READING HALL

THE DOORS OF WISDOM

THE ASSYRIAN EMPIRE. TABLE OF CONTENTS

CHAPTER VII

HITTITE CIVILIZATION

I.

THE HATTIC ART OF CAPPADOCIA

HITTITE civilization, during an activity which lasted at least a thousand years, was possessed and developed by several societies, differing in chronological periods and geographical location; also, probably, in race. If (as assuredly must be) a continuous and compact bloc of culture, distinguished throughout as Hittite, has to be fitted into the scheme of progressive civilization—a bloc as self-contained as e.g. the Egyptian or the Assyrian—within it was much local and temporal variety. All Hittites, by culture, were not Hatti; all Hittites did not possess the culture at one time or for the same period of time; all did not necessarily derive it from the same source; nor was its development governed by the cultural progress of only one society, or dependent on the existence of a single political domination, such as the transient Hattic empire of the fourteenth and thirteenth centuries BC. Too many Hittite remains have been assumed to be of the period of that empire and the work of Cappadocian Hatti, although the empire in question was at an end more than half a millennium before the civilization. The great majority of the Hittite monuments extant are post-Hattic or neo-Hattic.

The oldest extant remains of Hittite art are probably certain dado-reliefs which decorated the façade of the main gateway of Euyuk Alaja in north-west Cappadocia. Foundations of a gate or other structure, differently orientated, underlie this. Since sculptures of more developed style have been found decorating other parts of the gateway, the frontal slabs in situ may be assumed to be older than its present state, and to have been taken from that underlying building and re-used. As reset, they seem to have formed a double dado, for some slabs of a lower tier were brought to light in 1907, and others subsequently. Thus we have two stages of Hittite art represented, the one by the frontal dados, the other by the inner sculptures, of which the best preserved block shows a lion and ram, in the half-round, and of a fully developed archaic art. Probably the famous pairs of lion-sphinxes, which flank the two portals of the gateway, with the still more famous relief upon one of them, which shows a draped figure standing on a double eagle, belong also to the second stage. Like the eagle, which is derived from the blazon of Lagash, these sphinxes (whatever be the ultimate origin of the sphinx idea) proclaim, by their attitude, style and details of treatment, immediate affinity, not with Egypt, but with Mesopotamia.

The dado-slabs of the façade represent an art which owed something to the Sumerian; but even in this early stage it is sufficiently removed from that parent to suggest that it has been derived through some secondary medium. While this may well have been an early Sumero-Mitannian culture of Ashur, it cannot have been such Assyrian art as that of the Second or New Empire. The other parent must be sought in Inner Asia. However derived, Hattic art quickly achieved independence; for its earliest products, as illustrated at Euyuk, are marked by conspicuous individuality of treatment. Whatever it borrowed it transmuted to express distinctively local ideas. The ritual scenes represented, the details of dress, attributes, cult-furniture and accessories, belong to none but Hittite society. The earliest reliefs are executed in a single plane, which returns to the ground of the stone at an acute angle, hardly softened by any bevelling of edges; that is to say, no attempt is made to suggest anything behind the superficial silhouette. An effect results like that of appliqué work. This feature of execution, which distinguishes early Hittite glyptic intaglio as well as relief-sculpture, did not go wholly out of fashion till far on in neo-Hattic times. The human physiognomy represented is a peculiar one, which is reproduced both in Egyptian portrayals of Hatti encountered by the Pharaohs and also, though less pure, in the most primitive Hittite sculptures that have come to light in north Syria and north Central Mesopotamia. Cappadocian Hatti and their gods are shown with full hairless jowls, with inordinately long noses continuing the line of a rapidly-receding forehead, and with pointed brachycephalic skulls.

Subsequent to the style of these early Euyuk reliefs, with which, perhaps, a somewhat more elaborated dado-block found at Yarreh, in the Sangarius valley, might be grouped (but it is doubtless later), we find examples of two styles showing progressive artistic development. The second style is exemplified by the later Euyuk sculptures, and also by the decoration of the Yasili Kaia shrine which, is to be dated, by comparison with sealings on tablets, to the fourteenth— thirteenth century BC—the period of full Hattic empire. A third and last Cappadocian style has left us a fine gate-sculpture at Boghaz Keui in true and very high relief, representing the Hattie War-god (not an ‘Amazon’, as a Berlin bronze figurine, which shows similarly accentuated breasts, sufficiently proves). It must fall very late in the thirteenth century, and be a monument of the last years of the Hattic empire.

The early Euyuk sculptures can hardly, therefore, be dated after the middle of the fourteenth century; but how much older may they be than that? The artistic advance made by the Yasili Kaia style is not so considerable as necessarily to have required any long lapse of time; and the extant monuments of the earlier style, for their part, do not look like the first efforts of an art. While human figures are naively conceived, animals are represented with a freedom which bespeaks previous practice and some study of living models—such study as superstition often denied to the human figure. Moreover, much expert knowledge and more tradition are evinced in the schematic grouping of the elements of the subjects; one needs to look only at the slabs which show archers hunting a boar and stags to be convinced that discovery will some day reveal Hittite sculptures more primitive. If, then, the earliest Euyuk monuments precede Yasili Kaia by a generation or two, but hardly more, the beginning of the fourteenth century must be their upper limit of date; and another century, or not more than a generation or two, may be allowed for the development of their art out of the first local efforts of sculptors, inspired by some tradition transmitted directly or indirectly from Sumerian work.

