Central Asia in Antiquity: Interdisciplinary Approaches 9781407313115, 9781407342726

Table of contents :
Front Cover
Title Page
Copyright
CONTENTS
PREFACE
Exploring the Oases of Central Asia
Aryan warriors
First notes on a treasure from Mes Aynak
The Second after the King and Achaemenid Bactria on Classical Sources
Achaemenid Royal Women and Alexander
When Greece came to India
Ancient vardāna
The Italian–Uzbek archaeological mission at Uch Kulakh tepe
Alexander the Great in Sogdiana: A study in historiography
Nero, Parthia and the ‘Armenian Question’
The relations of the Sasanian Empire with the Eurasian Steppe
From the conquest of al–Andalus to the Muslim majority

Citation preview

Central Asia in Antiquity: Interdisciplinary Approaches Edited by

Borja Antela-Bernárdez Jordi Vidal

BAR International Series 2665 2014

ISBN 9781407313115 paperback ISBN 9781407342726 e-format DOI https://doi.org/10.30861/9781407313115 A catalogue record for this book is available from the British Library

BAR

PUBLISHING

CONTENTS

Preface

1

the EDITORS

Exploring the Oases of Central Asia A model of interaction between mobile livestock breeding and sedentary agriculture

3

Andreas ANGOURAKIS

Aryan warriors The Invented Heroes in the Ancient Near East

17

Jordi VIDAL

First notes on a treasure from Mes Aynak

23

Barbara FATICONI

The Second after the King and Achaemenid Bactria on Classical Sources

37

Manel GARCÍA SÁNCHEZ

Achaemenid Royal Women and Alexander

43

Clàudia ZARAGOZÀ

When Greece came to India. Herodotus and the Eastern boundaries of the world

53

César SIERRA MARTÍN

Ancient vardāna Results of the last archaeological investigations

61

Silvia POZZI

The italian-uzbek archaeological mission at Uch Kulakh tepe Ilaria VINCENZI

67

World is not enough: Alexander the Great in Sogdiana. A study in historiography

77

Borja ANTELA–BERNÁRDEZ

Rex armeniis datus? Nero, Parthia and the ‘Armenian Question’

85

Joan OLLER GUZMÁN

The relations of the Sasanian Empire with the Eurasian Steppe

99

Núria OLAYA MONTERO

From the conquest of al-Andalus to the Muslim majority Catalonia between the eighth and the tenth century Ramon MARTÍ & Mercè VILADRICH

105

PREFACE Central Asia is a wide subject of research in the archaeological and historical studies of the Ancient World. Scholars have been usually focused in the complex and diverse questions resulted from the analysis of the historical realities of this key region during Antiquity. Following this kind of interests, the purpose of the next pages is to undertake an approach to the polymorphic and multiple aspects of Central Asia in Antiquity from several points of view. The starting point is the confidence in an interdisciplinary perspective as the main way to understand the different aspects of the region in a very wide chronology: from the emergence of the cities and their relation with the nomadic populations, to the expansion of models and practices from Central Asia to the West during the campaigns and conquests led by the Islam. Concerning the subjects covered by the present volume, it deals with warfare, gender studies and historiography, mainly from an archaeological point of view. This book is the result of a workshop organized in the Autonomous University of Barcelona in 2012, in a date very near to the Ides of March (March 11 th). The aim of the organizer was to reunite the scholars that, around Catalonia, are working in questions related to the Ancient Central Asia, in order to get some discussion, to put questions together and, specially, to introduce to each other the own subjects of research. Moreover, as far as the Central Asia is a very attractive theme for students, the workshop had given them the opportunity to know the state of research on this question in their immediate academic framework. Bearing in mind this purposes, the workshop was, we think, a great success, and we can be very sure that it will have continuity, in a biennial format. This book reunites most of the contributions submitted during the meeting. However, and for several reasons, not all researchers involved in the workshop have been able to submit their papers. Nevertheless, these proceedings include some articles by leading scholars in Catalonia who could not attend to the workshop. To sum up, the final result is this book, which we think is a useful tool both for archaeologists, historians, classicists and everyone interested in Central Asia in Antiquity. Finally, this volume has been published thanks to the support granted by several persons and institutions. Mainly, the help provided by prof. Agustí Alemany (Autonomous University of Barcelona), both personal and economical, as coordinator and responsible of the Research Group Fontes Eurasiae Septentrionalis Antiquae et Medii Aevi (2009-2013) (2009SGR18), was essential to overcome many difficulties. In this sense, the edition of this volume has counted with two grants for junior scholars, Dr. César Sierra Martín and Ms. Ilaria Vincenzi (both at the Autònomous University of Barcelona), in order to help with the tasks of edition. Also, the workshop had being supported both by the Deanship of the Faculty of Humanities and the Department of Sciences of Antiquity and the Middle Ages, both of the Autonomous University of Barcelona. Editors are also in debt with the scholars who gently participate in the workshop (Sebastian Stride, Cl. Rapin, Agustí Alemany, Bernardo Rondelly, Jordi Cortadella and Dario Nappo) and those who trust in the book (César Sierra, Clàudia Zaragozà, Ilaria Vincenzi, Núria Olaya, Andreas Angourakis, Ramón Martí, Mercé Viladrich, Manel García Sánchez, Silvia Pozzi and Barbara Faticoni). Last, but not least, our families had to suffer sometimes the difficulties that the research about Central Asia in Antiquity had to face. B. ANTELA-BERNÁRDEZ & J. VIDAL, the Editors.

