1 Introduction

In the most optimistic accounts, Anatolian ‘history’ is made to begin in the Middle Bronze Age (hereafter MBA) and treated as a prologue to Hittite history (e.g., Bryce 2005:21–40), a western offshoot of Mesopotamian history (e.g., Veenhof 2017a), or an independent historical context (e.g., Larsen 2015). This is the period when written sources—the markers of ‘history’ against an illiterate ‘prehistory’—are for the first time produced and circulated among Anatolian urban centers in the frame of the Old Assyrian Kārum period (1950–1750 BCE). The Old Assyrian texts represent an invaluable source of information not only on the activities of Assyrian merchants, their relations with their homeland, and the itineraries they followed but also on the indigenous environment(s) in which they moved and interacted (see also Chapter 4). Nonetheless, the characterization of the MBA as a historical period is mostly a matter of spatiotemporal perspective that ultimately depends on our objectives. It is obvious in an Assyro-centric or Mesopotamo-centric historical framework. It is understandable in any attempt to find the roots of social institutions and ideological constructs that will inform the following centuries of the central Anatolian society that we call Hittite. However, the relevance of the Old Assyrian textual evidence must be briefly contextualized and its limits emphasized before fully considering the regional and interregional interactions involving Anatolia and their possible impact on the local sociolinguistic sphere.

Almost all the textual documentation on the Kārum period derives from the cuneiform tablets discovered at the site of Kültepe, known as Kaneš in Assyrian, and Neša in later Hittite sources (Fig. 4.1).¹ This site includes a ca. 24ha-sized mound, hosting the local royal palace, and a lower settlement, called the kārum (port or trading post), where the Assyrian merchants resided. The private archives of the kārum yielded about 22,400 tablets, and forty additional tablets have been found on the mound. From this large corpus, we understand that the Old Assyrian trade network was an imposing infrastructure, involving a large part of central and eastern Anatolia. At the apex of the Old Assyrian trade, there were, besides Kaneš, about other twenty major stations classified as kārūm, as well as almost as many minor ones, known as wabartum. Some of these stations (e.g., Hattuš, later Hattuša) have been explored archaeologically, yet the total textual evidence they produced, some hundred tablets, is dwarfed by the huge corpus found at Kültepe.

This concentration of most of the textual sources in a single center is coupled with a marked chronological unbalance. The vast majority of the Kültepe texts are from level II (ca. 1950–1836 BCE) of the kārum, while only about four hundred were found in the subsequent level, Ib (ca. 1833–1710? BCE). To the latter we may add most of the tablets found outside Kültepe. The latest kārum level (Ia; ca. 1710?–1685 BCE), which possibly stretched to the very beginning of the Hittite period (Kulakoğlu 2014), did not yield any textual evidence. Besides the stratigraphic distribution, the frequencies of attested dates, provided through the eponym (līmum) system in use in Assyrian chanceries, would indicate that most of the tablets were produced over a mere 40 to 50 years, corresponding to the first half of the 19th century BCE (Barjamovic et al. 2012:53–80, with references to additional literature).

In summary, available textual documentation for MBA Anatolia, despite relating to a vast area of central Anatolia, was mostly produced at a single site, Kaneš, during a very short time span. For linguistic purposes, the documents have further limitations in that they are codified in a language foreign to Anatolia, namely, the Old Assyrian dialect of Akkadian (hereafter: OA), and exclusively concern economic or legal matters. These factors are consequential for the vocabulary employed. Proper names, a few borrowings, and evidence for linguistic interference provide invaluable information on the linguistic substrata with which Assyrian merchants interacted. These data, however, can be hardly projected beyond Kaneš, or the Kızılırmak area at best.

Considering linguistic interactions as part of a broader set of cultural contacts allows us to introduce archaeological evidence into the picture. Our position in this respect, however, is just slightly better in terms of the spatiotemporal coverage of the evidence. Ironically, archaeological information is most abundant in central Anatolia, which is also the area more or less directly documented by the kārum texts. Important ancillary data also derive from Cilicia and the Konya plain, but entire patches of land, especially in the west, have been poorly explored if not ignored altogether relative to MBA frameworks.²

2 The Old Assyrian Merchants and Their Interactions with Anatolians

The Old Assyrian trade network in Anatolia was part of a large commercial enterprise financially supported by central institutions, namely, the palace and temples, as well as influential households residing in the Assyrian capital Aššur. The economic backbone of this organization is well known: Assyrians mainly exported tin and textiles to Anatolia, returning home with large volumes of silver and, to a lesser extent, gold (Fig. 4.1). Aššur functioned as a clearinghouse, insofar as neither of the two exported products was produced in the Assyrian homeland. Textiles were acquired from Babylonian merchants, and tin was likely imported to Assyria from the Iranian plateau, possibly in exchange for (part of) the silver and gold obtained in Anatolia. Therefore, unlike previous systems, such as those used in the Uruk and Ur III, the Old Assyrian trade was not aimed at channeling resources into a core area but was mainly a profit-oriented enterprise in which silver and gold functioned as proto-currencies that were (re)invested in one or the other market. Kings and temple institutions in Aššur could participate as investors in commercial activities and, in any case, extracted revenues from them through taxes imposed on the merchants. But the Assyrian trade can hardly be reduced to a state-sponsored initiative as it involved a variety of investors, including prominent citizens and households who seem to have pursued private interests.

Although the trading season only occupied the warmest part of the year, from late spring to early autumn, Assyrian merchants often resided in Anatolia all year long or even permanently, commissioning others to travel back and forth between Aššur and Kaneš. The Assyrians in Anatolia, therefore, formed a socially coherent community of expats, conscious of its ‘otherness’ and retaining strong bonds with Aššur, always called ‘the City’ (ālum). The main governmental institutions of the City were faithfully reproduced in the kārum. The merchants worshipped Assyrian gods and swore official oaths before the sacred dagger of the Assyrian supreme god, Aššur. Multiple texts attest to the presence in Anatolia of Assyrian cultic personnel assigned to temples dedicated to Aššur, Ištar, and other Assyrian deities (Highcock 2017, with references to additional literature). The use of Assyrian cuneiform as an essential means of communication required the establishment of scribal schools in the kārum.³

Although the merchants kept their ‘Assyrianess’ lively through practices, institutions, and religious beliefs, they were deeply integrated into Anatolian society. Although terms such as ‘colony,’ as a translation for kārum, and ‘Old Assyrian Colony period’ are still widely used, the idea of a top-down imposition of Assyrian sovereignty over indigenous Anatolian communities is now unanimously discredited. The trade organization was based on peer-to-peer negotiations, among different trading stakeholders and between merchants and local rulers, wherein all parties acknowledged each other in search of mutual benefits. Relations were juridically regulated by treaties, whereby Anatolian rulers committed themselves to keeping routes safe in exchange for a share of trade profits that was extracted through taxes and tolls on incoming goods.