If this dating of Euyuk seem altogether too low to those who are inclined, on the evidence of cuneiform archives, to presume a much greater antiquity for the Hattie civilization, they may be invited to consider the evidence for the Cappadocian use of the Hittite pictographic or ‘hieroglyphic’ script. Of this there is no trace on any sculptured monument at Euyuk and none on any stone on that site except one detached block which seems to have borne a few script characters in relief. Since the symbols on the upper part of the Yarreh block also are perished beyond recognition, the most primitive examples of hieroglyphic characters available for study, not only in Cappadocia but anywhere in Asia Minor, accompany the Fraktin rock-relief, some figures or groups at Yasili Kaia, the Kara Bel figures and, perhaps, the ‘Niobe’ near Manisa—the symbols on the last two monuments, as also on Nishan Tash at Boghaz Keui, being too much weathered to be of service. Others of early appearance are to be seen on the Kolitoghlu Yaila (Tyriaeum) block and on certain stones from Emir Ghazi (both places lie outside Cappadocia), but these are not associated with sculpture. Now, on the Fraktin relief, the characters carved have the most primitive forms imaginable; yet that relief is beyond question not earlier than the Euyuk dados, but rather of the Second Style of Cappadocian sculpture, though probably almost as early as the Yarreh block. Thus the available evidence leaves us with the most primitive example of the script —an example than which a more primitive is hardly conceivable —carved on a monument not older in all probability than the fourteenth century BC.

It is, however, well known that long before that century Cappadocia knew and used another system of writing, the cuneiform; and further, that this use was continued and developed by the Hattie monarchy of the fourteenth and thirteenth centuries. The facts, therefore, pose a twofold problem. Why and how did a ‘hieroglyphic’ script come to be introduced by Hatti thus late into a cuneiform-using country? and why and how was it that they used it so little? For it appears on less than a dozen monuments all told that have any good claim to be Hattic, and upon these its use is limited to the shortest of legends. At the same time the possibility that less-enduring vehicles of it have perished, or not yet been discovered, must not be lost sight of, especially since it has been found at Boghaz Keui and Carchemish used for graffiti on the shoulders of earthenware pots. It seems insufficient to conjecture that a cuneiform-writing people, on attaining independence or domination, deliberately invented, under a national impulse, a new script for monumental use alone. It is far more probable that rude warriors, equipped with no better means of written expression than pictographs derived from early Sumerian forms (of these some recently discovered tablets have shown us linear examples), brought them into Cappadocia when they first broke in on its cuneiform-using natives; and that, finding, as empire developed, their pictograph-hieroglyphs, which they continued to keep in monumental use, inadequate for political and domestic needs, the warriors enlisted scribes of the conquered peoples to conduct correspondence and keep records in the older and more convenient script, even as the Ottoman Turks, whom in certain other respects the Hatti often bring to mind, have used Armenian and Greek subjects. Such a practice, which naturally would have limited the use and retarded the development of the Hattie script proper, would account for the very small proportion of known Hittite inscriptions, which belong to the period of Hattic empire or have come to light in the Hattic homeland.

If, then, the Hatti attained comparatively late to those fruits of settled order, a native art and a native script, must we assume a long antecedent history of unsettlement and disorder in their historic home? Probably not; but each man’s answer will be affected by the measure of confidence or distrust that he attaches to the published decipherments and translations of ‘Kanesian’ records from the Boghaz Keui archives. In any case it is to be noted that one leading decipherer of these records finds only eleven kings before Shubbiluliuma’s successor, and, therefore, places Tlabarnash or Labarnash, the supposed founder of the Cappadocian Hattic dynasty, not earlier than the sixteenth century; and that there is no good evidence for Hatti in Cappadocia before that date, though a century or two may be allowed for the process of their settlement. No Semitic document of the region mentions them as a people, nor does the ‘Legend’ of the Cappadocian expedition of Sargon of Agade, of which a fragment was found at Tell el-Amarna.

At present it looks as though, during the third millennium and the first part of the second, the Semitized Cappadocian folk, who read and wrote cuneiform, and maintained intimate relations, commercial and probably also political, with Mesopotamia, had yet to become acquainted with Hatti in their midst. If this were so, most of the earlier history of the Hatti must have been enacted in some other region or regions; and there, rather than in Cappadocia, they must have received their first impulse to culture. But where in the Near East did this happen? Possibly in the Caucasus or Upper Armenia, as some argue from the characteristics of mountaineers that the historic Hatti betray and from their early possession of iron. These appeal for support to an early Caucasian art, derived in some measure from the Sumerian, and capable of having implanted in rude emigrants a fecund germ. But seeing that in much closer proximity to the Sumero-Babylonian area—in north Central Mesopotamia and northern Syria—evidence exists of an independent Hittite culture indirectly inspired by the Sumerian, one is inclined to suggest the south rather than the north for the earlier Hattic homeland; and to connect the retreat of the historic Hittite invasion which, in the eighteenth century BC, ended the First Babylonian Dynasty, with the appearance of Hatti in the north. That early sojourn of Hatti in Babylon may well have encouraged their rudimentary art and letters; and their ejectment have given the first impulse to a part of them to a northward migration which, doubtless with many stages and halts, e.g. in Hani (i.e. Mitanni?), where they deposited the Marduk statue that they had carried off, ultimately landed them in eastern Asia Minor. Such a Hattic Odyssey would have anticipated the course and history of Turk irruptions into the same region in after times. This theory, however, if or when it be proved, would but push the unknown a stage further back, leaving an earlier home of the Hittites and the first nursery of their art still lost in the heart of Asia.

