Anthropology of a Frontier Zone: Hittite-Kaska Relations in Late Bronze Age North-Central Anatolia Claudia Glatz Roger Matthews Institute of Archaeology Institute of Archaeology University College London University College London 31–34 Gordon Square 31–34 Gordon Square London WC1H 0PY London WC1H 0PY United Kingdom United Kingdom [email protected] [email protected] The northern and northeastern borders of the Hittite Empire of Late Bronze Age Ana- tolia hosted a loosely federated group of peoples known as the Kaska. Hittite texts tell us much about the persistent state of hostility between the Hittites and the Kaska, but there have been few serious attempts to understand the Kaska on their own terms. Here we employ a flexible interpretive framework, rooted in frontier studies, in order to re- view the textual evidence for Hittite-Kaska relations before treating the Kaska as an- thropologically approachable subjects. Issues such as ceramics, diet, and subsistence are explored by means of Late Bronze Age and Early Iron Age evidence from a range of archaeological sites. Finally, survey evidence from the Paphlagonia region is con- sidered in the light of Hittite-Kaska relations, and the importance of natural features as frontiers, especially rivers, is underlined. introduction: pecially in its later centuries. Their significance for looking north from hattusa the Hittites is indicated by the fact that they are men- tioned, as enemies, in every major historical work of the Hittites, as well as in many treaties, religious in- I f we stand today on top of Yerkapı, the great vocations and prayers, oracle queries, letters, and in- rampart that dominates the southern end of Hat- structions to border commanders (von Schuler 1965: tusa, the capital of the Hittites, a sweeping view 10–11). unfolds below as we face north across the city (fig. The episode of imperial formation, consolidation, 1). The entire town is laid out at our feet, the land and collapse that takes the form of the Hittite state is dropping dramatically some 300 m in height over a the dominant political event of Anatolia in the sec- distance of almost 2 km to the northern limit of the ond millennium b.c. From existing texts, principally city and the modern village of Bogazkale. Beyond from Hattusa but also from Mavat Höyük/Tapikka the steep contours of the Hittite city lie banks of roll- (Alp 1991), and from scattered archaeological in- ing hills and broad plains, today a patchwork of fer- formation, we already knew something of the nature tile fields traversed by streams, while beyond them and scope of relations between the Hittites and their more rugged terrain dominates the horizon. From northern neighbors, but no previous fieldwork had this viewpoint, we are facing what for the Hittites been conducted specifically in order to investigate was a combustible and contested frontier zone, for the archaeological evidence for these interactions. not far northward beyond these rolling hills lay the Between 1997 and 2001, five seasons of extensive territory of a group of peoples who caused more and intensive archaeological survey, under the title trouble for the Hittites than any other through their Project Paphlagonia (taking its name from the Ro- entire history—the Kaska. The loosely defined groups man province that covered this region), were con- of peoples called the Kaska feature prominently in ducted in the modern Turkish provinces of Çankırı political developments of the Late Bronze Age, es- and Karabük with the aim of investigating long-term 47 48 GLATZ AND MATTHEWS BASOR 339 Fig. 1. View of Hattusa looking north from Yerkapı. human settlement patterns in this distinctive array of published as a monograph by the British Institute of landscapes in north-central Turkey (Matthews, Pol- Archaeology at Ankara; (2) a study of the historical lard, and Ramage 1998; Matthews 2000) (fig. 2). A geography of north-central Anatolia in the Hittite pe- major contribution of Project Paphlagonia has been riod with reference to possible localization of known to shed new light on Hittite-Kaska relations and to toponyms, currently in preparation by the authors; generate and explore an important case study within and (3) a historical and anthropological approach to a contested imperial frontier zone. Across the entire the Hittite-Kaska frontier zone in the Late Bronze survey region, covering some 8,500 km2, about 30 Age. The present