, bone and ivory carving is still attested in North Syria as late as the early 20th century - cf. Franz Werfel, The Forty Bays of Mum Dagh, Engl, transl. by G. Dunlop, New York, 1934, p. 42.
*Cf. Care. Ill, PI. 71 f, and pp. 167 and 211, as discussed in I. J. Winter, "Phoenician and North Syrian Ivory Carving in Historical Context: Questions of Style and Distribution," Iraq, XXXVIII (1976) 16.
49 Christaller, Central Places, p. 20; Haggett, Loeational Analysis, p. 136.
50Christaller, Central Places, pp. 28 and 29.
CARCHEMISH &4 KI$AD PURATTI
187
major source of Assyrian revenue in the 9th century,51 can provide further information about the particular wealth controlled by the tributary state, and about the types of resources to which it had access.52
For example, it is clear that during the reign of Assurnasirpal, Carchemish and Patina in the 'Amuq provided the most abundant tribute of the North Syrian states.53 In many respects, the two give quite comparable amounts: both provide 20 talents of silver; one talent of gold from Patina is matched by a gold ring, gold bracelet and gold daggers from Carchemish; 100 talents of tin from Patina is balanced against 100 talents of bronze from Carchemish; both give linen garments with multicoloured trim, beds, thrones, couches and dishes (?) of boxwood or inlaid with ivory. However, Carchemish also gives a large number of goods not mentioned for Patina: a quantity of bronze pails, tubs and other vessels; alabaster; the elephant tusks cited above; a chariot of gold; and a gold inlaid couch. In addition, Carchemish provides two and a half times more iron (250 talents as opposed to 100), while Patina gives oxen and sheep that Carchemish does not.
Assuming that these commodities were accurately recorded, the differences noted can only be explained in terms of differing access to resources. The 'Amuq offered an extensive pasture land for flocks that Carchemish did not command. Presumably both states had access to sources of silver, presumably in the Taurus, while the tin given by Patina was probably its most valued resource. The quantity of bronze and iron offered by Carchemish surely suggests unrestricted access up the Euphrates to the copper and iron of Ergani Maden and other rich sites in the eastern Taurus.54 Nor can the greater amount of tribute paid by Carchemish be explained merely in terms of the state having been the more coerced by Assyrian forces. Rather the contrary, as no battles are recorded, and Sangara is never described as "taking fright" before Assurnasirpal in the same way as Lubarna of Patina, while Sangara's gifts of a gold ring and gold bracelet are far more appropriately seen as personal tokens to one perceived as an equal.55 Therefore, I feel we may assume that Carchemish was the wealthier and the more powerful of the two.
In the succeeding reign of Shalmaneser III, we find the two North Syrian states much closer to equal, except that now Patina provides three times more iron than Carchemish.56 It is possible that this may be a consequence of Shal-maneser's campaign of 858 against the North Syrian coalition, in which he not only marched across Kummuh, due north of Carchemish on the Euphrates, on his way to Sam'al, but seems also to have bound Kummuh into a client status with
51 See H. Tadmor, "Assyria and the West: The ninth century and its aftermath," in Unity and Diversity, eds. H. Goedicke and J. J. M. Roberts, Baltimore and London, 1975, p. 37.
52 This was attempted on a gross, regional level by N. B. Jankowska, "Some Problems of the Economy of the Assyrian Empire," in Ancient Mesopotamia, I. Diakonoff, ed., Moscow, 1969, pp. 253-76; R. Maxwell-Hyslop has culled references specifically to iron, in "Assyrian Sources of Iron: A preliminary survey of the historical and geographical evidence," Iraq, XXXVI (1974) 139-54.
SZARI2, § 584.
54 See map in Maxwell-Hyslop, Iraq, XXXVI, PI. XX and p. 148.
55 The best documented parallels for this exchange are those preserved in the Amarna correspondance, cf. C. Zaccagnini, Lo Scambio dei doni nel vicino oriente durante i secoli XV-XIII [Orientis Antiquii Collectio - XI], Rome 1973, esp. Ch. IV: L'articolazione sociale, pp. 183-4, for discussion of gifts to or between "Great Kings," including weapons, jewellery and chariots.