II

HATTIC SOCIETY

The Cappadocian Hatti, at whatever date they first appeared, enter written history as, by at latest the fifteenth century BC, a settled monarchical society, possessed of a civilization far from inconsiderable according to the standard of that age. Their art, from first to last, suggests a stolid warlike folk, largely self­educated and newly promoted to culture (the impression created by all that we know of Hattic history and remains being something like that left by the medieval Scots, rude vigorous fighters who gradually civilized themselves by preying upon elder civilizations). Hattic art is portentously heavy, solemn, and, while markedly independent and inventive, conventional. If it reflects faithfully the society which created it, this must have consisted of very dour uplanders. One can well believe that, when subsequently the Hatti pushed into the trans-Tauric lowlands in virtue of their native vigour and, probably, their use of iron weapons, and came into contact with Egypt and Mesopotamia, they were at some loss how to govern, and were constrained more and more to resort for secretarial and other bureaucratic services to more advanced peoples whom they had subdued, whether in Cappadocia itself or in Syria. Both the Hattushil-Ramses treaty, and the subsequent action of the Hattic king in taking personal conduct of his daughter to Pharaoh, betray, according to immemorial oriental usages, a consciousness of inferiority despite the established success of Hattic arms. One might compare the attitude of Macedonian conquerors towards Athens. If in communications with the Kassite monarchy of Babylon—itself of comparatively recent institution and, to judge by the paucity of its remains, of relatively indifferent culture—the Hattic king expressed himself with more assurance, he clearly used also no little circumspection.

According to the published translations of a Hattic Code of the thirteenth century, some Cappadocian king lifted social laws wholesale from Babylon. So wholesale indeed is this borrowing that it would be rash to assume that the actual level of social organization in Cappadocia was, at that time, at all on a par with that Code’s provisions. What we have is a foreign law, or perhaps only project of law, imposed, or about to be imposed, on a people of lower social development. It offers little warrant for presuming that such advanced ideas as distinction between premeditated and unpremeditated crime, or substitution, wherever possible, of a monetary fine for the death penalty, or inalienable rights of women and slaves, or condemnation of certain sexual acts as being against nature, were realized in the Hattic civilization of the fourteenth century. Nor, again, is there adequate reason for crediting the Cappadocian Hatti with any considerable knowledge of the arts and uses of com­merce. They certainly knew, possessed and could work several metals, notably silver, bronze and iron; but whether mined by themselves or by others we do not know.

So far as the German excavation of Boghaz Keui was directed to other objects than cuneiform tablets, it revealed little sign of any but very simple social apparatus. The enclosed site is spacious as early fortresses go, but in a bleak and ill-watered situation. Its fortifications are massive constructions, imposing enough by their bulk and megalithic methods to be quite worthy of an imperial city; but the remains of religious and palatial buildings, while illustrating considerable skill in masonry, have yielded little evidence of artistic refinement or any sort of luxurious furnishing. They suggest in these respects a life much ruder than was led in contemporary Thebes or Babylon. It must be remembered, however, that tombs, which would have informed us better on domestic matters than can a rocky denuded town site, were not found. The only early Cappadocian graves of which anything is known are some at Kara Euyuk (Kultepe), south of the Halys, the possible site of Kanes, which we have no reason to regard as a seat of the Hatti. The cremation custom illustrated by its burials may have been proper only to older Semitized aborigines.

Religious beliefs and practices of Hattic society in its imperial period are obscure matters. Notwithstanding the plurality of deities, either represented in the shrine at Yasili Kaia, or called to witness Hattic treaties, the cult of the Hatti themselves need not have been polytheist or even dualist; for though a Nature­goddess seems to have had equal honour with a Sun-god, she was the patroness of another city, Arinna, adopted, perhaps, for politic reasons by the Hatti, and she is of a type found almost everywhere in Asia Minor. Doubtless her cult in Cappadocia was of very ancient local origin, and long antedated the Hattic settlement. In the Hattic society of the fourteenth and thirteenth centuries she probably represented an indigenous Cappadocian element. If this were so, then the Hattic deity proper was the Sun-god. He advances at Yasili Kaia behind the goddess of Arinna. A suggestion was made long ago, but is now somewhat discredited, that this whole scene is concerned with the formal union of the deities of a conquering and a conquered people, and that the Yasili Kaia shrine is to be regarded as a chapel of reconciliation. Others hold that it perpetuated, year by year, the sacred marriage of sky and earth which brings the spring. Much depends on the identification of the figures at the heads of the two confronted processions. Those on the right, which stand on wild animals, must be deities; but the leaders on the left, who stand on human supports or mountain peaks, need not be so, though they bring in their train winged and other divine beings. It is, perhaps, more likely that the leader is the Hattic king, attired like his god, and that among his followers are gods of conquered cities. If so, the scene represents a triumph —a returned conqueror presenting himself and his conquests to the deities of his home. As for other divine personages who make up what has been called ‘the Hittite pantheon’, they were almost certainly local deities of other cities and districts federated or subject to the Hatti, and they could have been recognized by the head of the League without prejudice to its own fundamental henotheism.

Elsewhere the Hattic god appears (e.g. at Kara Bel, near Smyrna) as a warrior armed with sword, bow and long spear; or, in the south country (e.g. at Malatia and in many neo-Hattic representations of Syria and Mesopotamia) he stands in the act of striking on a bull. Evidently he came to be assimilated to Hadad of the Semites and Teshub of the Mitanni; we may compare also the Egyptian Sutekh. What his appellation was in the earlier Hattie society of Cappadocia is not known. Possibly it was Tarku, or some similar form: but our evidence points rather to this name having belonged to the local god of other societies than the Hattic. The conservatism of Anatolian cult-usage will go far to justify an assumption that the historical eunuch-priesthood of Cappadocia dates back to Hattic times; but in view of the prevalence of a hairless fleshy physiognomy in representations of Hatti, whether warriors or not, no good argument for hieratic castration can be based on the aspect of their priests.