It is commonplace to say that without their cuneiform tablets, the Assyrians would be barely recognizable in the material record. The main commodities imported to Anatolia from Assyria, that is tin, and textiles, are no longer visible or preserved. The kārum hosted a substantial number of Anatolians together with Assyrians, yet all of the houses hitherto excavated at Kaneš as well as at other sites where Assyrians traders had a presence conform to Anatolian building traditions irrespective of the ethnolinguistic affiliation of their occupants.⁴ Also, the ceramics and other small finds belong to local types.

Old Assyrian language and script are the clearest evidence of an Assyrian presence at Kaneš. These were chiefly used in communications among merchants. As we shall discuss later, some Anatolians were familiar with Old Assyrian cuneiform, and the number of Anatolians involved in record keeping steadily increased between Kārum II and Ib.⁵ To the later period also belong a number of records written in Old Assyrian by local kings, including the famous letter sent by Anum-hirbe of Ma’ama to Waršama of Kaneš (Balkan 1957) and the recently discovered letter written by Wiušti of Hattuš (KBo 71.81). Despite this, the available evidence suggests that cuneiform remained a foreign feature that was never fully assimilated by the local population throughout the Old Assyrian period.⁶

Apart from the tablets themselves and some seal impressions (see below), very little that can be acknowledged as genuinely ‘Assyrian’ has been identified so far in artifact assemblages. In more general terms, the sphere of interactions between the foreign and indigenous groups has been conceptualized as a ‘middle ground’ on which respective identities were juxtaposed and renegotiated. As Lumsden explained, the Assyro-Anatolian middle ground is “not a process of acculturation, of one culture becoming more like the other, but of a more nuanced form of encounter in the ‘in-between’ space of the middle ground, which results in something completely new.”⁷ A well-known textual example for this kind of cultural negotiation was an appeal made by the merchant’s guild for the release of a colleague (called, as usual, ‘a brother’), who was imprisoned by the royal couple of an unnamed Anatolian polity in relation to an affair of espionage.⁸ The rulers who were addressed were begged to let the poor man plead his innocence ‘before the sword of Aššur,’ according to the Assyrian custom, or perform a river ordeal like a native (ana id lilīk kīma DUMU ālika, literally, ‘to go to the river like one of your citizens [referring to the royal couple]’).⁹

The middle ground finds a precise material correlation in patterns of hybridization that are best exemplified in glyptic traditions. According to a long-held typological classification, MBA seals/impressions present in Anatolia are arranged in four groups or styles: Old Assyrian, Anatolian, Old Babylonian, and Old Syrian. The archives of the Kaneš kārum show a preponderance of the first two styles, which have often been equated with the ethnicity of their owners. The situation, however, is now understood to be more complex. Hybridized middle-ground styles, in which Anatolian motives were assembled in an Assyrian or Mesopotamian matrix or vice versa, have been identified. Like Anatolian-style seals, these were produced by local workshops and, as far as we can judge from the attested owners’ names, they were used by both Assyrians and Anatolians.¹⁰

Middle-ground negotiations involving trade stakeholders and local authorities, as well as the hybridization of glyptic styles, belong to the public domain of commercial interactions. However, the prolonged stays of Assyrian merchants in Anatolia inevitably resulted in a more private form of cross-cultural integration, that is, interethnic marriages. Most commonly, Assyrian merchants contracted second marriages in Anatolia with indigenous women. Over time, as communities became more mixed and Assyrians a permanent presence, marriages between Anatolian men and women of Assyrian descent are also attested.¹¹ The Old Assyrian marriage contract allowed bigamy, provided that the two wives did not enjoy the same status and did not live in the same place. As a rule, the first marriage granted the wife the status of aššātum, ‘main wife,’ while the second wife was called amtum—literally, ‘female slave.’ The distinction seems to have been a matter of temporal order, irrespective of ethnicity. Therefore, during the first generation, we mostly see Anatolian amtum wives as opposed to Assyrian aššatum, that is, the wives that the merchants had left in Aššur when moving to Anatolia. The opposite situation (Anatolian aššatum vs. Assyrian amtum) became more common during later generations.¹²

The seemingly derogatory term amtum derives from the Mesopotamian practice of acquiring female slaves to employ as surrogate mothers in case of anticipated childlessness. However, this is hardly a role that fit the Anatolian amtum, who were not slaves and seem to have enjoyed rights broadly comparable with those of aššātum wives. Heffron (2017) considers the amtum marriage as a form of middle ground negotiation between the trade diaspora and local communities seeking mutual advantages in close cross-cultural ties. The amtum-aššatum semantic shift is thus explained as a case of cross-cultural misunderstanding, whereby classifications compliant with the Assyrian marriage laws were reformulated and reshuffled into a new form acceptable to both parties in the mixed community abroad. This is possible if we observe that Anatolian women may have had a different, less degrading understanding of the slave-wife status than the Assyrians did. The evidence would suggest that marriages between free and unfree in central Anatolian common law entailed conditions not very dissimilar to those expected from fully free marriages. Purely Anatolian couples of the Kārum period practiced a community of property in which wealth was equally split between the partners in case of divorce.¹³ According to the later Hittite Laws (§§ 31 and 33), this rule also applied to cases of divorce between free men (LÚ EL-LUM) and slave women (GÉME).¹⁴

3 The Peoples and Languages of Anatolia during the Old Assyrian Period

From the description proposed in the previous paragraphs, it is evident that the type of society the kārum documents describe is a mixed one, with traces of a diachronic evolution that seems to have modified the relationship between the Assyrians and locals. However, what remains unclear and to some extent debated in trying to describe the cultural, social, and sociolinguistic strata of the population of Anatolia in the Kārum era is the composition of the local populations with which the Assyrians merged. When, as in this case, the material culture does not provide solid hints for a fine-grained differentiation, the procedure most commonly used to attempt such an assessment is an analysis of the linguistic profile of the people mentioned in the texts. Exemplifying this process is Gernot Wilhelm’s influential article ‘Hurrians in Kanesh’ (2008), in which, in the absence of other types of data, Hurrians are recognized based on the fact that they bore Hurrian names. Of course, this approach is not unproblematic. A traditional approach to reconstructing the cultural geography of the early phases of the Anatolian area that took into consideration its linguistic history would be based, at least in part, on speculation about the so-called Indo-European migrations. This, when combined with the archaeological evidence for the prehistorical age, could result in a dangerous attempt to match the main Indo-European and non-Indo-European components with given material cultures.¹⁵

Speculations on the original Indo-European culture, when they are pushed too far, tend to produce poor models, especially when one tries to use them to describe the roots of the situation attested in later historical times. While the common genealogical origin of the Indo-European languages is certain, the earliest phases of the separation of the individual languages from the Stammbaum are lost in the fogs of pre- and protohistory and, by the time the linguistic communities were settled into areas (e.g., Anatolia, Greece, Italy, Iran, and India, to mention some of the earliest ones), the coexistence and cohabitation with local(?) pre-Indo-European communities such as the Hattians in Anatolia, Minoans in the Aegean, Elymians in Sicily, Elamites in the West of Iran, and possibly the Indus Valley civilizations in northwest India) had been established for long enough to make it virtually impossible for us to distinguish between original Indo-European and non-Indo-European cultural components.