article addresses the third of these sites of Late Bronze Age date were located, identi- areas of research and has the following aims: fied as such by the presence on their surfaces of ce- ramics known from excavated sites in central Turkey • to construct a conceptual framework for ap- to date to that period (definitive dating of these sites proaching north-central Anatolia in the Late within the second millennium is in progress and will Bronze Age, be featured in the forthcoming final publication of • to review the historical evidence for Hittite- the fieldwork). Kaska interactions, Output from these researches, as regards the Late • to employ an anthropological approach to the Bronze Age, will be tripartite: (1) a full presentation Kaska peoples, and of archaeological data and interpretations in the final • to review the nature of Hittite-Kaska relations in report on the field project, currently being compiled the light of archaeological results from Project and completed by a range of contributors, and to be Paphlagonia. 2005 ANTHROPOLOGY OF A FRONTIER ZONE 49 possible to track the intricacies of a complex relation- ship between an imperial power and one of its imme- diate neighbors, the Kaska peoples. How best might we conceptualize and research this relationship? Border and frontier studies have become increas- ingly sophisticated in recent years and now encompass a broad range of approaches (De Atley and Findlow 1984; Kimes, Haselgrove, and Hodder 1982). A recent review article by Lightfoot and Martinez encourages a view of frontiers not so much as clear-cut bound- aries between neighboring communities, but rather as “zones of cross-cutting social networks” (Light- foot and Martinez 1995: 471). Prominent traits of such a framework, as identified by Lightfoot and Martinez, include a merging and blurring of material culture traits at boundary zones, the existence of so- cial and political networks spanning communities on Fig. 2. Map of Anatolia, showing extent of Project Paph- both sides of borders, and the development of seg- lagonia survey region. mental and factional groups within such communi- ties. The aim here is to demonstrate that such a model, characterized by fluidity, overlap, and persistent com- conceptual and physical promise, most accurately encompasses the situation frameworks: frontier studies of Inner Paphlagonia in the Late Bronze Age. In sub- sequent sections of this article, we consider how the and north-central anatolia historical and archaeological evidence sits within Concomitant with a burgeoning interest in empire such a framework. studies, frontiers and borders have received increas- The predominance of a colonialist perspective in ing archaeological and historical attention in recent frontier studies, broadly identified by Lightfoot and years, as researchers have recognized that processes Martinez (1995: 473), has hitherto been unquestioned of cultural definition and hybridization occurring in in the case of relations between the Hittites and their such regions can shed light on broad issues of so- neighbors, the Kaska. One reason for this is that all ciocultural practice and identity. The region of Inner the textual, and almost all the archaeological, evi- Paphlagonia, encompassed by the modern provinces dence originates from the Hittite side of the relation- of Çankırı and Karabük, is an ideal arena for the ex- ship. Not only are there no texts from the Kaska side, ploration and application of approaches to the study but there are almost no excavations that reveal the of frontier issues. At the tectonic level, the region is nature of their settlements, cemeteries, and material traversed by still-active faults, in particular the North culture. It is hardly surprising, then, that terms such Anatolian Fault Zone, which shape the associated as “aggressive,” “wild,” “barbarian,” and, more origi- geology and geomorphology. Geographically, Inner nally, “nemesis from the north” (Gorny 1995: 80) are Paphlagonia spans the transition from the rolling routinely used to describe the Kaska in the light of steppe of the Anatolian Plateau, stretching far to the Hittite history. It is because the Kaska can only be south, to the severe mountain ranges of the Pontic seen through the lens of Hittite history, using Hittite region to the north (fig. 3). Several major rivers, in- primary sources, that such terms seem appropriate. cluding the Kızılırmak (Hittite