56 Luckenbill, ARAB 1, § 601 (Kurkh Monolith).
188
ANATOLIAN STUDIES
Assyria.57 At any rate, the state took no part in the subsequent battle, and it is likely that the political vicissitudes of Kummuh affected the northerly access of Carchemish to its former metal sources. Nevertheless, Carchemish still appears to have had a significant stockpile of metal goods, as the state does provide 500 (presumably iron) weapons to Shalmaneser, as contrasted with 1000 copper vessels provided by Patina.58
Unfortunately, later kings were not as precise in listing specific commodities in association with specific places;59 but on the basis of the 9th century accounts, it may be suggested that a major source of the wealth of Carchemish was related to its access north and west to the rich metal deposits of Anatolia, and particularly to sources of iron.60
Despite the fact that we do not have an extensive body of texts directly attesting to trade in metals (or any other commodity) at this time, we can only assume that trade was the basis for the acquisition of these resources.61 This may be supported by circumstantial evidence: first, as Carchemish cannot be demonstrated either to have these resources within its own political territory or have acquired them by conquest; and second, on the grounds of the coastal installation at Al Mina, near the mouth of the Orontes, which - whether inhabited predominantly by Greeks, Cypriotes or Syrians - cannot be explained on any other basis than as a funnel for North Syrian goods and resources moving west and Aegean goods and resources moving east in an extended exchange network that had to include the major North Syrian states as suppliers if not active merchants.62
Further evidence can be adduced in support of this reconstruction. The occurrence of Carchemish in an Ebla text of the mid-third millennium in a form indicating the etymology of its name to be kar-kamis, or kämm of K., "Quay of (the god) Kamis," attests to the "trading-station" nature of the site at that time.63 The prominent role of Carchemish in trade and commercial activities is also attested for the second millennium, in texts from Mari, Ras Shamra and
57 Ibid., § 599; and see Hawkins, Iraq, XXXVI, pp. 79-80.
58 The large quantity of copper vessels from Patina, seen in conjunction with Cypriote pottery found in 'Amuq sites just at this time and the corresponding intensity of activity at coastal Al Mina which begins in the third quarter of the 9th century (J. Du Plat-Taylor, "The Cypriote and Syrian Pottery from Al-Mina, Syria," Iraq, XXI (1959) 62-92), suggest that Patina may have recently acquired access to Cypriote copper, which then accounted for a large portion of its wealth. Archaeological evidence thus supports the details given in the Assyrian tribute lists of 9th century kings, and encourages us to put some weight on their veracity.
See, for example, the accounts of Tiglath-pileser III which include only generalized lists of tribute or booty from a group of subservient rulers — ARAB, I, § § 772, 801.
60 In addition to the study of Maxwell-Hyslop on Assyrian sources of Iron, cited above, fn. 79, see also S. Mazzoni, "Gli Stati siro-ittiti e l'"eta'oscura": fattori geo-economici di un sviluppo culturale," Egitto e Vicino Oriente, IV (1981) 311—41, esp. 324-5, in which the wealth of all of the North Syrian states are attributed to their proximity to sources of metal.
61 See now, Hawkins and Morpurgo-Davies, in Festschrift Gunter Neumann, pp. 94-5 and 98—9, where at least we have the Neo-Hittite Luwian words for buying and selling, and transactions of land-sale are attested (in re Care. A.4a), if not "trade".
62See on this, C. Roebuck, Ionian Trade and Colonization, New York, 1959; J. Board-man, The Greeks Overseas (2nd edition), Baltimore, 1980.
63Hawkins, "Karkamis," RLA V, p. 426, citing G. Pettinato, "Carchemis Kar-Kamis," Or. Ant., 15 (1976) 11-15.
CARCHEMISH $A KI$AD PURATTI
189
Bogazkoy,64 and it is hardly likely that such activity would cease without some extensive historical explanation, which cannot be provided before the conquest of Carchemish by Sargon II in 717. In addition, the "Carchemish mina" was not only a known weight standard in the early first millennium; it was actually adopted in Assyria during the 8th century B.C. As early as 1901, Johns had noted that all of Shalmaneser Ill's minas conformed to the heavy standard; the light (Carchemish) mina was first adopted by Tiglath-pileser III65 - precisely at a time when the West was having a very strong impact on Assyria that included the increasing use of Aramaic as a lingua franca and many elements in the visual arts as well.66 The Carchemish mina was subsequently used under Sargon and Sennacherib also. Sargon was the first ruler to apply the phrase "of the king" to the light mina, possibly related to the conquest and incorporation of Carchemish into the empire during his reign. The occasional specification on weights: "mina of the merchant", instead of "mina of Carchemish", further attests to the commercial activities of the state as the basis for the weight standard's adoption in Assyria. And finally, one of the Harper letters, ABL 186, written by an official of the Assyrian king (probably Esarhaddon) at Nineveh, refers to the murder of a "merchant, a native of Carchemish."67 It is not unfounded to suppose, therefore, that at least in this period, Carchemish had generated a class of merchants; and that some of them resided outside of their native city, presumably engaged in their profession.68
Thus, a picture of the productivity and the economic role of Carchemish as a state, prior to Assyrian conquest, begins to emerge. The state was likely to have been engaged in the production of sculpture and of small luxury goods -both artisans and resulting objects likely moving beyond the limits of the city as well. The abundance of metal vessels and weapons given to Assyria as tribute suggests stockpiling, if not production in metal. Large quantities of raw metal given as tribute further suggest access to and exploitation of resources at some remove from Carchemish. Such acquisition could only have been based on trade. The commercial role of Carchemish is further attested by the prominence of the Carchemish weight system, and its ultimate adoption by Assyria, and the existence of individuals known as merchants from the city.