Liverani (1988:314) described general similarity in the historical contexts of the introduction of Indo-European linguistic (and demographic?) elements in a way that does not imply long-distance migrations and applies to other apparent cultural turn-overs (think, for instance, of the emergence of the Akkadian and Amorite elements in Early Bronze Age Mesopotamia or the West Semitic elements in the Levant during the Iron Age):

Quel che sembra comunque accettabile, nella ricerca di un collegamento […] è che la crisi delle culture urbanizzate dell’Antico Bronzo, pur avvenuta per motivi di dinamica interna, abbia aperto […] dei ‘vuoti’ demografici e politici, che sono stati occupati da genti contigue che per modo di vita e tipologia economica erano pronte e adatte a subentrare nello sfruttamento di certe risorse e di certe aree.¹⁶

In general, when dealing with the problem of describing the type of multiculturalism and multilingualism of the Anatolian region in the Middle Bronze Age, the most appropriate approach consists in working with the available evidence only, refraining from any attempts at reconstructing former non-documentable processes.

The virtual impossibility of distinguishing Indo-European from previous(?), indigenous(?) elements is also encountered in examining the historical corpora from the later and properly Hittite phase. In particular, Klinger (1996:4) pointed out the enormous importance of the Hattian component in the culture of the Hittite kingdom, which results in mixed and entangled social, cultural, and religious facies. Although, as previously stated, the historical processes that turned Old Assyrian Anatolia into the Old Hittite kingdom are still partly obscure, it is important to observe that the intense degree of multiculturalism that makes Late Bronze Age Hatti look like a melting pot of Luwian, Hattian, Hurrian, and Mesopotamian cultural features seems to be consistent with the similarly mixed situation that emerges from the Old Assyrian documents. In other words, the Hurrian and Luwian cultures and literature during the XV and early XIV century pre-imperial ages did not penetrate a consistent and monolithic ‘Hittite Anatolia’ but rather represented a new wave of multiculturality in an environment that was already the complex result of longue durée interferences.

The main pieces of evidence emerging from the large corpus of cuneiform economic and legal documents found in Kültepe that may help describe the Middle Bronze Age social and sociolinguistic situation are:

1. The existence of the designation nuwā’um, which was employed by the Assyrians to indicate the local populations and may have related to the designation of a specific nisba;

2. Anatolian or otherwise ‘local’ anthroponyms and toponyms, some of them associated with people who owned archives;

3. Lexical items that allow for the identification of cultural and social features, including some that are directly or indirectly related to the local administration and polities;

4. Grammatical mistakes in texts composed by scribes who were not Akkadian speakers (especially in archives owned by or related to Anatolians, only a part of which, unfortunately, has been published).

The first point in the list is only indirectly a matter of linguistic analysis, and should, in our opinion, be approached from several angles. Assyrians referred to locals with the noun nuwā’um. Goedegebuure (2008, based only in part on a hypothesis by Carruba 1992) proposed that the word was an Anatolian loan that referred to the Luwians (luwiya), suggesting Hurrian mediation to account for unexpected phonetic features such as the initial consonant (/n/ instead of /l/, which does not occur word-initially in Hurrian), and providing an insightful, albeit speculative, historical justification for this path of diffusion of the word. Goedegebuure’s proposal may be correct. While the cases of l/n alternation, already listed by Carruba 1992, are not many, it is not true that the phenomenon would be “awkward” (as claimed by Kloekhorst 2019:45), because it would depend on specific features of Hurrian phonotactics. As for the historical settings Goedegebuure (2008) does not claim the word was originally created in Kaneš to refer to local Luwian speakers (pace Kloekhorst 2019:45–46). Nonetheless, the hypothesis is far from being proved. If the word needs to be explained in terms of contact, instead, an alternative explanation may exist that could make the diffusion more linear. As the population of Kaneš was, in all likelihood, mostly Hittite, it is conceivable that the word luwiya- was employed in the area by the Hittites to refer to the west Anatolian ‘strangers’; this could relate to the existence of Western trading routes connected to the kārum system via the common gateway of Purušhanda.¹⁷ The Assyrians may therefore have borrowed the word and extended the use of this designation of foreigners to all local non-Assyrian peoples, including the Hittites themselves. A meaning of ‘stranger’—more generic than ‘Luwian’—for the borrowed word would also better explain why in later phases, when the Anatolian model word was no longer recognized by the Hittites, the Hittite correspondence in the list MSL XII 206, 8f. is in fact dampupi-, which, with Soysal (2006:131), apparently had also the meaning ‘stranger, foreigner’ beside the standard one, which is normally given as ‘unskilled, untrained’. Both this and Goedegebuure’s solution, however, would fail to account for the large diffusion of the word nuwa’um outside of the Assyrian of Anatolia, which makes a local borrowing less likely. Therefore, one should certainly consider also the possibility that the word nuwa’um was merely an Assyrian onomatopoeic (if not etymologically Semitic, cf., e.g., Akkadian nawûm, ‘to become wild’) designation of ‘strangers’, similar, in its origin and function, to the Greek designation for ‘barbars’. All things considered, Kloekhorst (2019:46) is certainly right when he highlights that the word has no specific ethnolinguistic connotation.

Moving to the other points in the list above, Anatolian personal names, lexical loans, and morphological adaptations of non-Akkadian words were identified early in the study of Hittite. In the 1950s, just a few decades after Hittite was deciphered, Bilgiç (1954) examined these features in a long article that was a milestone in the study of the kārum texts. Bilgiç identified over 60 foreign words and 17 uncertain ones in Old Assyrian. In the following decades, these figures changed, and some of Bilgiç’s interpretations can be challenged, especially because of the mixed methodology he employed. He analyzed some words as Indo-European based on true Hittite or Luwian comparanda, but above other interpretations on reconstructa that must be approached with caution. For instance, the connection between the kamsu textile and the Latin camisia (Bilgiç 1954:71, mistakenly: camisa) is hardly credible. In other instances, words listed by Bilgiç appear to be almost certainly non-Assyrian but do not admit an Indo-European or Anatolian interpretation. As was highlighted in studies published in the decades that followed (cf. Dercksen 2007, with references), most non-Assyrian words that can be analyzed as Anatolian and Indo-European seem to be analyzable as Hittite, with only very few cases of Luwian loans. As will be discussed later in this chapter, the situation that emerges for personal names is comparable: there is a fairly large set of non-Assyrian names in the kārum onomasticon but within this set, the subgroup of Hittite names is large, whereas Luwian names are at best a small minority.

Borrowings on the lexical level and mixed onomastics hint at an area of intensive and stable contact but do not prove it. Lexical exchange is far from uncommon even between languages that are in loose or indirect contact with each other, and personal names do not necessarily reflect the synchronic cultural background of an area and demography. Luckily, however, other types of interference exist that are much more telling for the linguistic profile of the local population of Old Assyrian Anatolia (or, at least, for the linguistic profile of its scribes). Grammatical mistakes that may depend on interference occur in some documents, especially in those from archives that belonged to individuals who bore Anatolian names. Some of these errors, as will be discussed later, can be regarded as depending on the native language of the scribes, who must have been non-Assyrian and, in all likelihood, were Anatolians in an Indo-European sense.