Marassantiya) and Were it possible to write a Kaska history independent the Devrez Çay (probably Hittite Dahara) (fig. 4), cut of Hittite sources, doubtless the Hittites would seem through the region, further strengthening its capacity to be the aggressors, destroyers, and intruders on the to function as a border zone. Through many periods Kaska stage. Our inability to compose such a history, of its past, there is ample evidence that the region and to compare it with the familiar Hittite version, fulfilled its role as a contested frontier zone. But it should not prevent us from attempting to construct a is especially during the Late Bronze Age, with the more balanced and nuanced view of the intricacies of Hittite Empire at the height of its powers, that it is the Hittite-Kaska relationship. 50 GLATZ AND MATTHEWS BASOR 339 Fig. 3. Paphlagonian landscape. The Hittite historical sources make clear the spe- extent within sometimes brief time-spans. The em- cial nature of their frontier with the Kaska peoples, pire appears to teeter on the brink of collapse at sev- differing in many respects from other Hittite borders eral points in its history before a total, irrevocable with great powers such as Egypt and Mittani. With collapse at the end of the Late Bronze Age, around the Kaska, there was no possibility for the Hittites 1200–1180 b.c. One undoubted factor in this impe- to deal with a single, all-powerful leader for, as we rial fragility was the proximity of the Hittite’s capi- shall see, Kaska society was not structured in that tal, Hattusa, to the vulnerable northern frontier of the way. Furthermore, there is no indication in the Hittite empire where the Kaska had their home, barely three sources that the Kaska ever accepted an imperial way days’ march away (Bittel 1970: 11). of life dominated by Hattusa, unlike regions and Rich textual evidence, principally from Hattusa, polities to the south and east. In territorial terms, the supported by occasional archaeological evidence, Hittite-Kaska frontier was essentially static, a no- provides a detailed picture of the nature and range man’s-land of constant mistrust and mutual misun- of interactions between the Hittites and the Kaska derstanding at the most basic levels. Coping with over a period of several centuries. From the written such a frontier made exiguous demands on the po- sources, whose appreciation is subject to the reser- litical and military structures of the Hittite state. vation that they originate exclusively from one party in a two-party dialectic, it is possible to reconstruct the historical evidence several elements in the structure of this fraught re- for hittite-kaska interactions lationship. First, it is clear that there was consider- able variability in the ways that the Hittites sought to The Hittite state is characterized throughout its deal with the Kaska problem. Hittite approaches to existence by dramatic swings in its fortunes, with the the Kaska veered from attempts at total domination, total territory under its control fluctuating wildly in including military conquest, to efforts to agree and 2005 ANTHROPOLOGY OF A FRONTIER ZONE 51 Fig. 4. View of Devrez Çay-Dahara River. sign treaties establishing mutual rights and obliga- “Hatti appears to be nothing but an Akkadian allo- tions between the two parties. Second, the well- graph of the name Hattusa,” unequivocal evidence attested frequency of Hittite campaigns against the for the supreme role of the capital city as an emblem Kaska throughout the duration of the empire indicates of the Hittite state. This strong sense of geographical an overall failure of any Hittite policy ultimately to attachment lent a singular importance to the core eradicate the Kaska problem. Third, despite that fail- region of the Hittite state as a defining context and ure, the Hittites did largely succeed, through tireless motif for the state. The core area included not only campaigning and other procedures, in establishing a the capital city and its hinterland, but a broad and delicate balance of power in north-central Anatolia ill-defined (for us and perhaps also for the Hittites) that endured for several centuries, with some fluctu- swath of territory in central Anatolia, partly en- ation around a fragile equilibrium. closed by the great Marassantiya River (today the At the base of and permeating this relationship lay Kızılırmak). Conceptually attached to the core re- a