As for ritual apparatus and cult-attributes, some evidence can be derived from the monuments. These sometimes represent the deity, and perhaps also the deified dead, at a ritual repast, to which the worshipper contributes by libation. A very singular form of table-altar, whose stem suggests the lower half of a figure draped from waist downwards, sometimes appears before the deity. It carries an object atop which the bad preservation of the sculpture seldom allows us to identify with certainty; but sometimes it is, clearly, a bird—an attribute common to many Near Eastern cults. In other representations the table-altar is supported by legs crossed camp-stool wise. Various offerings are borne by those approaching the deity, and music evidently was an adjunct of ceremonies.

An art which showed such manifest initiative and individuality as the Cappadocian Hattic may safely be presumed to have reproduced (so far as it represented scenes of ordinary life) the real aspect of the society of its locality and period, or at any rate, the aspect of the governing class. We may picture its men in pointed or round caps, short bordered tunics falling in a point before and behind, and tiptilted shoes like those of an Albanian hillman. Women and certain priests, possibly eunuchs, wore long robes, and some priestesses, mitres of ‘mural crown’ form (more usually the great Goddess herself wears a hood drawn forward). These features of dress will be found to be maintained, by religious conservatism, in the representations of deities long after the fall of the Hattic empire; but not in those of mortals.

III

SOUTHERN HITTITE ART: ZENJIRLI

It was said, earlier in this chapter, that evidence of an independent but not necessarily contemporaneous Hittite offshoot from the aboriginal cultural stem is offered by southern regions outside Cappadocia. This evidence has been furnished chiefly by excavations conducted at Tell Khalaf, near the source of the Khabur, in mid-north Mesopotamia; and at Zenjirli (Samal), Sakjegeuzi and Jerablus (Carchemish) in the northern part of Syria. On the Mesopotamian site were found rudely sculptured dado-blocks, said by their discoverer to have been reused in a structure of later date, and now, for the most part, owing to a chance of war, in the British Museum. While other blocks of later style show brief cuneiform legends, none bears any Hittite inscription; but the Hittite character of their rude art is none the less obvious. From Sakjegeuzi a series of slabs of much more developed art, representing a hunting scene, was conveyed to Berlin in the last century after long lying partially exposed. At Zenjirli scientific excavation, begun in the nineties and continued in the present century on behalf of the Berlin Oriental Society, has laid bare dado-slabs and other sculptures of Hittite style. The most primitive-seeming of these represent a stage of art distinctly earlier than the Sakjegeuzi slabs; the others, examples of a somewhat more developed work, show significant innovations. The earliest sculptures on all these sites betray the same parents as the Euyuk reliefs, namely Sumerian art, transmitted by some medium, and an art of Inner Asia. But, since in some respects they differ considerably, it will be well to con­sider each site separately.

On the series of slabs from the South Gate of the town at Zenjirli the human male type represented is not unlike the beardless men of Euyuk; but neither is it absolutely the same, nor are the general style, details of human dress, and treatment of animal forms such as could be mistaken for Cappadocian Hattic. On the other hand, by many details (e.g. treatment of lion paws) and by the character of the subjects, these reliefs fall into the Hittite class. On another series of dado-slabs from the Citadel Gate, of later, but not very much later, date, male figures of Hattic profile wear beards; and while the Cappadocian (originally Chalaean?) pigtail, tip-tilted shoe, and conical mitre or skull-cap persist, the short tunic, when worn, is bordered or fringed and does not fall in points before and behind. This second series, however, is too near in style to two statues found on the site (the later is dated to about 800 BC by a Semitic dedication) for its period to be reckoned earlier than theirs by more than about a century; and since the first series cannot reasonably be divided from the second by a much longer interval, we have to consider a date for the most primitive—the most Hittite—of the Zenjirli dado sculptures, which is hardly before 1000 BC—a date, that is to say, divided from that of the Cappadocian First Hattic style by some four hundred years. Yet, while abundance of later sculpture was found at Zenjirli—latest examples, executed after the Assyrian occupation, about 740, becoming so contaminated by North Semitic style and features as to lose much of their distinctively Hittite character—none was found certainly, or even probably, earlier than the South Gate dados. The rudest of the Gate lions need be referred to no older date, nor need the earliest Palace of ‘khilani’ type or the earliest works of fortification. The mass of the smaller objects found by the excavators in the Zenjirli soil has never been published. It is, therefore, not possible to use the evidence of pottery, etc., for dating the beginnings of culture on that site; but certainly it looks as if no social condition that can fairly be called civilized prevailed there before the twelfth century.

This inference is supported by the more meticulously excavated—or more conscientiously published—site at Sakjegeuzi, some twenty miles to the east. Its monumental history seems to have been exactly parallel, period for period, with that of Zenjirli; but before the earliest of its sculptures, pottery gives evidence of long habitation by a society using a culture' derived from the earliest Sumerian, but of much ruder achievement and not, by any distinctive feature, Hittite. Tell Khalaf shows sculptures of ruder style and execution than any at Zenjirli; but their primitive features are the result, without doubt, not of superior antiquity, but of the barbarian ineptitude of remote copyists. They must rank, as regards date, with the reliefs from the Town Gate at Zenjirli.