Cf. J. Sasson, "A Sketch of North Syrian Economic Relations in the Middle Bronze Age," JESHO 9 (1966) 161-81; G. Dossin, "Aplahanda, roi de CarkemiS," RA 35 (1938) 115-21; H. Klengel, Geschichte Syriens im 2 Jahrtausend v. u. Z., Teil I - Nordsyrien (D. Akad. der Wiss. zu Berlin, Institut für Orientforschung, Veröffentlichung Nr. 40), Berlin 1965.
65 C. H. W. Johns, Assyrian Deeds and Documents, Cambridge, 1901, pp. 264, 268-70. And cf. CAD 'M', vol. I (1977), manu, p. 220: Al.b, citing ADD 35:3 & 41:2.
66 See on this most recently, P. Garelli, "Importance et role des Arameens dansl'admin-istration de l'empire assyrien," in Mesopotamien unde seine Nachbarn [RAI XXV, Berlin 1978], eds. H.-J. Nissen and J. Renger, Berlin, 1982, pp. 437-8; H. Tadmor, "The Aramaiz-ation of Assyria: Aspects of Western Impact," in ibid., pp. 449-70; and I. Winter, "Art as Evidence for Interaction: Relations between the Assyrian Empire and North Syria," in ibid., pp. 355-82.
67 L. Waterman, Royal Correspondence of the Assyrian Empire, Ann Arbor, 1930—36, no. 186.
68 Another, unfortunately unspecified reference to natives of Carchemish residing in the Assyrian capital is to be found in the "Nimrud Wine-lists," (J. V. Kinnier-Wilson, TheNimrud Wine Lists, London, 1972, p. 91, dated to the reign of Adad-nirari III). Rather than being prisoners of war, as suggested by Kinnier-Wilson (pp. 91, 93), Tadmor has speculated that these may be merchants, along with others of foreign origin included in the same list {Unity and Diversity, p. 42).
190
ANATOLIAN STUDIES
The favourable geographical position of Carchemish — both on the Euphrates and with easy access to land routes — certainly was a significant factor in facilitating this trade. Once again, modern economic theory would emphasize the importance of routes in the formation of a significant "central place," providing both access to resources and also means of subsequent distribution of goods — i.e., the redistribution of the raw materials plus the marketing of added products of local manufacture.69 And in addition, control of routes and particularly river crossings is itself likely to generate wealth, in the form of quay and crossing fees for parties moving across the territory.70
The prominence of Carchemish due to the routes and resources it commanded, and its resultant wealth, must be fully understood in order to assess properly the role played by the state in the history of the early first millennium B.C. Further, I would suggest that it is only through an understanding of these factors that one can properly evaluate Assyrian policy with respect to Carchemish, and indeed with respect to the entire westward expansion of the Assyrian Empire at this time.
We may now come to the third and last point of the present article. For. once one sees Carchemish as a crucial factor in Assyrian foreign policy in the 9th and 8th centuries B.C., an astounding number of facts and events fall into place.
In his initial move westward. Assurnasirpal II marched across Bit Adini, collecting tribute en route; then forded the Euphrates in flood, crossing into the land of Carchemish. Presumably he used the Carchemish crossing itself, as he does not report any other; and once on the west bank, he received the tribute of its ruler, Sangara.71 Although Assurnasirpal subsequently engages in a number of skirmishes in Patina, he does not seem to directly engage Carchemish in any military encounter, and one must wonder if this is not because he considered Carchemish too powerful an adversary to take on in this initial foray.
That his interests were at least in part commercial, and that he was thinking long-term of means for tapping directly the sources of Syrian wealth without being dependant upon continued enforced tribute, has been suggested by Tadmor as the only plausible explanation for the establishment of a "colony" of resident Assyrians at Aribua, a city of Patina, after its capture.72
These three phenomena: the collection of raw materials and goods as tribute, the planting of an Assyrian population well beyond areas of Assyrian political control to tap into these goods and resources, and the avoidance of any direct confrontation with Carchemish, all provide perspective for the actions of Shal-maneser III.
Shalmaneser also avoids a head-on confrontation with Carchemish; in fact, he seems to do his best to avoid Carchemish altogether. In his first campaign against the North Syrian coalition, he crosses the Euphrates to the north, his itinerary leading him through the state of Kummuh into Gurgum and down to
69Christaller, Central Places, pp. 20, 42.
70 It is presumably for this reason that Sargon II grants the city of Assur freedom from quay duties in the so-called "Assur Charter" (H. W. F. Saggs, "Historical Texts and Fragments of Sargon II of Assyria I: The 'Assur Charter'," Iraq, XXXVII (1975) p. 17, 11. 36a, 38a, 36b) as a special privilege, in order to "make (his) dynasty firm".