Hittite sense of self-identity. (We can say almost gion, and therefore of equal significance, were adja- nothing about Kaska sense of self-identity, due to cent frontier zones and scattered holy cities, some of the Hittite origin of the textual sources.) While argu- which, such as Nerik, lay well beyond the physical ably seeing themselves as strangers in a strange land borders of the core region (Haas 1970; Dinçol and (Van De Mieroop 2000: 138; contra Gurney 1979: Yakar 1974; Houwink ten Cate 1979; Macqueen 153), the Hittites defined their social identity not so 1980). The location of Nerik within territory almost much in terms of shared language, culture, or his- always under Kaska control, along with its im- torical experience, but more by their physical geo- mense cultic importance to the Hittites as home of graphic context, expressed in the term “people of the the Storm-God of Nerik (frequently a divine witness land of Hatti” (Bryce 1998: 19). Gurney (1979: 153) to Hittite treaties; Beckman 1999: 7), is a fundamen- has pointed out that the term “Land of Hatti” may tal structuring principle of the Hittite-Kaska relation- equally be rendered “Land of the city Hattusa” as ship, at least as perceived in royal circles, and one 52 GLATZ AND MATTHEWS BASOR 339 Fig. 5. View of Kızılırmak-Marassantiya River. that can only be appreciated by accepting the signif- priate cultic practices, including animal sacrifice, is icance of cultic practice in Hittite daily life. attested for the entire period. The Hittites’ sense of belonging to a core region These activities can be broadly grouped in three manifested itself in attitudes toward non-Hittite lands. overlapping categories: Such lands could be either friends or enemies: “neu- trality was not an option” (Beckman 1999: 1). The 1. Military/strategic land of Hatti, the Hittite core region, lacks sharply de- • Campaigns led by the king or his proxy from fined physical boundaries. The Marassantiya can be Hattusa into Kaska territory forded without difficulty at almost any point (fig. 5), • Foundation and maintenance of garrison towns, and challenging mountain ranges occur only well and system of routes, outposts, and watchtowers to the north, beyond effective Hittite control during in the frontier zone most periods. The openness of the land of Hatti was • Capture and fortification of Kaska-held towns in particularly significant as regards the location of Hat- the frontier zone tusa, hundreds of kilometers away from the sophisti- • Forced exaction of agricultural and other tribute cated subject territories of the south and the associated from Kaska areas under Hittite control trade routes of north Syria, Upper Mesopotamia, and • Attempts to recapture, rebuild, and reoccupy the the Levant. Cultic attachment to their capital city, holy city of Nerik (located within Kaska territory) home to a thousand gods, and its surrounding sacred • Fortification of Hattusa and other towns in the landscape, appears to have kept the Hittites pinned core region down in a region open to attack from several sides, • Relocation of the capital from Hattusa to a safer and especially from the north, the home of the Kaska. region in the south (Tarhuntassa) The intention here is not to itemize and address every textual attestation of Hittite-Kaska interac- 2. Diplomatic tion (for which see von Schuler 1965; Klengel 1999). • Attempts to agree on political concessions and The aim in this section is to provide an overview of treaties with Kaska leaders, including safe pas- modes of interaction, which can serve as an inter- sage to Nerik for Hittite cultic processions, the pretive device for approaching the archaeological banning of Kaska from entering border towns evidence of Paphlagonia in the Late Bronze Age. such as Tiliura, the granting to the Kaska of spe- Table 1 presents a synthesis of textually documented cific grazing rights within the frontier zone, and Hittite strategies on the Kaska frontier. More or less the occasional acceptance of Kaska settlement constant military campaigning and pursuit of appro- within Hittite territory 2005 ANTHROPOLOGY OF A FRONTIER ZONE 53 Table 1. Chronological Overview of Hittite Attempts to Deal with the Kaska King Strategy/Events Source Old Hittite Kingdom, 1680–1450 b.c. Hantili I (II?) Fortification of Hattusa; “Sammeltafel” (CTH 11), Empire-period copy Building of garrisons; Treaty of Hattusili III and Tiliura (CTH 89) Telipinu