In view of this dating, one cannot suppose the art of Euyuk and Boghaz Keui to have been the immediate parent or inspirer of Hittite culture on any of those three sites. Had the impulse been given by Cappadocia, during the imperial expansion of the Hatti, we should have found at Zenjirli examples not only of the second Cappadocian style, but also of the Cappadocian script. As it is, both the style and the script in question make complete default. Though the local use of Semitic script shows Hittite tradition by its choice of relieved characters in more than one of the Zenjirli texts, not a scrap of Hittite writing has come to light on any one of the three sites just described—a fact, which, if the scale on which Zenjirli has been excavated be considered, tells strongly against the Hittite script having been used by their societies after they had attained to settled civilization. These societies, to judge by their monuments, spoke and wrote Semitic tongues and were predominantly of Semitic race. The Zenjirli society must have been mainly Aramaean; and it is more than probable that the other two also resulted chiefly from the ultimate settlement of that great Aramaean wave of migration, which, as we know from Assyrian records, was flowing over north Mes­potamia into north Syria in the twelfth and eleventh centuries BC. Historical questions, which these data suggest, had best stand over till the evidence of a more important site—that of Carchemish—has been called to assist. As regards the problem raised by the use of the North Semitic alphabet at Zenjirli, see below. But a few words may be said here on the general character of the Aramaeo-Hittite civilization which prevailed there.

Its princes lived in a circular town whose walls enclosed a small, low hillock standing out of marsh. Their palaces occupied a fortified citadel, irregularly oval. The living chambers, opening off a hypaethral courtyard, were arranged in a self-contained block, laid out on a scheme which Assyrians called ‘a khilani according to the Hatti’. Though this plan has been found at Zenjirli and also at Sakjegeuzi, no such building has been observed in Cappadocia. The scheme implies an oblong stone structure, which was entered in the centre of one of its long sides through a pillared gateway giving access to a narrow ante-room with a small chamber at either end, and beyond this to the main hall, bordered at the back and one side by a range of small rooms, including a bath chamber. It is a simple compact scheme such as might well have originated in the chieftain’s booth of a nomadic tribe. Citadel and town, in both of which domestic structures were built of brick imposed on foundation and ground courses of masonry, were defended by massive stone walls, built dry, and pierced by double gateways planned as in Cappadocia—that is, they were flanked by projecting towers on the inner and outer faces, the salient being much the greatest on the latter face, and between the outer and the inner portals was enclosed an oblong hypaethral court. This arrangement is not found in Assyria. Stone lions kept guard at the portals, and the internal walls might be ornamented with orthostatic reliefs set dado-fashion.

What state and style the prince maintained within his fortress, with what apparatus of life his household was equipped, the publication of Zenjirli, for a reason given already, fails to inform us. Still less do we know how the common folk lived in the lower town. At a late period the citadel wall was lined, round part of its circuit, with a series of small parallel chambers which the excavators call casemates; but we are not told if anything found in them revealed that their use was similar to that of the ring of chambers round the Lower Temple at Boghaz Keui.

IV

CARCHEMISH

In considering the cultural remains of north-western Syria we have thus far neglected the most important, those at Carchemish. The reason is that these remains suggest a history differing in important respects from that of the other sites. The Bronze Age culture of the town and neighbourhood, which is abundantly illustrated by tombs on the Acropolis, cemeteries in the plain, and floor-deposits, shows from first to last none of the features that properly characterize Hittite culture. Whether by its ceramic or its glyptic art, or again by its apparatus of daily life, it recalls entirely the south Mesopotamian cycle. Not long, however, before the close of the second millennium, it was replaced abruptly by a culture not merely Hittite, but, in important respects, Hattic, whose introduction seems to date from a destructive invasion which ruined the pre-existing town, and was followed by rebuilding upon not only the Acropolis hill but a large semicircular area below. The pottery, immediately underlying the new foundations and afterwards characteristic of the restored city and its cemeteries, is wholly different from preceding ceramics. Iron replaces bronze. The method of burial has been changed from inhumation to cremation. Objects very similar to Cypriote products of the Early Iron Age make their appearance. Local glyptic art, previously almost indistinguishable from Sumero-Babylonian, takes on distinctively Cappadocian characteristics; and plastic monuments of comparatively advanced, but impure, Hattic style begin. The society had evidently received some strong alien infusion which stimulated it to cultural advance per saltum

Hattic colour is pronounced from the first. The hieroglyphic relieved script appears at once with the forms of its characters developed and their arrangement and spacing much improved in comparison with prior Anatolian examples. It is used freely for long texts disposed in several registers; and an incised linear reduction of the pictographic characters, evolved through a stage of sunk relief, whether or not after an interval or where originated we have no means yet of determining, makes its appearance and quickly becomes common, obviating any need of cuneiform for cursive writing. Plastic art is represented by monuments never wholly Cappadocian Hattic, but strongly reminiscent, in the conception and execution of figures and groups, of the Yasili Kaia reliefs, divine types being especially so. For example some small cloisonnd images of gold, found in a cremation-grave at Carchemish, might well have been reduced from models of those reliefs. The new burial custom marked a change to what had been a Cappadocian, but not certainly a Hattic, custom. The decoration of the new pottery belongs to an Anatolian and Cypriote dark-on-light family; but this again we do not know to have been proper to the Cappadocian Hatti.