4 The Geography and Scope of Old Assyrian Trade

The kārum and wabartum making up the Old Assyrian network were all associated with indigenous urban centers whose names, in most cases, broadly correspond to toponyms attested in later Hittite texts. Despite ongoing debates on individual localizations, the comparison between the two geographic corpora permits a fair reconstruction of the scope of the Old Assyrian trade (Fig. 4.1). The caravans started from Aššur and made a long journey to reach the Euphrates at Hahhu. Thence they proceeded to Kaneš across the Antitaurus in the Maraş-Elbistan area. Kaneš served as the hub for caravans departing on several routes that led across Anatolia in an area extending from the Taurus to the Pontus and the Black Sea, and from the Phrygian highlands to the northern Konya plain.⁴¹

As mentioned above, investments in tin and textiles imported from Mesopotamia to Kaneš produced a return in silver and gold that was eventually shipped to Aššur. But that constituted only the main trunk of a complex endeavor that involved intermediate local circuits focusing on other commodities and models of exchange (Fig. 4.1). A key node in the Anatolian network north and west of Kaneš was the kārum of Durhumit (the Hittite Durmitta), a regional market where specialized Assyrian firms traded large quantities of tin, wool, and textiles in exchange for copper. It has been argued that Durhumit was a major hub for the collection of copper shipments from a network that controlled the copper-rich ores of the Pontic area.

View Full Size

Figure 4.1

Map of MBA Anatolia, with the main trajectories of Old Assyrian trade (dashed line) and key places mentioned in the text

• Download Figure
• Download figure as PowerPoint slide

Suggested localizations for Durhumit/Durmitta range across the whole Kızılırmak bend, from the eastern sector (between Sivas and the upper Delice basin), to the north (between Zile and Merzifon), and from the southwest (at or around Büklükale in Kırıkkale province) to the northwest (the mid-lower Kızılırmak around Çankırı).⁴² The last proposal seems the most reasonable on account of the arguments recently advanced by Kryszeń (2016:343–387). In this geographic collocation (Fig. 4.1), the role of Durhumit/Durmitta as an important gateway for the Old Assyrian copper trade, and later for Hittite engagements with the Kaška and the northwestern peripheries, makes much more sense. The Çankırı area, in fact, is connected to the east and southeast through the Çorum and Delice Çay basins and the north, west, and south through the Çankırı and Kızılırmak valleys.

The copper trade fuelled by the kārum of Durhumit did not have an endpoint in the Assyrian market itself as the copper purchased was immediately reinvested in other Anatolian circuits to increase bargaining power over the silver trade.⁴³ The main target of the copper circuit was the city and kārum of Purušhattum (in Hittite, Purušhanda, Paršuhanda, or Parašhunta). This centre was long identified with Acemhöyük (Fig. 4.1), an excavated site in Cappadocia that yielded the vestiges of a prosperous MBA settlement that was intensively frequented by Assyrian merchants and entertained contacts with Old Syrian centers.⁴⁴ In recent years, however, Barjamovic advocated a more westerly localization that would seem more attractive or at least closer to a realistic scenario (Barjamovic 2011:357–378).⁴⁵ Barjamovic also argued that Purušhattum was an interstitial market of the Old Assyrian trade, lying at the juncture of multiple systems of exchange. In fact, transactions attested in Purušhattum suggest that this city was responsible for the inflow in Assyrian markets of huge quantities of silver and gold acquired from a network beyond the reach of Old Assyrian merchants. In exchange, the Assyrians could allow imports of equally large amounts of copper acquired in Durhumit into Purušhattum and allied networks. This evidence, in addition to fluctuations in prices conducive to a market economy of investment and profit, suggests that Purušhattum was, like Aššur, a gateway of commerce straddling multiple exchange networks (Barjamovic 2011:372–375). It seems clear, therefore, that Kaneš, Durhumit, and Purušhattum represented nodes in a sort of tripartite system that was meant to support the Old Assyrian network in Anatolia and in turn connect it with other circuits of exchange.

This organization of the Old Assyrian trade seems to have changed in the last phase documented by Old Assyrian sources, which corresponds to the Kārum Ib level at Kaneš.⁴⁶ Exchange during this period seems to have increasingly focused on local circuits in copper and wool, with a more limited inflow of tin and textile from Assyria. Likely related to this development was the growing number of Assyrians residing permanently in Anatolia (wašbūtum). At this stage, they became a group of their own, distinct from the Assyrians involved in the overland trade. Based on textual attestations of toponyms, scholars have also argued for a general contraction of the Kārum Ib network, especially its western branches. The relative paucity of Kārum Ib records, compared with those available for the previous generation, and the growing number of toponym attestations in this period, including a mention of Purušhattum (Kt 90/k 359), suggest caution against too pessimistic interpretations, which might be biased by the scattered nature of the findings. The apparent marginalization of western nodes can nonetheless be contrasted with the increasing importance of the Kızılırmak area and the eastern branches of the Old Assyrian trade (e.g., in Ma’ama, in northern Syria). This is perhaps linked with the coeval expansion of Syrian markets, chiefly Mari, which could have created new opportunities in this direction.

5 The Late Kārum Period and the Anitta Text (CTH 1)

The bulk of the Old Assyrian sources relate to the Kārum II period, roughly corresponding to the late 20th and 19th century BCE. Based on the eponym lists, the destruction of Kārum II at Kaneš can be situated around 1830 BCE.⁷² A documentary hiatus of about a decade followed, perhaps due to a temporary retreat of Assyrian merchants from the colony, together with their archives.⁷³ During the subsequent Kārum Ib period, the Kaneš colony was resettled and the Assyrian archives became active again, although far less so than before. Notwithstanding their small number (about 500), the Kārum Ib texts offer some rare glimpses into Anatolian historical facts, sometimes with a specificity unseen in the previous period. In particular, for the first time, several Kanešean kings and their highest officers come to be known by name. These details have been gleaned most often from notarization formulas in legal transactions, the so-called iqqāti documents.⁷⁴ Through comparison with other available sources, these attestations allow the reconstruction of a dynastic sequence that is likely complete for the Kārum Ib period at Kaneš: Hurmeli (?), Harpatiwa, Inar, Waršama (Inar’s son), Pithana, Anitta (Pithana’s son), and Zuzu.