Last king to reach towns close to Nerik; Extensive Annals of Mursili II (CTH 61) Middle Hittite Kingdom, 1450–1380 b.c. Tudhaliya I (II) Military campaign; Annals of Tudhaliya (CTH 142) Arnuwanda I Oaths of Kaska leaders; Prayer of Arnuwanda and Asmunikkal (CTH 375) Agreements with Kaska; Oracle (CTH 137) Treaties (CTH 138–40) Instructions for officials and (CTH 257, 260–61) Grenzherrn (border chiefs); Tudhaliya II (III) & Instructions about civil and Mavat texts (Alp 1991) prince Suppiluliuma military measures against Kaska, who steal cattle and wine and threaten Hattusa and Upper Land; Regular military campaigns; Deeds of Suppiluliuma (CTH 40) Repopulation of fortified frontier Deeds of Suppiluliuma (CTH 40, frg 13) towns; Empire period, 1380–1200 b.c. Suppiluliuma I Fortification of Hattusa; Resettlement of border regions; Deeds of Suppiluliuma (CTH 40, frg 28) Mursili II Garrisons; Ten Year Annals of Mursili II (CTH 61) Resettlement of Tiliura; Treaty of Hattusili III and Tiliura (CTH 89, Vs II 3) Muwatalli New capital at Tarhuntassa; Hattusili, petty king of Hakpissa; Hattusili III Fortification and repopulation of Treaty of Hattusili III and Tiliura (CTH 89) frontier; Regulation of interaction with “friendly” Kaska; Treaties regulating Kaska access to Hittite towns; Use of Kaska to usurp Hittite throne Apology of Hattusili III (CTH 81, 11) • Attempts to maintain political and military • Repopulation of abandoned towns in the frontier agreements with potential allies in the north-cen- zone with Hittite subjects or transportees from tral region elsewhere in the empire • Appointment of governors in buffer regions of Hittite tactics in dealing with the Kaska are illu- Pala-Tumanna minated by the archive of 200 texts of early 14th- • Use of “friendly” Kaska to police the frontier century b.c. date from Mavat/Tapikka (Alp 1991), zone and for internal political ends, including re- mainly letters from the Hittite king to officials and cruitment of troops from Kaska regions commanders living at the site. These letters are strik- 3. Demographic ing in demonstrating the involvement of the king in • Slaughter of Kaska population in captured areas the fine detail of guarding and securing a relatively • Shift of Kaska population to other regions small and remote border town against the constant 54 GLATZ AND MATTHEWS BASOR 339 threat of attack from the Kaska. Regular activities Hittites and Kaskans in the region as a whole” (Bryce undertaken by the garrison commander included post- 1986–87: 92), and in treaties there seems to have ing sentries along roads, closing gates at night, main- been provision for some Kaska traders to enter Hit- taining the fortifications, and providing food, water, tite towns to conduct their business. and firewood. The letters hint at the grim realities of Within Hittite society, the overriding significance life in a frontier garrison town: a commander, using of the great king of Hatti is indisputable, “the linch- a hunting term, boasts of “bagging” 16 Kaska pris- pin of the universe, the point at which the sphere of oners (Hoffner 2002: 67). Some captives are de- the gods met that of human beings,” as he has been scribed as being blinded after capture and then set aptly characterized (Beckman 2000: 135). It is this to hard labor in mills, an indication of the often se- linchpin role, incapable of delegation, that underlies vere consequences for the Kaska of capture by the the king’s intimate and exhaustive involvement in Hittites. Others are listed along with the ransom re- every aspect of the attempt to keep the Kaska at bay, quired for their release. Thus, a certain Tamiti of as the Mavat/Tapikka letters show, and in almost Taggasta, “who can see” (i.e., has not been blinded), every military campaign into the harsh mountains to has a ransom price of two boys and one man (Hoffner the north and northeast of Hattusa on an almost an- 2002: 67). The ransoms include animals, usually nual basis. oxen and goats, but interestingly never pigs (see be- It is important to stress the small window of op- low). There can be no doubt that Hittites captured by portunity for military campaigning by the Hittites Kaska had an equally hard time of it, and there is evi- (Houwink ten Cate 1984: 63). The severe weather of dence for temple personnel serving as slaves for the the north restricted campaigns to a few months in the Kaska (von Schuler 1965: 72). late spring