At the same time this new culture at Carchemish, if Hattic, is so with a difference; and what distinguishes, at the outset, its sculpture from the Cappadocian Hattic is not merely evidence of artistic and stylistic development, such as some lapse of time would explain, but features of human physiognomy, hair-fashion, dress and the like. The excavation of the site, being incomplete, has unfortunately not supplied such stratification evidence as would serve to date classes of monuments independently of their style; and it is not till Assyrian influence has become strong that we get a relatively fixed point by comparison of such reliefs as line the Palace staircase with monuments of Ashur-Nasir-Pal and Shalmaneser III, and find ourselves in the ninth century. Thereafter down to the epoch of the destruction of the town, near the close of the seventh century, comparison with monu­ments at Zenjirli avails much. Up to a certain point the cultural history of the two places seems to have run on parallel lines during some centuries; in both culture was Semitized, first by infiltration of Aramaeans, then by Assyrian occupation. But though Assyrian influence appeared earlier at Carchemish (as was natural where Assyrian armies had raided a century before they reached north-west Syria, and the main Mesopotamian trade route to the West passed), neither it nor the Aramaean infiltration, which must also have begun early—since monuments prior to the Assyrian Conquest show bearded figures—succeeded in overwhelming the ‘Hattism’ of Carchemish. This resisted obstinately and largely with success, no doubt because Carchemish, in its capacity of a capital city, received it in greater volume at the first and had it more constantly reinforced. Of the persistence of its Hattism a sufficient proof is its conservation of the Hattie script. As has been said already, its use of this marks an advance both quantitative and qualitative on all previous use. No trace has been found on the site of script-characters so ill-formed and arranged as those on Cappadocian monuments; and by the early ninth century its lapicides had attained to very fine surface detail, arrangement and decorative effect. Subsequent decline is marked by summary treatment and crowded arrangement of characters, through which decorative effect was lost and the minuscule hieroglyphs came to look like worm-casts on the stones. Progressive decay of their artistic quality may be traced, not only at Carchemish, but also on such monuments as the Marash lion and the Ivriz reliefs.

If it is comparatively easy to trace the cultural history of Carchemish after, say, 850 BC, and to arrange its later monuments in a due sequence, we can say little more about other classes of its monuments than that, from their innocence of Late Assyrian influence, they must certainly be earlier than that date. But by how much earlier, and which of them should be before or after the other, can hardly be determined. Such are the reliefs from the ruined Water Gate, probably the oldest; the mythological series of dado-slabs, which lined the inner hypaethral court of the ‘King’s Gate,’ whose scenes are all of Sumero-Babylonian derivation, while their bearded figures wear Hittite shoes and often the Hittite pigtail (this hair-fashion was known earlier to Sumerian art); and the series representing a seated goddess, with long train of priestesses and attendants, in the outer Palace Court. Such, again, is a group of slabs, mostly not in situ, which show curiously slender human figures, with un-Hattic hair-fashion and dress, while the scenes in which they appear recall Cappadocian compositions. Finally, two negative facts are worth attention. First, that no monuments of purely Cappadocian Hattic appearance have yet been found at Carchemish. Second, that there is a like lack of sculptured monuments of very primitive appearance; that is to say, there are no examples of a local infancy of Carchemishian Hittite art. What is perhaps the earliest sculpture found—a lion-supported column base, dug out of a deep trial trench in the oldest part of the site, the Acropolis—shows a fully formed archaic style, affiliated to the south Mesopotamian, but already far removed from its parent.

The whole body of evidence agrees in suggesting that Carchemish and its district experienced, towards the close of the second millennium, invasion by iron-using men who had acquired the full Hattic culture. Probably they came from eastern Asia Minor, and may have been either actual north Cappadocian Hatti forced southwards after the fall of their empire, or more southerly inheritors of the Hattic civilization. On the other side of the Taurus, in the extreme south-east of Cappadocia, monuments found near modern Malatia show that a Hittite culture, slightly contaminated with some southern, perhaps Semitic, infusion, flourished at a time when Hattic artistic traditions were still very strong. This Melitenian art betrays some degradation of the latest Cappadocian Hattic style; and the hieroglyphic script-characters, which are carved on the monuments, mark an advance on the north Cappadocian use. On both accounts the group may reasonably be regarded as post-Hattic, if not long posterior to 1200 BC. On the south side of the Taurus also, at and near Marash, which lies at the mouth of a much-used pass from Malatia as well as of two easy passes from central Cappadocia, occur monuments more purely Hattic in character than any at Carchemish, and much more so than any at Zenjirli or Sakjegeuzi—monuments which may well cover a long duration of time, ranging from, perhaps, the close of Hattic empire to the decay of Hittite culture in Syria. It looks therefore as if the Tauric lands offered to survivors of the imperial Hatti a refuge whence, after some interval, a strong contingent moved on to Carchemish, either before or after that raid into Babylonia which was repulsed by Nebuchadrezzar I. If and when such a neo-Hattic contingent seized Carchemish, it must have established itself as only a dominant minority, not greatly diminishing the subjugated native Syrian folk. For not only do we note, some centuries later, evidence of native pre-Hittite customs having survived and regaining prevalence, but we find plastic work of the new order at Carchemish showing from the first more contamination by Semitic influence than does the Melitenian.