* Grayson, ARI, 2, § 584; and see fn. 12.
72 In Unity and Diversity, p. 37, re ARI, 2, f 585.
CARCHEMISH $A KI$AD PURA TTI
191
Sam'al.73 The next series of campaigns waged by the Assyrian king in the West were against Bit Adini, culminating in the taking of Til Barsib on the east bank of the Euphrates and the re-naming of it as Kar-Shalmaneser, the king's own port.74
The two sets of campaigns must be seen together to understand what is happening to Carchemish at this time. Again, Carchemish itself is likely to have been too formidable an opponent to engage. Shalmaneser does claim the destruction of several of Sangara's dependent cities;75 but as long as the Assyrian king had control of alternate routes west, he could leave Carchemish alone.
The northern alternative, Kummuh, not only afforded uncontested passage to the armies of Shalmaneser, but in addition, took no part in the battle, providing tribute in 858, 857 and again in 853.76 The neutral, if not "client" status of Kummuh had to have been very important to Assyria. That state's control of strategic river crossings provided the Assyrians with both access to the northwestern routes toward Gurgum and down the "back corridor" into Syria, and also access into the resources of the Taurus — since the same metals of value to Carchemish would have been sought by the Assyrians as well, particularly iron.77 The southern alternative, Bit Adini, was equally essential. Not only did the Til Barsib crossing give the Assyrians a direct route west while by-passing Carchemish, but here, too, Tadmor would impute economic as well as strategic motives, suggesting that the very use of the term karu(m) for the new installation had to convey the traditional connotations of associated trading activity in addition to its literal meaning of "port" or "harbour".78
This combination of Assyrian alliance with or domination over Kummuh and the securing of Til Barsib served as a serious check on the power of Carchemish, boxing that state in between areas of Assyrian control. The consequences of this for Carchemish would have been threefold. First, it was likely to have limited the direct access of Carchemish to the Anatolian metal sources that were a significant basis of its wealth — as partially indicated by the amounts of iron available in the tribute list of Shalmaneser — while at the same time giving Assyria independent access into the Taurus. Second, with an Assyrian installation at Til Barsib, access to the south could have been limited for traffic coming out of Carchemish, while interruption or heavy taxing of river traffic moving up from
See on this the itinerary provided in the Kurkh monolith of Shalmaneser HI (ARAB, I, § 599), and the comments thereon by N. Na'aman, "Two Notes on the Monolith Inscription of Shalmaneser III from Kurkh," Tel Aviv, 3 (1976) 89-106, esp. pp. 92-7. The next major crossing of the Euphrates north of Carchemish is Birecik; however, there is no direct evidence that the Birecik crossing was at that time part of Kummuh, and not at the northern reaches of the territory controlled by Carchemish. With the discovery of a rock relief of Shalmaneser III on the west bank, near Kenk Gorge, north of Birecik, at a place where even today it is relatively easy to cross the river by goat-skin raft, we must consider this crossing as well (cf. 0. A. Tasyiirek, "A Rock Relief of Shalmaneser III on the Euphrates," Iraq, XLI (1979), 47-54), although the relief itself was carved after this campaign, as the inscription refers to Shal-maneser's third campaign and the taking of Til Barsib (ibid., p. 49). Finally, if Hawkins is right in connecting the mound of Samsat with Neo-Babylonian Kimuhi and Assyrian Kummuh ("Kummuh", RLA, VI/5-6), then Shalmaneser could have crossed well to the North. Somehow, for arrival in Gurgum, however, this does not seem the most efficient route, and one would prefer to seek a more direct road via either Birecik or Kenk Gorge.
ARAB I., § 599. 75Ibid., §'s 559, 567,601.
76 Ibid., §'s 599, 601, 610.
77 Cf. Maxwell-Hyslop, Iraq, XXXVI, pp. 148-9.
78 Tadmor, Unity and Diversity, p. 38.
192
ANATOLIAN STUDIES
the south would also have cut into the economic life of the state, as would the provision of an alternative river-crossing that would deprive Carchemish of the duties constituting another of its sources of wealth. A third consequence would have been the isolation of Carchemish from former allies within North Syria. With not only Kummuh and Bit Adini, but also Bit-Agusi and Patina under the suzerainty of Assyria by the end of the reign of Shalmaneser III, Carchemish would have been forced to seek political alliance with lands well beyond its own borders — particularly to the northeast and northwest - in order to maintain its independent position with respect to Assyria.