and summer, a season when manpower Within military and social contexts, there is con- was also in heavy demand for the annual harvest and siderable evidence for fluidity and mobility between for mud-brick manufacture (with freshly available the Hittite and Kaska sides. An individual called straw and sufficiently mild weather to dry the bricks). Kassu appears to be a Kaska turncoat who rose to a Current efforts at reconstructing a section of the position of authority in the Hittite army (Ünal 1998). city-wall at Hattusa, under the direction of Dr. Jür- Texts from the reign of Hattusili III (reigned 1267– gen Seeher, are shedding new light on the mechan- 1237 b.c.) indicate that Kaska men could serve as ics and exigencies of mud-brick manufacture and wall troops in the Hittite standing army, although there construction. This annual concatenation of demands were restrictions on their movements. In that capac- on human labor—campaigns, harvest, construction— ity they might serve on campaigns or be put to labor may help to explain the considerable evidence for a in military construction projects such as roads or Hittite concern with storage. The storage of large fortification work (Beal 1992: 42–43). Hittite tex- quantities of commodities, such as water and cereal, tual references indicate that the Kaska maintained a as attested by silos and reservoirs at Hattusa and standing force of regular troops, which could be sup- elsewhere (Seeher 1997: 320–3), would provide an plemented by levies when required (Beal 1992: 68– element of flexibility in the distribution of human 69). The Hittites themselves acquired Kaska troops labor across the spectrum of tasks at any time. for the Hittite army as contributions from conquered Once on campaign, the Hittite army would carry provinces; provinces that failed to provide agreed with it bulk supplies of bread, flour, and other com- quotas of men were punished (Beal 1992: 82). Such modities (Beal 1992: 130), which may have been Kaska conscripts could be deployed against other supplemented in some cases by supplies maintained Kaska forces. Other obligations imposed by the Hit- at so-called seal house cities of grain (Beal 1992: tites on subjugated Kaska territories included a re- 131). Such supply centers, administered by an offi- quirement to fight any hostile force marching through cial (AGRIG) accountable directly to the king, ap- their land and to assist the Hittites in repulsing en- pear to have been restricted to the core provinces of emy forces (Beal 1992: 124). Some Hittites fled from the Hittite state, and their military, as opposed to the Hittite state and lived with the Kaska (von Schuler cultic, significance is not at all clear (Singer 1984). 1965: 72). There they must have met with Hittites al- While hesitating to see any of the newly discovered ready held prisoner or there of their own free will. Late Bronze Age sites in Paphlagonia as “seal house Texts make it clear that there were “relatively close cities of grain,” it is worth commenting on the fact political, commercial and social dealings between that all of them are located in close proximity to 2005 ANTHROPOLOGY OF A FRONTIER ZONE 55 natural sources of water, in the form of springs or tite border settlements before farmers and their ani- streams, with ample expanses of arable land adjacent mals were given the all-clear to proceed to their to the site. Each town would have been capable of agricultural holdings for the day’s labor and grazing. producing food and water and perhaps of storing We can assume that lookout posts or towers, proba- these commodities for the benefit of themselves but bly manned day and night (Beal 1992: 270), were also of an army passing through, which might num- situated on high ground with maximum visibility ber several thousand strong (Beal 1992: 283). The over roads and approaches to and from frontier set- Hittite army marched with a considerable baggage tlements. Not surprisingly, lookouts were also em- train, including ox- and horse-drawn carts, donkeys ployed by Kaska forces, as the following lines from carrying fodder, water, supplies, and no doubt tents the Annals of Mursili II make clear: “Because their and basic domestic equipment. It was precisely such [the enemy’s] lookouts were standing [at their posts], baggage that Mursili II left behind at Altanna when and because, if I had tried to surround Pittagga- he made his night march against one element of a talli, Pittaggatalli’s lookouts would have seen me, major Kaska force (von Schuler 1965: 48). An army and he would not have waited for me, but would have on the move, and led by the king, would also be slipped away before me . . .” (Beal 1992: 265). carrying with it all the paraphernalia of the royal Hittite-Kaska interactions in military and diplo- camp, sufficient to maintain the dignity and sacred matic contexts, as textually attested, seem well de- aloofness of the king even while on campaign (Bryce scribed by Lightfoot and Martinez’s phrase (1995: 2002: 15). 