An important historical source for the Kārum Ib period is the famous letter sent by Anum-hirbe of Ma’ama to King Waršama of Kaneš, written in Old Assyrian cuneiform and found in the so-called Waršama palace on the Kültepe mound.⁷⁵ This text relates to a diplomatic crisis between the two kings that was triggered by one of Waršama’s vassals, the ruler of Taišama. This personage is accused of having promoted unlawful alliances and raided Ma’ama’s territory, taking advantage of Anum-hirbe’s temporary weakness after a recent defeat. Taišama’s behavior is compared by antithesis to that of Anum-hirbe’s vassal state Šibuha, which never harmed Kaneš. The political geography reflected by this situation was radically different from that which was portrayed in the previous generation of OA sources. During the Kārum II period Assyrian merchants mostly moved in a landscape of small city-states that occasionally formed loose coalitions, whereas Anum-hirbe’s letter reveals a system of larger hegemonic formations that projected their power over one or more client states.⁷⁶ This process of territorial integration is somehow mirrored in coeval Mesopotamia by the more ambitious expansionistic projects attempted by Šamši-Addu in the north and, later, Hammurapi and his successors in Babylonia. In the case of Ma’ama and Kaneš the scale was certainly smaller, but still impressive relative to the Anatolian scenario. At Mari and in later Hittite historical legends, the same Anum-hirbe is attested as the king of both Zalwar and Haššu. This information, in addition to a much later mention of Anum-hirbe’s monument on Mount Atalur that was reported in the annals of Salmanassar III, argue that Anum-hirbe’s domain should be localized in the Antitaurus area, between the Kara Su River valley and modern Maraş (Fig. 4.1).⁷⁷ If so, regardless of the positions of Taišama and Šibuha, which cannot be ascertained, the hegemonic spheres of Ma’ama and Kaneš must have been quite large, perhaps encompassing on one side or the other such major centers as Šalahšuwa or Luhuzattiya.⁷⁸ Incidentally, Anum-hirbe also informs us of an earlier war conducted by Waršama’s father Inar against the land of Haršamna, which probably lay at a short distance from Kaneš.⁷⁹ Anum-hirbe and his deeds had a later echo in Hittite scribal circles, as Hittite historical legends mentioning this king in connection with the cities of Zalwar and Haššu are preserved in a few fragmentary tablets (CTH 2).⁸⁰

The most important historical source about the Kārum Ib period is preserved in a Hittite account of military deeds (CTH 1) that is told in the first person by Anitta, son of Pithana.⁸¹ According to this text, Anitta and Pithana, natives of the city of Kuššar, conquered Kaneš (spelled Neša), and Pithana sat on its throne after overturning the local ruler. After succeeding his father, Anitta used Kaneš as a power base to launch military campaigns over a vast area that comprised the Kızılırmak bend and the territory of Zalpuwa, on the Black Sea. After prevailing over the latter, Anitta took its king, Huzziya, to Kaneš as a hostage, together with a statue of the Kanešean Sun god, that a former king of Zalpuwa named Uhna had abducted as booty. Within the Kızıl Irmak bend, a major rival defeated by Anitta was Piušti, the king of Hattuša (LUGAL ^URUHATTI). Hattuša was taken by night, destroyed, cursed, and made symbolically unproductive by sowing weeds on its terrain. In a second phase of expansion, Anitta, now claiming the status of ‘great king’ (LUGAL GAL), concentrated his efforts on the western side of the Kızılırmak, finding strong opposition from the ‘Man (LÚ) of Šalatiwara.’ This ruler eventually crossed the river Hulanna and unsuccessfully besieged Kaneš. A brief excursus on Anitta’s building and hunting activities is inserted in the account of these events. Finally, Anitta turned to Purušhanda, whose ruler is presented again as a ‘Man.’ This time, however, the conflict was resolved peacefully, as the Man of Purušhanda acknowledged Anitta’s power and became his ally.⁸²

The Anitta Text would thus convey the memory of an attempt at constructing a hegemonic polity in central Anatolia at a point predating the expansion of Hattušili I. The composition is preserved in several manuscripts composed in Hittite during the Hittite period (KBo 3.22, OH/OS, is the oldest), but it likely reflects, at least in part, authentic accounts dating back to the Old Assyrian period. Indeed, were it a tradition invented from scratch for the sake of Hittite propaganda, it would be unclear how to fit it with the extant evidence.⁸³ Clear analogies might have sparked the interest of Hittite governing elites in transmitting the Anitta text: both Anitta and the founder of the Hittite kingdom, Hattušili I, were from Kuššar and founded successful regimes elsewhere. Departures, however, are equally evident and can hardly be reconciled with Hittite ideological narratives. The passage on the curse of Hattuša, already present in the OH/OS manuscript, is the clearest example: what is the political gist of it for a Hittite subject, considering that Hattuša was the Hittite capital from Hattušili I on? Hattušili I not only founded his kingdom in the city that Anitta had destroyed and cursed but proudly took his name from it, imitated by at least one successor (Hattušili III).

The historicity of Anitta and his father Pithana as pre-Hittite Anatolian rulers is confirmed by independent evidence, as are some individual details of the Anitta Text. A sequence of two kings named Pithana and Anitta is attested at Kaneš during the Kārum Ib period. Related sources include a series of iqqāti documents, some drawn up under the supervision of Pithana and/or Anitta, and a bronze spearhead marked as belonging to ‘the palace of Anitta, king’ (É.GAL ^mA-ni-ta ru-ba-im). Other iqqāti documents attest Anitta’s kingship at Amkuwa, confirming his hegemonic rule over more than one urban center of the Kızılırmak area.⁸⁴ In one of these documents (OIP 27 49a/b), Anitta bears the OA title of ‘great king’ (ruba’um rabûm), that matches the title LUGAL GAL that he claimed in his text (KBo 3.22 obv. 41).

A ruba’um named Wiušti, likely identical with the Piušti mentioned in the Anitta Text, ruled at Hattuš during the Kārum Ib period as is attested by the OA letter KBo 71.81.⁸⁵ This letter was found in a storage room of an administrative building located in the kārum quarter of Boğazköy-Hattuša. The eponym date of another OA document found nearby (KBo 71.95) provides a terminus post quem for the destruction layer sealing this building of 1748 BCE,⁸⁶ which is compatible with the supposed date of Anitta’s destruction of Hattuša (ca. 1730 BCE).⁸⁷ In sum, there are no reasons to be too skeptical about the historicity of the political scenario depicted in the Anitta Text.

The question remains of how the memory of Anitta’s deeds reached the Hittite court and was then transmitted during the Hittite period. The Anitta Text appears to be a compilation that drew on multiple sources. A boundary between at least two different original archetypes is found in KBo 3.22 obv. 33–35, in which Anitta concludes an account of the wars against Neša, Zalpuwa, Hattuša, and other localities with this statement:

These words […] from a tablet at my gate. In the future may no one destroy this tablet. Whoever destroys it, let him be an enemy of Neša

Since this formula would be out of place in the middle of a coherent composition, it is generally interpreted as the concluding paragraph of a separate inscription dedicated by Anitta himself, presumably on the gate of his palace or a city gate at Kaneš. Most scholars, therefore, maintain that this was an original inscription of Anitta that was later blended with another account of his exploits (ll. 36–79) to form the Anitta Text as we know it.⁸⁸ As others propose, however, the Anitta Text might have had a more complex redactional history, originating from the combination and further manipulation of three or possibly even more original sources.⁸⁹

Closely connected with the issue of transmission is the question of the language in which the Anitta Text was originally composed, which is part of a broader debate over the early stages of Hittite literacy. Apart from very few outliers in forms of Akkadian possibly linked with Syrian or other Mesopotamian circles (see below), all known records of the Kārum period were produced using the OA script and language. Old Assyrian was not only used by both Assyrian and Anatolian merchants but also was the official language of Anatolian chanceries, as exemplified by the iqqāti documents and letters treated above. Therefore, most scholars assume that the inscription ‘of the gate’ referred to in ll. 33–35 and the other source(s) conflated in the Anitta Text, were first recorded in Old Assyrian Akkadian and then translated into Hittite.⁹⁰ The later step would have occurred sometime during the Hittite period, that is, the time when the earliest known records written in Hittite were composed.