It may be objected here that, since there is no absolute certainty about the higher limit of Hittite culture at Carchemish, it might easily be pushed back a century or so into the period previous to 1200 BC, when the Cappadocian monarchs were holding the place as a fief, and perhaps occupying it. Thus the introduction of Hattic culture would readily be accounted for. But two counter­-objections, taken together, render preferable at present the theory given above. First, that if direct Hattic influence had been exercised in the fourteenth and thirteenth centuries, some monuments of pure Hattic style ought to have occurred at Carchemish; second, that the pre-Assyrian classes of sculptured monuments actually found do not seem adequate to fill anything like all the centuries which would require artistic content, if the original establishment of Hattic empire had been directly responsible for introducing Hittite culture to Syria. Neither at Carchemish, nor anywhere else south of the Taurus (except, conceivably, at Marash), is there evidence of that empire having imposed its own culture during its lifetime. The scattered Hittite monuments of north Central Syria, at e.g. Aintab, Aleppo, Hamah, and Restan, as well as those distributed along the eastern bank of the Euphrates from Birejik to below Tell Ahmar, are, none of them, of earlier appearance than classes which, at Carchemish, are to be ascribed to the tenth or the ninth century. As for the earliest Hittite art of Zenjirli— to return to that site—a comparatively late date is virtually assured for it, and it is not to be affiliated to Cap­padocian Hattic art. Strictly speaking indeed, both there and at Sakjegeuzi we have to do with monuments not of Hittite but of Semitic culture, inspired originally by Hattic tradition. Where and how it came so to be inspired we can only guess. The Aramaean tribes, before their final settlement, probably assimilated no more culture of any sort than Bedouins do; and, if so, they can have possessed none capable of producing works of art till, at earliest, the eleventh century. Probably, too, this stage was first reached by them on the eastern side of the Euphrates. The rude sculptures of Tell Khalaf are witnesses, if not to the beginnings of Aramaean Hittite art, at least to its slow and painful development in a lean land. That it had come into being through contact with surviving Mitannian kindred of the Hatti is the most probable explanation of its genesis; and, if as has been suggested, a certain inscription found to the south of Zenjirli is in a Mitannian language expressed by Semitic characters, this Mitannian origin was not forgotten in a subsequent age. When the Aramaeans passed the Euphrates and settled down in fat territories at Zenjirli and Sakjegeuzi, they received, probably from neighbouring Marash, a fresh Hittite stimulus, and thus, in the earliest artistic fruits of their Syrian settlement, came to outstrip at once the stage in which they had left predecessors at Tell Khalaf. Such, for lack of better, must be our tentative explanation of that Hittite-Semitic blend which constituted the north Syrian civilization of the first half of the last millennium BC.

Post-Hattic culture was developed under kinder skies than the Cappadocian and in closer relation to elder luxurious societies of the south and of the sea. Naturally, therefore, it achieved a social advance. Its remains in Syria convey a hint of amenities, to which Cappadocian society had remained strange. The portentous ugliness of the northern monuments is relieved by more gracious types, human and divine; and scenes of festivity become frequent. Such are the musical concerts, and the children at play on the dado of the ‘King’s Gate’ at Carchemish. Some sweetness and light have been introduced into the dour culture of the earlier Hatti, the introduction being, no doubt, due to the racial mixture. But, however numerous the Aramaeans who filtered into north Syria, its population, particularly in the riverain strip, continued to be called Hattic by the Assyrians, so long as these exercised domination west of the Euphrates.

Beyond a general impression of the culture of Carchemish one may hardly go, while no document of its Hittite period found on the site, except a very few fragments of cuneiform, can be read. We know the place to have been a great focus of trade both before the coming of the neo-Hatti and after its subjection to Assyria. That, on that account, it had been important in even earlier days is suggested by the fact that a Babylonian measure of weight took its name from it. Assyria kept officials there, one of whom, Akhi-Ilai, is named as eponym on a contract of Natanyau (Nethaniah) of Gezer, c. 650 BC.

As to the religious practice of Carchemish, the monuments indicate that the neo-Hatti brought down with them both the Hattic Sun-god and the Cappadocian Goddess; but that the former soon took on the similitude and attributes of the Semitic Sun-god, and, with the ever-increasing Semitization of the place, came to be indistinguishable from Hadad of Zenjirli. By what name he may have been called at Carchemish we do not know; but it is unlikely that he was the Mitannian Teshub, who should belong to an earlier period and a disappearing cult. By the ninth century the religious iconography became conspicuously Semitic.

There have not been found at Carchemish, as at Boghaz Keui, archives of a Foreign Office, or any written documents at all, whose contents throw light on the society of the place. Assyrian records, however, make it clear that here no question of ‘empire’ affected daily life. Carchemish was but one—though probably the richest—of several small Syrian states, which concluded, on occasion, alliances with one another or with states of south-eastern Asia Minor, but maintained no permanent federation. It was doubtless because of a commercial rather than a political connection that cultural relations with Cyprus have left so plain a mark on its products; but it was not the chief buyer and seller in this commerce, but rather a transmitter; for there is good evidence of Cypriote intercourse, late in the second millennium, with a point much farther on into Asia, namely, the city of Ashur. Penetration by Phoenician influence becomes evident in the seventh century; but not that of the Greek communities before the sixth, when Carchemish had fallen to low estate, and any culture that could be regarded as distinctively Hittite had been submerged by the ever-flowing tide of Semitism.

V.

OTHER SITES

About the Hittite culture of other societies, whether south or north of the Taurus, there is little to say, no excavation having been conducted on any site within their borders. What digging has been done on the site of Kadesh (Tell Nebi Mend), at the southern end of the Lake of Homs, is negative in its results, indicating that the place lay beyond the area of Hittite culture. But, as far south as Hamah and Restan, Hittite culture must be presumed to have spread by the ninth, or at latest the eighth, century. This is said not because a king of Hamath, named Yaubidi, is called ‘of the Khatti’ by Sargon—for his is no Hattic name—but because of the occurrence of Hittite inscribed monuments at both places, and a promise of more from the great flat-topped mound of Hamah. But it must be remembered that those two sites lie on a great international route up the Orontes valley, and that a culture, which penetrated thither from the north, is not likely to have spread laterally in steppe and desert areas up to anything like the same parallel. Aleppo on the one hand and Antioch on the other are the probable southward limits of the general diffusion of Hittite culture in Syria. It threw forward no more than outposts down the central road towards Palestine.