Evidence that this last was indeed the case may be inferred from subsequent historical records. Carchemish would have had every reason to welcome the realignment of Kummuh with Urartu during the complex and apparently weak period for Assyria that followed the death of Shalmaneser.79 That same period seems to have found not only an Urartean political presence in North Syria generally,80 but perhaps even a direct connection between Urartu and Carchemish. At least, Yariris boasts of his linguistic and scribal abilities, including possibly Urartean;81 and this connection, plus the re-establishment of access into the Taurus it would have afforded, would provide a good context in which to view the renewed prosperity and building activity under Yariris and his ward, Kamanas.
In the same text, Yariris also states that he was known among the Lydians and the Phrygians;82 and it was the apparent political overtures from Carchemish to Phrygia that were taken by Sargon II as a pretext for finally invading the state in 717.83 The reconstruction of an "Urartean connection" and a "Phrygian connection" for Carchemish in the 8th century B.C. thus becomes meaningful, not as isolated or accidental phenomena, but rather as aspects of highly-organized policy.
79Sarduri II (ca. 764-735 B.C.) was able to defeat Malatya, and then impose vassalage upon Kustaspi of Kummuh, after destroying the latter king's royal city of Halpa (modern Halfeti) on the Euphrates (cf. F. W. König, Handbuch der chaldischen Inschriften [AfO Beiheft, 8J Graz, 1955-57, p. 124, Inscr. 103, rev. pt. 9, § iv). See also, Hawkins,irag XXXVI, p. 80; and M. van Loon, "The Euphrates Mentioned by Sarduri II of Urartu," in Anatolian Studies Presented to Hans Gustav Güterbock, ed. K. Bittet, Istanbul 1974, pp. 187-y4.
80 There seem to have been no campaigns south of Kummuh, but this is not to deny the influence of Urartu upon the rest of North Syria during this time. Rather, it took a form other than military intervention — i.e., political alliance. In a fragment of a treaty found at Nineveh that may be part of a treaty between Assur-nirari V of Assyria and Mati'el of Arpad, the state immediately to the south and southeast of Carchemish in the 8th century, it is announced that: "... if the Urartean envoys come, you shall not receive them. .." (A. R. Millard, "Fragments of Historical Texts from Nineveh," Iraq XXXII [1970] 174). Clearly, then, the diplomatic presence of Urartu was a reality that had to be addressed. Similarly, when Tiglath-pileser III began at the beginning of his reign to reclaim the North Syrian territories, he found Urartu allied with most of those states against him (Luckenbill, ARAB I, § 769 and §§797, 813).
81 Cf. Hawkins, Anat. Stud. XXV, p. 150, Appendix 2, re Carchemish A.15b, 4; and Anat. Stud. XXIX, p. 157. Although the passage is badly damaged, it is also possible that Yariris came into direct conflict with Assyria at this time; at least, he refers to the Assyrian king (carrying?) away the Storm-God of Aleppo, at which time he (the Storm-God) retaliated against the land of Assyria (with fire?) — cf. Hawkins, Iraq, XXXVI, p. 72, re Carchemish A.24, and personal communication regarding the final logogram for a word generally meaning "red", hence, fire(?). A second fragment, A.6, refers to one "[X]-atanas, Assyrian king," which may refer to Assur-dan - ibid., p. 73. At any rate, conflict with Assyria at a time when Carchemish was allied with Urartu would not be unexpected.
82 Hawkins, Iraq, XXXVI, p. 68; Anat. Stud. XXV, pp. 150, 152.
83 Luckenbill, ARAB II, § 8.
CARCHEMISH $A KISAD PURATTI
193
as Carchemish attempted to secure its resources and maintain its economic and political life in opposition to Assyrian expansion.
In the period between the regency of Yariris at Carchemish and the conquest by Sargon II, Tiglath-pileser III had systematically re-taken those North Syrian states which had fallen away from the domination of Assyria during the first half of the 8th century, pushing Urartu back into the highlands, and re-organizing the West into a series of provinces and vassal states.84 By the end of the reign of Tiglath-pileser, with the conquest of Unki/Patina and Arpad and the submissive status of Sam'al, Kummuh and Bit Adini, Carchemish was again (and this time, irrevocably) isolated from its former network of allies and resources. Assyrian accounts seem to suggest a canny strategy designed precisely to isolate the state and wear it down. One must then see the overtures of the last king, Pisiris, to Mita of Mushki (Phrygia) as a last desperate attempt to keep connections up on the Anatolian plateau, as well as an attempt to organize common cause with a state not yet subsumed, but equally threatened by, Assyria.
A situation similar to that of Carchemish is recorded in the Cylinder Inscription of Sargon, describing Ambaris of Tabal, who "put his trust in the king of Urartu and the land of Mushki (with their) powerful armies."85 In other words, Carchemish was not the only state among those still holding out from Assyria to have counted on alliances with Urartu and Phrygia.