This appears to be the most reasonable scenario, judging from the evidence available so far. Although not conclusive, some positive evidence in the narrative suggests a phase of elaboration of the text within an Akkadian scribal environment.⁹¹ The emphasis on building activities and the motif of the hunting king was not part of the usual Hittite apologetic repertoire, at least until the Late Empire period, but better matches Mesopotamian traditions.⁹² To this observation, one may add Dercksen’s (2010) proposed comparison between the gift presentation made by the Man of Purušhanda and diplomatic protocols in use in Middle Bronze Age Syria and Mesopotamia. A linkage with OA sources is suggested more directly by the geography of Anitta’s actions, which relates to a map of Anatolia probably more familiar to Old Assyrian merchants than Hittite scribes. Šalatiwara, in particular, was a prominent node in the Old Assyrian network and well attested in the related corpus (cf. OA Šalatuwar). By contrast, this toponym is almost absent in the Hittite corpus, and the few attestations outside the Anitta Text are also suspected to stem from a knowledge of Old Assyrian compositions.⁹³ These texts and other sporadic sources, including the Anum-hirbe legends mentioned above, prove that Hittite scribes had an access to Old Assyrian material, perhaps including narratives of Anitta’s deeds.⁹⁴

In his edition of the Anitta Text, Neu (1974) rejected the hypothesis of a translation from Old Assyrian sources, finding no sign of interference with Old Assyrian in the vocabulary and grammar of the text. He argued therefore that Anitta had the account of his deeds written directly in Hittite, that is, nešili, the native language of the inhabitants of Kaneš. Considering the lack of secure traces of record keeping in Hittite before Hattušili I, Neu’s reconstruction did not receive wide acceptance. In a recent article, Kloekhorst and Waal (2019) discuss a group of “cushion-shaped” tablets written in Hittite, including the earliest known witness of the Anitta Text (KBo 3.22; OH/OS), arguing that they were composed outside Hattuša, before the Hittite Old Kingdom. Although this is theoretically possible, there is no fundament in the suggestion of the two authors that these tablets date to the Kārum Ib period.⁹⁵ More realistically, we can suppose that Hittite narrative styles and forms, after reaching a mature stage, influenced the translation of the Anitta Text to the point of masking any legacy of Old Assyrian. Moreover, it is theoretically possible, although not provable, that, during the transmission of the Anitta Text, OA material concerning Anitta was knitted together with other sources written (or orally transmitted?) in other languages. This would have introduced further distancing from the language of any original sources.

The last king known at Kaneš is Zuzu, estimated to have succeeded Anitta around 1725 BCE and ruled until the end of the 18th century.⁹⁶ The last dated text from Kültepe was produced around 1710 BCE. Sometime after this date the settlement of Kārum Ib came to an end for unclear reasons. During the following occupation phase, Kārum Ia, which may have lasted until the rise of the early Hittite kingdom, Kültepe was much reduced in size and did not yield any texts. After this phase, the settlement was abandoned until the Iron Age, perhaps due to permanent flooding caused by a rise in the water table.⁹⁷ Elsewhere—at Acemhöyük and Konya-Karahöyük—the last Kārum-period settlements were violently destroyed sometime during the 18th century. This would signal an intensification of political competition and warfare that would accord well with Anitta’s account. As mentioned above, the Kārum-period settlement of Hattuš(a) was also destroyed around 1730 BCE, likely due to Anitta’s conquest. Giving credit to Anitta’s claims to have cursed the city and banned resettlement, scholars have long assumed that Hattuš(a) was abandoned for some generations after its destruction. Recent investigations, however, reveal that the kārum area was immediately resettled, albeit in a more haphazard fashion, in the so-called Zwischenphase, radiocarbon dated after 1720 BCE.⁹⁸

6 Non-Old Assyrian Commercial Networks

As already stressed (§ 1), the textual evidence does not permit tracking sociocultural developments in a pan-Anatolian perspective as the kārum network only occupied the eastern and central part of the peninsula. However, it is clear that the Assyrians were not alone in the commercial landscape of the Near East but part of a wider system of interlocking networks that connected central Asia with southeastern Europe.⁹⁹ In this context, Aššur was just one among several hotspots in Near Eastern trade, others being in Lower Mesopotamia (Sippar), the middle Euphrates (Mari and Karkemiš), the Levant (Ebla, Aleppo, and Ugarit), and the Aegean (Minoan Crete). Likewise, there is reason to think that several interacting circuits or commercial spheres operated within Anatolia (Fig. 4.2).

View Full Size

Figure 4.2

Old Assyrian and other main competing commercial spheres argued in the text

• Download Figure
• Download figure as PowerPoint slide

Available textual evidence for non-Old Assyrian trading activities in Anatolia is scanty but probably represents the tip of the iceberg of a larger corpus now lost or still buried in undiscovered archives. Some OA texts inform us about contacts with Ebla, whose merchants were involved in the Anatolian copper trade.¹⁰⁰ In addition, an interesting small group of texts from Kaneš displays linguistic and paleographical features foreign to OA traditions, which can be linked to non-Assyrian scribal environments in north Syria and Upper Mesopotamia.¹⁰¹ Within this outlier group, the letter Kt k/k 4, sent by a certain Ehli-Addu to an Unap-Še—both Hurrian names—concerns affairs taking place in Tunip (northern Levant), and closes with a list of ‘witnesses of the city of Haššu’ (šibū ša ālim Hašši), located in the vicinity of Maraş or Gaziantep.¹⁰² Neither Tunip nor Haššu were involved in the Old Assyrian trade. Finally, an oft-cited OA text warns an Anatolian ruler against dealing with ‘Akkadian’—that is, Babylonian—merchants,¹⁰³ whose likely presence in Anatolia is attested by the diffusion of Old Babylonian-style seals and impressions.¹⁰⁴

Relations between Mari and Anatolia are also documented.¹⁰⁵ The Mari letter T.135 attests exchanges with Purušhattum during the reign of Yahdun-Lim, contemporary with the Kārum II period at Kaneš, and, probably somewhat later, a sealing belonging to a daughter of Yahdun-Lim reached the Sarıkaya palace at Acemhöyük.¹⁰⁶ Contacts continued during the reign of Šamši-Addu, a period when the activities of the Old Assyrian caravan enterprises are poorly documented. Several sealings of this king or his officials feature in the Sarıkaya archives at Acemhöyük and lie at the core of ongoing debates on Bronze Age chronology.