North of the Taurus, post-Hattic culture is not better known than in those southern societies; and hardly more can be said of it than has been said already. Besides the state of Milid (Melitene), the only Hittite political organization of the later period in that region, whose remains are such as to excite lively curiosity, is the Mushkian of the Tyanitis. This society evidently maintained cultural relations with north Syria not less close than we know its political relations to have been at one epoch—the last part of the eighth century: and there can be little doubt that thence it derived its post-Hattic art and script. Three of the Tyanean monuments—those of Ivriz and Bor, which, on comparative evidence from Zenjirli, should all be placed in the ninth century—suggest a royalty which maintained pomp and luxury: a fourth, by its proximity to the rich silver-lead deposits of Bulgar Maden, indicates whence some of the wealth of that royalty may have been derived. That the Mushkian state grew into a stable power with considerable capacity of defence may be inferred both from its immunity late in the eighth century, after aiding and abetting Pisiris, king of Carchemish, to rebel against Assyria, and also from the absence of any indication that Assyrian armies were able subsequently to deal effectually with it, though they tried once and again.

Where it excites our curiosity most keenly, however, is on another side, the north. An inscription in Phrygian script, in which the name Mita (Midas) occurs, has been found built into a pier of the ruined aqueduct of Roman Tyana, whose great mound has never been excavated, and may well conceal more such monuments. Far away in the north of Cappadocia, at Euyuk Alaja, have been found other inscriptions in the same script; while, at an equal distance from both of these points, to the west, in the middle basin of the Sangarius, are well-known monuments similarly inscribed. These last are always assumed to mark the home of the Phrygian power, known to Greek tradition. Was, then, all the triangular heart of Asia Minor, which is contained by these three points, under one rule at the period of those several inscriptions—say, late in the eighth century? If so, its rulers, called Mita of the Mushki by Assyrians, must have been identical with the Midas kings of Phrygia known to the Lydians and Greeks. Was the centre of their power at Tyana or in the Sangarius basin? Or, while their winter capital was at the former place, was their summer residence in the wooded well-watered uplands of the latter? And, granted all or any of these possibilities proved (they are in fact far from being so), what about centuries before the eighth? Was the Tyanean Hittite culture with its Hattie script and Aramaeo-Hittite art spread northwards as early as the opening of the first millennium, by which date we know the Mushki to have wielded a formidable raiding power? All that can be urged yet in favour of such an hypothesis are the following considerations: that, as suggested above, Hittite monuments round Mazaca-Caesarea offer evidence of a post-Hattic culture of Tyanean type spread up to the Halys, within moderate distance of Euyuk Alaja; that Hittite monuments, as well as Phrygian, are known in the Phrygian highland, though no search under its soil has yet been made; and that certain Phrygian sculptures, e.g. the two great lion­reliefs of Ayazinn, show south Mesopotamian influence similar to that shown by early post-Hattic monuments of Carchemish.

The occurrence of Phrygian inscriptions in old homes of the Hittite script, if it cannot yet be fully explained, at all events suggests the period at which Hittite culture lapsed in eastern Asia Minor, and another took its place. Presumably about the close of the eighth century an alphabetic system, brought from the north-west, dispossessed the syllabic system, which, in its incised linear form, had been in common local use. For although no examples of the Hittite linear script have been found north of the Halys, they abound in south Cappadocia and the Tyanitis—being, indeed, much more abundant there than examples of the relieved hieroglyphic script of any period. As the linear system accompanies Tyanean sculptures whose late post-Hattic date is beyond question, we may rest assured that it continued in use up to the epoch at which the Phrygian alphabetic script appeared. It has usually been supposed that linear Hittite was peculiar both to the post-Hattic Age and to the northern half of the Hittite area. The second of these suppositions has been demonstrated false by recent exploration at and about Carchemish, where linear Hittite inscriptions have proved plentiful, among them being some of a kind apparently transitional between the relieved and incised scripts, and others which show more elaborate, and probably earlier, forms of the linear characters than any observed north of the Taurus. If the incised system was, indeed, of late origin in the Hittite world, Carchemish rather than Cappadocia should be credited with an invention, which made the Hittite script a convenient medium. But in view of remarkable analogies between elements of this system and pre­cuneiform Sumerian characters, it does not seem safe to presume it so late an invention, unless some day it should be proved that Sumerian pictographic characters were continued in use in Kassite Babylonia.

In fact, a problem of the same sort and difficulty has arisen in regard to the Hittite script as that which arises from the Sumero-Babylonian affinities of Hittite art in Syria—how to bridge the wide chronological interval which divides the latest known Sumerian monuments from the earliest known Hittite. It is a problem not yet seriously attacked by any critic, and not yet soluble for lack of sufficient examples of Babylonian work, at once post-Sumerian and pre-Assyrian, to establish any conclusion, whether positive or negative.

We stand on the threshold only of knowledge about both the origins of Hittite art and its influence upon art contemporary and later. It is clear that it cannot be explained fully by any earlier art whose products are yet familiar to us. Until, therefore, much more has been learned about the art of Mitannian Assyria and its relations to the Sumerian, and also about early art or arts of inner Asia, whose existence and quality may be inferred with certainty from the features of Siberian and South Russian products, from the characteristics of early Persian art, and to some extent from certain Chinese remains, any conclusions about the parentage of the Hattic style cannot be more than provisional or better than nebulous.

CHAPTER VIII

THE KINGDOM OF VAN (URARTU)