471), “zones of cross-cutting social networks.” There The brevity of the campaigning season and the dif- is evidence for mobility between the two sides as ficulty of the terrain meant that objectives in the north well as for one party recruiting factions of the other had to be limited. Most annual campaigns involved side for its own ends. It is highly probable that other attempts solely to recover territory lost in the months aspects of social fluidity between Hittites and Kas- since the previous campaign and to reopen routes of ka, such as intermarriage and peaceful cohabitation, communication and supply. Only Hattusili III and were commonplace and, for that very reason, failed Tudhaliya IV appear to have tried to expand their to find their way into the highly attenuated historical campaigns into major incursions into enemy terri- record. tory, attempts doomed to failure by the inability of the Hittites to maintain adequate supply lines over an anthropology such an inhospitable and extensive landscape (Yakar of the kaska peoples and Dinçol 1974: 98). Two historical fragments appear to refer to watch- towers, and they are frequently mentioned in the In- As mentioned earlier, it is difficult to construct structions for the Commander of the Border-Guards a sense of Kaska self-identity, or even to evaluate (Houwink ten Cate 1984: 65). One text reads, “Let whether such a concept has any meaning, largely due the scouts [occupy] the look[-outs] on the main road. to the Hittite origin of the relevant texts that are the [As they] scanned the forefield down from the town only written source for the Kaska. The potential for —[after they occupy the look-outs] let them ca[re- bias in such a one-sided source base need hardly be fully scan] the forefield [from there likewise]” stressed, but must be kept in mind (von Schuler (Goetze 1960: 69; KUB XIII 1 i 12–14). Another 1965: 1). With due caution, however, it is possible states, “Then let the scouts who hold look-outs re- to glean some hints concerning Kaska society from turn to the town, bolt the gates and exits and let down those same sources. In the Annals of Mursili, for the bars” (Goetze 1960: 70; KUB XIII 1 i 23–28), example, we learn that “Pihhuniya did not rule in the and “Whereas the roads are covered, the scouts will Kaskan manner. But suddenly, where in the Kaskan bring word immediately they see a sign of the enemy” town the rule of a single man was not (customary), (Goetze 1960: 71; KUB XIII 2 i 5–6). These refer- Pihhuniya ruled in the manner of a king” (Bryce ences demonstrate that a systematic and careful orga- 1998: 215). This text suggests that the norm for nization of guards and lookout posts was an integral Kaska rule was through loose consensus rather than feature of the Hittite defensive frontier against Kaska domination by a single leader. In the late 14th-cen- attack and incursion. They also reveal that scouts tury text, the Deeds of Suppiluliuma, Mursili II de- were deployed at dawn in the landscape around Hit- scribes how the Kaska were organized into 12 tribes 56 GLATZ AND MATTHEWS BASOR 339 (Güterbock 1956: 67). Additionally, in the Hittite cation and continuity makes Kaska settlements hard view, the Kaska were grouped into three distinct re- to locate archaeologically, even apart from problems gions (von Schuler 1965: 61–62): western, central, in identifying Kaska material culture, as discussed and eastern (the focus of Project Paphlagonia coin- below. cides with the western Kaska zone). Historical ignorance of Kaska identity is well The Kaska were capable of putting a sizable matched by the state of our archaeological knowl- military force into action: 5,000 to 9,000 warriors, edge of the Kaska. Early explorations by Burney with up to 800 chariots, are typically mentioned in (1956) located possibly relevant sites in the north- Hittite