A key port of entry into Anatolia from the Middle Euphrates and the Jazira was the city and kingdom of Karkemiš (Gaziantep province). The best documented king of Karkemiš during the MBA is Aplahanda, who sent sealed goods to the palace of Acemhöyük. Control of western trading routes via Karkemiš likely was a main motivation for the expansionist policies of Šamši-Addu, who subjugated Aplahanda, perhaps together with the countries of Haššu and Uršu, and tried to establish diplomatic relationships with Zalwar, generally identified with Tilmen Höyük in the Amanus region (Fig. 4.1).¹⁰⁷ After the death of Šamši-Addu, when the Old Assyrian trade restarted in Kārum Ib, Aplahanda regained his independence in Karkemiš, as did Zimri-Lim in Mari. In this period, a trade-oriented partnership between Mari and Karkemiš thrived. The respective chanceries corresponded on commercial matters, including shipments of goods from Kaneš, Hattuš, and even as far west as Šalatuwar. However, some records unveil Mari’s efforts to establish a direct commercial link to Anatolia that would bypass Karkemiš.¹⁰⁸ These attempts entailed engaging in direct competition with Assyrian merchants or, alternatively, seeking their complicity. Some Mari letters bear witness to a proposal for a marriage alliance that was advanced by an Assyrian merchant residing at Kaneš and addressed to the chief of Mariote merchants, Iddin-Numušda, alias Iddiyatum.¹⁰⁹

These few hints concur with the OA sources in depicting the involvement of Anatolia in a diverse landscape of competing commercial agencies that operated across partially overlapping spheres of interest. Compared with the directional caravan trade documented by the OA sources, the little information available on Syrian enterprises seems to reflect more flexible and indirect patterns of trade that relied on the mediation of multiple, independent interstitial markets. The OA ban on Babylonian merchants mentioned above, as well as similar attested territorial restrictions,¹¹⁰ have been used to argue that relationships between the different networks responded to mercantilistic logic and were aimed at the creation of areal monopolies.¹¹¹ Nonetheless, the fact that limitations had to be imposed through specific regulations may suggest that the markets involved in these circuits had normal aspirations to attract an array of firms aiming to increase demand. This would be conducive to the development of permeable boundaries between different circuits—a situation that, for example, is apparent in the strong spatial overlap and stylistic interpenetration of the Anatolian, Old Assyrian, Old Babylonian and Old Syrian glyptic styles.¹¹²

Archaeological evidence complements textual sources in suggesting multiple intersecting channels of exchange between Anatolia, Syria, Mesopotamia, and the Aegean, which unfolded in part along trajectories already established in the EBA.¹¹³ Notwithstanding differing historic-geographic views, scholars agree that the Old Assyrian network only marginally touched upon areas south of the Tuz Gölü and did not extend west beyond the Sakarya river basin. Significantly, this areal extent bypassed several regions known for their relevance in interregional connectivity throughout the Bronze Age. The most striking case is Cilicia, which was a well-known natural passage and had been one of the main gateways in overland connections between central Anatolia and the Levant since early prehistory.¹¹⁴ Convincing evidence points to Cilician involvement in one or more non-Assyrian networks. As mentioned above, Zalwar/Tilmen Höyük, at the eastern entrance of Cilicia, was a key node in commercial and political interactions with Mesopotamia and the northern Levant. Tending to confirm this picture are the Old Syrian-style sealings with inscriptions in OB ductus that have been found on this site, seemingly attached to exchanged goods.¹¹⁵ Tilmen lay close to the Amanian Gates (nowadays the Bahçe Pass) that gave access to Plain Cilicia and controlled traffic in this direction. In Cilicia, MBA evidence from Sirkeli comprises a number of finds testifying to contacts with northwest Syrian centers, as well as an imposing upper- and lower-town complex that was comparable in size to the citadel and kārum compounds found at Old Assyrian trading posts.¹¹⁶ Cylinder seals conforming to MBA northwest Syrian traditions have been found at Tatarlı Höyük,¹¹⁷ and Tarsus has yielded an Old Babylonian seal.¹¹⁸

The clearest testimony of a material-cultural convergence between Cilicia and its eastern neighbors during the MBA is represented by the so-called Syro-Cilician, or Amuq-Cilician, ware. Named after its main distribution and production area, which straddled the Amanus Mountains, this is a class of both wheel- and handmade ceramics characterized by distinctive matte-painted decorations that include geometric, floral, or zoomorphic motifs.¹¹⁹ While this ware was found in Mesopotamia and through the Levant down to Egypt, it is the main diagnostic feature of MBA ceramic traditions in Plain Cilicia.¹²⁰

Cilicia was well connected with central Anatolia through a circuit that interlocked with the Old Assyrian network. Stamp seals and impressions following native central Anatolian traditions featured at Sirkeli and Tilmen Höyük and found their way to Ebla and Tell Bi’a in Syria.¹²¹ Crescent-shaped loom weights, typically used for weaving in central Anatolia, were also common in Cilicia.¹²² Additionally, this region might have acted as a channel for the diffusion of Old Syrian- and Old Babylonian glyptic traditions to the north and west.¹²³ Syro-Cilician ware imports crossed the Taurus, reaching the Central Plateau.¹²⁴

Due to its central position and rich settlement history, the Konya Plain must have been another major area of interaction during the MBA.¹²⁵ This region was only marginally involved in the Assyrian trade through the wabartum of Uš(š)a, which is, however, poorly attested in OA sources.¹²⁶ The published archaeological record for the MBA in the Konya Plains, almost exclusively limited to the glyptic corpus excavated at Konya-Karahöyük, offers some supplementary material.¹²⁷ According to Barjamovic (2019:75), the main commercial partner of the inhabitants of the Konya Plain could have been Ebla, a hypothesis supported by the finding of several Old Syrian-style cylinder seals and impressions at Konya-Karahöyük. If so, the Konya Plain would have been closely tied with Cilicia, possibly along trajectories of contact developed in the EBA (Chapter 3).

Western Anatolia was also involved in exchange circuits that interacted at various levels with the Old Assyrian and other networks. Use of Aegean weighting systems is documented at Kültepe and further south and east, at Ebla and in Upper Mesopotamia. This would point to a local adaptation or acquaintance and close interactions with western mercantile practices, likely through both maritime and overland routes.¹²⁸ Archaeological evidence for the early second millennium in the whole region west of the Kızılırmak bend is scattered and problematic. Coupled with the almost complete lack of relevant textual information, this situation allows for only a vague assessment of possible trajectories of contact in the area.¹²⁹ Sandwiched as it is between two major catalysts of scholarly attentions, namely, the proto-Hittite and Minoan core regions, western Anatolia has not been seen until recently as a subject in its own right as far as the second millennium is concerned. It is symptomatic of this state of the art that only a few sites have been excavated and even fewer fully published, and that the great majority are located in the coastal areas under strong Aegean influence. The major excavation carried out in the 1950s at Beycesultan, in the upper Meander, was chiefly aimed at investigating Arzawa, a composite political entity best known as a rival of the Hittite kingdom. Comparatively little has been published and little is known about the second millennium BCE in the inland intermediate areas of western Anatolia.¹³⁰