It is telling that the years following the taking of Carchemish were devoted to campaigns against the more recent allies of the state: against Urartu, which resulted in the well-known destruction of Musasir in 714; and then, in 709, raids against Mita of Mushki, led by the Assyrian governor of Que, which resulted in the first overtures of tribute from Phrygia to Assyria.86 The important letter written by Sargon to his governor in Que in year 12 or 13, further illustrates this phase in political relations — making reference with pleasure to the overtures of Mita, and at the same time commenting upon the interception by the governor of a (presumably subversive) embassy sent by the former ruler of Que to Urartu.87 This reference, plus the one to Ambaris of Tabal, in effect strengthens the view that the earlier activities of Carchemish should be seen as part of a larger picture of concerted political efforts to resist Assyrian domination.
With the capture of Carchemish, the territory was annexed at last to Assyria, an Assyrian governor established over the city, the king and rebels carried off, and Assyrians settled in their place.88 Although bricks found in a Roman hearth on the acropolis, inscribed "Palace of Sargon, king of nations, king of Assyria," and
84 Cf. reference in fn. 80; and discussion in M. Astour, "The Arena of Tiglath-pileser Ill's Campaign against Sarduri II (743 B.C.)," mAYsur, Vol. 2, No. 3 (October, 1979).
ARAB II, § 118; & see, also Display Inscriptions from Khorsabad, ibid., § 55. 86ARAB II, §§ 12-22,42-3,71.
87 Cf. most recent publication by C. N. Postgate, "Assyrian Texts and Fragments," Iraq, XXXIV (1972) 90-8.
88 A brief account of the capture of Carchemish is given in the annals, inscribed on the walls of Sargon's palace at Khorsabad (ARAB II, § 8), and on a prism-fragment found at Nimrud in 1952 (C. J. Gadd, "An Inscribed Prism of Sargon II from Nimrud," Iraq, XVI (1954) 173—201). The longest account was on a basalt slab from Nimrud which stood at the entrance to Room U of the Northwest Palace of Assurnasirpal, restored by Sargon, in which the treasures of Pisiris of Carchemish were said to have been stored (cf. ARAB II, §§ 137-8, and Gadd Iraq, XVI, p. 181). Sargon's rage at the "evil words, lies and vile talk" spread by Pisiris about Assyria which led to the attack, annexation of the territory and establishment of a governor over the people of Carchemish, has been discussed by Tadmor ("The Campaigns of Sargon II of Assyria: a chronological-historical study," JCS 12 (1958) 22 3).
194
ANATOLIAN STUDIES
a single Assyrian-style relief fragment, are the only certain evidence that Sargon ever built at Carchemish,89 the taking of the city and the installation of an Assyrian governor must have afforded enormous satisfaction as well as security to the Assyrian king. By Sargon's death in 705, the North Syrian states and most of the trans-Taurus states as well had been tightly bound into the administrative system of the Assyrian Empire. The important question underlying all these events is then whether there had been reasons other than mere "snowballing expansion"90 dictating Assyrian policy, and how those reasons related specifically to Carchemish.
I have cited Tadmor, and martialled what evidence there is, to argue that Assyria's interest in the West was commercial as well as territorial.91 I have further suggested elsewhere that one explanation for the lengthy focus on North Syria, particularly Carchemish, was competition for the land routes which provided access to metal resources — hence, direct commercial competition.92 That the restriction of commercial activity unfavourable to Assyria was a factor in later treaties is evident in the agreement between Esarhaddon of Assyria and the Phoenician city of Tyre.93 Yet, with the Phoenician cities, until revolt made it absolutely necessary, their viable economic life was not dismantled, presumably because they had access to the sea trade for which Assyria was not equipped on her own. With regard to the North Syrian states, and again, particularly Carchemish, by contrast, the Assyrian administrative machine could easily take over control of the land routes, and therefore had no similar need to maintain their independent economic activity.
From later correspondence, it would appear that the area - despite the former prestige of the Carchemish mina and the state's likely economic role -was essentially turned to agricultural production.94 In any event, Carchemish itself was not entirely destroyed, as witnessed not only by the albeit meagre evidence for Sargonid building on the citadel, but also by the fact that two governors of Carchemish are attested in the Eponym lists for 691 and 649 B.C.95
Care. Ill, pp. 211, 265. A fragmentary cuneiform text that seems to be part of an Assyrian royal inscription and may well belong to Sargon is also published (ibid., pp. 265 and 280, and PI. A.33m). J. D. Hawkins confirms (personal communication) that this is a fragment of an annalistic account, most likely belonging to Sargon if compared to other known inscriptions, such as his Hamath stele. The only distinctive phrase preserved is 1.4: NU-MU-Sk-puratti, lit. "Widow of the Euphrates," which could conceivably refer to Carchemish after its conquest.