texts. Larger armies were pulled together central region, to some extent expanded on in Sam- from several tribes (von Schuler 1965: 73). The so- sun province and elsewhere by Yakar and Dinçol cial and political structures necessary to bring these (1974); however, work specifically targeted at issues forces together and coordinate their arming, training, of Kaska archaeology has been minimal. Excavations and deployment should not be underestimated. The at Kınık-Kastamonu (Emre and Çınaroglu 1993; tendency toward cohesion among scattered Kaska Gates 1997; Bilgen 1999) have revealed an inten- groups was doubtless stimulated and developed by sive level of metal production, including metalwork- the very presence of Hittite armies on a regular basis. ing kilns, that may tentatively be associated with a According to the Hittite texts, Kaska sites were often Kaska presence, but without publication of associ- protected by fortifications but also took advantage ated pottery it is not possible to date or situate this of the natural defenses afforded by the largely moun- site precisely within a broader cultural milieu. The tainous terrain. famous hoard of silver items from this site (Emre and The social structure of the Kaska shaped the ways Çınaroglu 1993) need not indicate a Hittite presence, in which the Hittites could deal with them, in war as we know from texts that the Kaska frequently and peace. In war, it proved impossible for the Hit- looted such items from Hittite temples and carried tites to engage a significant proportion of the Kaska them off to their mountain strongholds (Goetze in forces in battle and thus to deal a fatal blow to their Pritchard 1969: 399–400). total military capability. A defeat of one Kaska force Apart from the possibility of Kınık-Kastamonu, would be followed by attack from another Kaska no Kaska sites have been excavated, no cemeteries force. It is notable that the only Kaska leader known have been found (Yakar 2000: 300), and there is little to have engaged in open battle with the Hittites, the clue as to what constitutes a Kaska material culture. same Pihhuniya mentioned above, was also the only As Genz (2003: 189) has recently put it, “The prob- Kaska leader to rule “in the manner of a king” (Bryce lem is that we do not know what the Late Bronze 1998: 215). He thus ruled like a Hittite king and Age pottery in the Pontic mountains looked like, fought in battle like a Hittite king. Unsurprisingly, whether it looked like Hittite pottery, or whether it he lost the battle and his kingship at one and the had a tradition of its own.” In fact, the position is same time. But Pihhuniya was the exception, and even bleaker: we do not even know whether the other Kaska rulers kept wise counsel and fought their Kaska were using pottery at all. One suggestion is battles at places and times of their, not the Hittites’, that the Kaska assumed a material culture identity choosing. While not at war, treaties agreed to by the from the Hittites and are therefore archaeologically Hittites with one element of the Kaska tribal confed- indistinguishable from them (Özsait 2003: 203), but eracy need have had little currency with other ele- this argument fails to account for the lack of typical ments. Hattusili III’s attempt to accommodate and Hittite pottery over much of the Pontic region in the exploit some Kaska as “friendly,” using them in his Late Bronze Age, in regions known from texts to usurpation of the throne, was hardly a lasting solu- have been inhabited by Kaska. tion to the problem (von Schuler 1965: 58–59). The lack of excavation of Kaska sites dating to Many Kaska towns are mentioned in Hittite texts the Late Bronze Age necessitates an indirect ap- using the designate URU (town) (von Schuler 1965: proach to the archaeological question of Kaska iden- 75). Most of them are not likely to have been major tity. It has been reasonably argued that the Kaska urban centers, and none of them will have remotely were involved in the final collapse of the Hittite approached Hattusa in scale or complexity. Kaska Empire around 1200–1180 b.c. (Bryce 1998: 379). It towns were frequently destroyed by the Hittites but remains unclear, however, whether or not we can as- appear to have been rapidly resettled or rebuilt else- sociate traces of Early Iron Age occupation among where. This considerable flexibility in settlement lo- the ruins at Hattusa, on Büyükkaya, with an ephem-