Chronological uncertainties further contribute to the general fuzziness, interfering with attempts to correlate materials from different sites in broader regional and supraregional perspectives. The two stratigraphic pillars for the second millennium sequence, Troy and Beycesultan, are located some 500 km apart and belong to distinct cultural horizons. At Troy, the MBA has long been a sort of phantom, squeezed between settlements VI and VII, which are assigned to the LBA, and settlement V, which is often considered an appendix to the cultural developments of the EBA. Recent chronological reassessments, however, squarely bracket Troy V between the 20th and the mid-18th centuries BCE, that is, within a span broadly parallel to Kārum II to Ib in central Anatolia.¹³¹

In the large, two-mounded site of Beycesultan, early excavations by Seton Lloyd and James Mellaart exposed two levels (V–IV) assigned to the MBA chiefly through relative dating and associations with historical events.¹³² The main feature of level V is the Burnt Palace, an imposing building with an area of over 4.5 km² and multiple rooms organized around a main courtyard. After the violent destruction of this building, the area was occupied by squatters in level IV. Mellaart associates this transition with the conflict between Hattušili I and Arzawa that he dates to 1750 BCE following the now abandoned High Chronology. The next level, III, featuring another institutional edifice, the Little Palace, was dated to the LBA (15th–13th centuries BCE), primarily based on the finding of a fragment of a Mycenaean LH IIIA/B stirrup jar, “embedded in a platform of Late Beycesultan III date.”¹³³ Various scholars argued for an earlier dating of Beycesultan level III considering some ceramic similarities with Kārum Ib ‘Old Hittite’ traditions from central Anatolia.¹³⁴ The excavations restarted in 2007 under the direction of Eşref Abay produced a revised stratigraphic sequence supported by absolute radiometric dates that confirm this early date or perhaps argue for an even earlier dating.¹³⁵ In this updated framework, the MBA occupation(s) at Beycesultan includes levels V, IV, and III, covering the 19th to the late 18th century BCE.¹³⁶ The single Mycenaean sherd is hardly significant in the face of the recent more solid evidence: it should probably be interpreted as an intrusion from later layers. Based on comparisons with Troy V, Beycesultan levels VII to VI likely date from the early MBA, around the 20th century BCE.¹³⁷

Notwithstanding stratigraphic shifts, the cultural frameworks in Troy V developed without major breaks from the local traditions of the late third millennium BCE (Troy IV). Red Cross Bowls, a class of carinated bowls so-called for their distinctive decoration, constitute the only remarkable innovation, and provide an interesting anchor point for connections with Anatolian contexts farther east.¹³⁸ Close comparanda for this ceramic class are found from northwest Anatolia to the Kızılırmak bend, including Kültepe (Kārum IV–III) and southeastern Anatolia. However, while Red Cross Bowls are found throughout the MBA sequence at Troy, elsewhere this ceramic is conspicuously limited to early MBA or transitional EBA–MBA contexts. Particularly close matches are found in Cilicia, specifically in Mersin XIb, Tarsus EBA III–MBA Transitional, and Kilise Tepe Vf-e.¹³⁹ At Beycesultan Red Cross Bowls are abundant in levels VII to VI but absent in later MBA deposits.¹⁴⁰ This pattern seems to indicate that, early in the second millennium BCE, contacts with western Anatolia continued along overland routes broadly consistent with the Great Caravan Route evidenced for the EBA III (see Chapter 3), having a main axis between Cilicia and the Troad. The later retreat of Red Cross Bowls traditions within Troy V might indicate a shift toward other trajectories during subsequent phases of the MBA, but the locations of those hypothetical trajectories remain undetermined.

Kārum-style and early LBA traditions would find a northwestern limit at Gordion, where MBA layers have been investigated in test pits under Megaron 10 and 12 and the so-called Hittite Cemetery.¹⁴¹ Many have also pointed to an influence of Kārum-style traditions in the ceramic production of Beycesultan V to IV, allegedly manifested in pitchers with pronounced beak spouts, basket-handled jars, trefoil jugs, teapots, and carinated bowls.¹⁴² Mellaart insists that these shared traits have a common areal heritage in EBA III wheel-made pottery horizons rather than developing out of synchronic interregional contacts.¹⁴³ However, closer—if still tenuous—interactions with central Anatolia are unanimously identified in Beycesultan III assemblages and are now understood to reflect the late MBA phases on the site.¹⁴⁴

The complex architectural layout of the institutional buildings uncovered at Beycesultan levels V to III, with their annexed storage facilities, points to a centralized socioeconomic organization not dissimilar from those observed at Kültepe and other kārum sites. Likewise, the participation of the MBA Beycesultan community in commercial activities probably was a catalyst for the wealth accumulation necessary to sustain this complex organization. Cultural influences associated with Beycesultan MBA horizons traveled along the Meander and Hermos valleys (e.g., Aphrodisias), reaching the coast at Miletus, Liman Tepe, and Panaztepe.¹⁴⁵ Here the overland routes met maritime networks, chiefly signaled by Minoan(-style) artifacts.

The sites of Miletus, Çeşme-Bağlarası, and Iasos of Caria likely hosted Minoan settlements, and Minoan imports feature in all of the MBA assemblages along the Aegean coast from Knidos to Troy.¹⁴⁶ No trace of direct Minoan influx has yet been found further inland. However, Aegean trade with central Anatolia, albeit mediated by interstitial markets, is evidenced by the distribution of Aegean weighting systems that was mentioned above. At this juncture, it is also worth mentioning a single Anatolian-style stamp seal impression found in an MBA deposit at Phaistos in Crete that has close comparanda at Konya-Karahöyük.¹⁴⁷

For the 19th to 17th centuries BCE, cuneiform sources from Mari, Alalah, and Babylon bear the earliest textual evidence for the overseas exportation of metals, chiefly copper, from Cyprus (ancient Alašiya).¹⁴⁸ Significantly, these references concur with the archaeological record to show an increasing involvement of Cyprus in Near Eastern trading networks during the MBA. Exotica imported or showing influences from the Levant, Egypt, and Anatolia increased during this period on the island, reciprocated by a widespread distribution of Cypriot exports, chiefly ceramics, across the eastern Mediterranean.¹⁴⁹ This exchange was accompanied by a boost in metal production on Cyprus, likely in response to external demand.¹⁵⁰

In Anatolia Middle Cypriot White Painted III–IV vessels have been recovered in a Kārum Ib context at Kültepe and at another unspecified location in the Kayseri region. In contrast with later periods, no MBA Cypriot influx is found in Cilicia, barring a single White Painted III–IV style sherd from an unclear chronological context at Sirkeli Höyük. Therefore, Cypriot contacts probably reached central Anatolia through Syrian mediation and then spread overland across the Antitaurus.¹⁵¹ One way or another, the availability of Cypriot copper in Anatolia may have diminished the purchasing power of the Pontic copper that had hitherto dominated the Old Assyrian markets and, from thence, other interlocking circuits.¹⁵²