W. G. Lambert, "The Reigns of Assurnasirpal II and Shalmaneser III: An interpretation," Iraq, XXXVI (1974) 106.
91 Such a reading for the West can perhaps be supported further by arguments martialled for trade as the primary motivating force behind military and political actions by both Assyria and Urartu in the East as well (cf. L. D. Levine, "East-West Trade in the Late Iron Age: A view from the Zagros." in Le plateau iranien et l'Asie centrale des origines ä la conquete islamique [Colloques internationaux du Centre National de la Recherche Scientifique, No. 567], Paris, 1976,pp. 171-86.
92 Winter, Iraq, XXXVIII (1976) 20.
93 R. Borger, Die Inschriften Asarhaddons, Königs von Assyrien (AfO, Beiheft 9), Graz 1956, pp. 107-9.
*Cf. H. W. F. Saggs, "The Nimrud Letters, 1952 - Part VII," Iraq, XXVII (1965) 27, re ND 2671, regarding the failure of the grain crop in the region of the border between Arpad and Kummuh; also, Waterman, Royal Correspondence, letters 500 and 1082, with reference to grain production in Adini and Hatti; and discussion thereof in Winter, Unpublished PhD. diss., "North Syria in the Early First Millennium B.C.....", pp. 457-8.
95Hawkins, RLA, V/6, "KarkamiS," p. 445, § 16.
CARCHEMISH$A KI$AD PURATTI
195
In fact, Carchemish still survived to be attacked by Nebuchadrezzar, in the power struggles between Egypt and Babylon at the very end of the 7th century B.C.96 That the geographical position of Carchemish remained strategically important is suggested by its selection as the base of Egyptian operations in this stand; and not less by the modern Ottoman railroad bridge that bisects the site to cross the Euphrates today. But with the capture of 717, the state ceased to conduct its own affairs, or to control its own destiny.
In conclusion, I would suggest that of all the states of North Syria and Southeast Anatolia, it is Carchemish that held the position of primary economic importance prior to Assyrian take-over; and I would further suggest that Carchemish played a pivotal role in the determination of Assyrian military policy up to its capture. As had been the case in the second millennium against the Hittites,97 Carchemish was the last major North Syrian state to hold out against an enemy with territorial ambitions. Ultimately, the state had to succumb to the military force of Assyria — it did not have the hinterland to muster large armies. Indeed, the North Syrian states had never been able to organize successfully into large political-cum-military entities, but rather functioned best as independent centres, their wealth derived mainly from position, access and resultant trade, and their power vested rather in the establishment of strongly-bonded networks of interrelations. Military coalitions were attempted in the face of Assyrian pressure; but as individual states were picked off one by one, it was ultimately only a matter of time before Carchemish as well became too vulnerable to withstand absorption.
Thus, we must see the period up to 717 B.C. as one of the gradual wearing down of Carchemish. When we first met the state through Assyrian sources in the early 9th century B.C., it was at the height of its power, its prosperity and status reflected in the tribute provided to Neo-Assyrian kings and in its architectural monuments, which constitute the most complete sequence of any of the North Syrian states. In addition, study of the reliefs of Carchemish in comparison with works from other sites leads us to propose that the artistic workshops of Carchemish either greatly influenced or even provided workmen for monuments from other states.98
On the basis of the foregoing analysis, Carchemish was not only a centre of Neo-Hittite/Luwian culture and cultural production, but also a centre of economic activity — including the manufacture of luxury goods and the acquisition, processing and distribution of major metal resources. As such, until incorporation by Assyria, Carchemish meets all of the requirements for a "central place": a supportive hinterland of agricultural production and subordinate settlements; radial routes of access and control of traffic; access to resources; production of central goods and services, with its corollary, art production and a distinctive aesthetic; influence, wealth, status and power.
96 Ibid., p. 446.
97 See H. Klengel, Geschichte Syriens im 2 Jahrtausend v. u. Z., Teil I - Nordsyrien (D. Akad. der Wiss. zu Berlin, Institut für Orientforschung, Veröffentlichung Nr. 40) Berlin, 1965.
98 In fact, once it was clear that the reliefs of Suhis II and Katuwas were to be dated prior to Assurnasirpal, it became a possibility that Carchemish itself, if not North Syria in general, could have served as a stimulus for the very development of Assyrian architectural relief (cf. Winter, in Mesopotamien und seine Nachbarn, p. 356).
196
ANATOLIAN STUDIES
The art work provides the basis for the primary theoretical conclusion to be drawn from the particular case of Carchemish. For, since Central Place Theory derives from economic geography, emphasis has been on the production of goods rather than art. However, what is "art" here, if not on the one hand luxury goods (i.e., the pyxides and whatever decorated metal vessels and textiles that have not been preserved except in text), and on the other hand political and/or religious goods (i.e. public monuments, the messages of which are carried in visual rather than verbal form). So that, on the basis of the foregoing discussion, one might say that, if our reconstruction of the role of Carchemish with respect to sculpture and luxury production is correct, we may add another dimension to the definition of the primate centre: one in which art is (a) likely to be produced and consumed; (b) likely not to be derivative once the state is fully established; and (c) likely to be exported, in the form of goods, craftsmen or "influence".
Developing centres may well incorporate external elements as they move to forge an appropriate visual vocabulary; but they will ultimately assimilate the stimuli and make of them their own." Once established, mature centres may well continue to import luxury goods, art works, and even craftsmen for their own "display" purposes; and they may periodically absorb external stimuli at significant historical moments as well.100 But they will also be producing and exporting works and craftsmen; and will export their "aesthetic" through these works and craftsmen, to the extent that elements of their style and/or iconography are absorbed elsewhere. The degree of "influence" exerted by such centres will be in direct proportion to the prominence of the centre, the quality of its goods, the extent of its "interaction sphere", and the varying degrees of dependence and receptivity of its clients; while it is essentially the centre's own "aesthetic" that will dominate at home.
As a primary historical conclusion, once we acknowledge the role trade had to have played in the economic life of Carchemish, we are led to see the state as a significant factor in Assyrian policy, and conversely, to see Assyrian policy as a primary factor in the determination of strategies at Carchemish.
At the beginning, assuming our reading of the gifts proferred to Assurnasirpal by Sangara to be correct, the king of Carchemish seemed to be attempting to deal as an equal with another "great king". If subsequently Carchemish seemed to set itself, more than any other North Syrian state, in opposition to Assyria - allying with Urartu to the northeast and with Phrygia to the northwest up to the very last minute of conquest by Sargon II - it seems likely that this was the case precisely because Carchemish, as the most powerful of the North Syrian states, had the most to lose in incorporation by Assyria. And these fears would have been well-grounded. Principles of economic competition demanded that Assyria ultimately reduce the position enjoyed by Carchemish once the state was taken. For, only when its former activities and sources of wealth had been diverted to Assyria could the larger kingdom itself assume the "centre".
Carchemish sa kisad puratti was thus subsumed into the Assyrian Empire: the seat of a provincial governor, whose status ranked well below the governor of neighbouring Til Barsib; the river crossing it once commanded reduced to one
Ibid., pp. 357,364-5.
E.g., Assyria under Tiglath-pileser III and again under Sennacherib, cf. ibid., pp. 366—7.
CARCHEMISH SA KI$AD PURATT1
197
of several; its weight system and trade transferred to Assyria; its luxury production curtailed, as artisans presumably moved to where the power and patronage now lay. And to the prophet Isaiah (Isa. X:9) barely 20 years later, Carchemish was just one more of the Syrian states whose fate was held up to Judah as a warning of what one could expect at the hands of Assyria.
[Addendum. A fragment of the lower part of the Carchemish statue referred to on p. 182 (with footnote 26) — showing fringed garment hem, tassel falling from belt, and feet on low plinth originally socketed to the base — has been mentioned without illustration by H. G. Giiterbock, Guide to the Hittite Museum in the Bedesten at Ankara (Istanbul, 1946), pp. 61—2, which strongly supports the suggestion of direct parallels — if not identity — with the Zincirli figure. I am grateful to Professor Giiterbock for calling this fragment to my attention. ]
PLATE XLIV
(a) View of Carchemish, from East bank of the Euphrates looking northwest.
(Photo courtesy The British Museum)
(b)   View from the citadel of Carchemish looking south, including ruins of the Inner Town, Outer
Town and the Plain of Jerablus.
(Photo courtesy The British Museum)
PLATE XL V
(a) Til Barsib, Stele A of Hamiyatas. (Louvre AO 11501; photo courtesy of M. Pierre Amiet,
Musee du Louvre)
(a) Carchemish, Storm God and Goddess, relief from the Long Wall of Suhis II. (Ankara 104; photo courtesy The British Museum)
(b) Carchemish, Two Gods slaying Lion, relief from the Herald's Wall.
(.Ankara 9666 + BM117909; photo courtesy The British Museum)
(c) Carchemish, Bull-base, Temple of the Storm God of Katuwas.
(Ankara 10103: photo courtesy The British Museum)
PLA TE XL VII
(b—c) Carchemish, large steatite pyxis.
(BM 116122; photo courtesy The British Museum)
PLATE XL VIII
(b) Carchemish, Hero vs. Animals, relief from the Herald's Wall.
(Ankara 9665; photo courtesy The British Museum)
(c) Carchemish, Musicians, relief from King's Gate of Katuwas.
(Ankara 141; photo courtesy The British Museum)
PLATE XLIX
(c)   Carchemish, Head of Goddess (Kubaba?), relief    (d)   Nimrud, long ivory pyxis of North Syrian from the Long Wall of Suhis II. style.
(Ankara 103; photo the author)     (BM 1181 75; photo courtesy The British Museum)