Exploring the Oases of Central Asia* A model of interaction between mobile livestock breeding and sedentary agriculture ANDREAS ANGOURAKIS ERAAUB–University of Barcelona ABSTRACT Central Asia is an area in which the main variants of pre–industrial economic productions coexisted for several thousand years. The relation between the different economic activities is known to have been unstable, with abundant examples of conflict at different geographical scales and most places having seen shifting patterns of land use change right up to the early 20th century. Here we present and discuss a model of interaction (implemented using Agent–Based modeling) between mobile livestock breeding and sedentary agriculture, the Musical Chairs model. The model is. To complement a previous publication (Angourakis et al. 2014), this article deepens on the analysis of assumptions, implications and stable states (i.e., attractors) involved in this model. KEYWORDS Land use – Oasis – Central Asia – Agent–based modeling – Coevolution – Herding and Farming Modeling land use in oases of Central Asia 

The problem and its context For at least three millennia, Central Asia was the stage of frequent interaction between the main variants of pre– industrial food production, ranging from fully nomadic pastoralism to modes of agriculture highly reliant on irrigation. Given that most of this subcontinent is too arid to sustain sedentary agriculture, this interaction was not ubiquitous, but in fact concentrated in what we came to call oases. Regarding Central Asia and most of the arid Eurasia, oases are equivalent to alluvial cones and plains, whose rivers are mainly feed by distant rainfall and the seasonal melting of glaciers and snow cover (Oberhänsli 2012).



This contribution is part of the activities of the Equip de Recerca Arqueòlogica I Arqueomètrica, Universitat de Barcelona (ERAAUB), as a research group consolidated by the Generalitat of Catalonia (2014 SGR 845). The author is in debt particularly with the SimulPast project (Consolider Ingenio 2010, CSD2010–00034), funded by the former Spanish Ministry of Science and Innovation (MICNN). Among numerous collaborators, this work is mostly obliged to Bernardo Rondelli (Complexity and Socioecological dynamics–CaSEs, Istitució Milà I Fontanals–Spanish National Research Council), and Sebastian Stride, Josep Maria Gurt and Verònica Martínez (ERAAUB), who conveyed their experience in the archaeology of Central Asia.

Oases themselves attest that the aforementioned interactions were not at all stable. Both historical and archaeological records revealed several situations in which land use radically changed (Stride 2005, in Southern Uzbekistan; Nesbitt and O’Hara 2000, in Southern Turkmenistan; Alizadeh and Ur 2007, in North– Western Iran; Abdi 2003, in Central Zagros; Newson 2000, in South–Eastern Syria). Starting with the very diffusion of agriculture, it appears that farming–i.e. crop cultivation and small–scale livestock breeding—was able to progressively claim the use of oases. As oases are first and foremost natural havens for animals and water– demanding plants, many wild species and the livelihoods relying on them–such as hunting, fishing and gathering— found themselves competing for (and losing) these key sites against this new sedentary lifestyle and its constructed ecosystem (Rosenberg 1998). The latter development and implementation of technologies to control the flow of water, although widening the oasis itself, have further empower this process, rendering also marginal lands more and more attractive to cultivation. Contrasting with its relationship with hunting, fishing and gathering, farming displayed more complex interactions with the different varieties of mobile livestock breeding. The latter, generally called pastoralism, emerged in a secular process alongside farming itself, driven by the relentless demand of animal products of the growing sedentary population. As the productivity of arid lands is

ANDREAS ANGOURAKIS highly constricting to livestock breeding and only a reduced number of animals can be sustained throughout the year in the same locality, mobility proved to be the best strategy to meet this demand at the long run. In fact, the ‘herding solution’ was probably not given by any human enlightenment, but it presented itself as the obvious way to proceed with animals that were already highly migratory in the wild (e.g., Martin 2000: 22, on gazelle behavior).

fertility and identity of the land. Farming and herding activities are indeed qualitatively different regarding land use, involving different logistics, ecological constraints and timings. Although our research pursues the conditions in which there is competition for land use between these two livelihoods, they are not necessarily mutually exclusive, as they can even draw upon each other (e.g. fodder for manure), provided that they are sufficiently coordinated (Stride et al. 2009). We should also keep in mind that the value of land for these livelihoods is diametrically different: herding implies visiting different pastures throughout the year, often within a vast territory, with little or no investment on particular patches of land; farmers on the other hand will only be interest in the few areas where cultivation is feasible, which shall be altered through intensive–and costly—activities.

Even though livestock breeding is an integrated part of any farming economy across Afro–Eurasia, the increasing specialization in the mobile varieties cause some people to develop specific interests, which often diverged from those engaged in farming. Although a complete independence between these livelihoods is hardly the case–and for most scholars, it is virtually impossible (e.g., see Khazanov 1994)—, this divergence do produces very tangible cultural distancing. A Sumerian fable, telling of the dispute between Enkimdu the farmer and Dumuzi the herder for the hand of the goddess Inanna (Pritchard 1955: pp.41–2; Moscati 1960: pp. 42–3), exemplifies how socially significant this situation can be, even within the same cultural community. Once Inanna declares to favor the proposal of Enkimdu, Dumuzi defends himself by comparing their products:

Despite these caveats, there is no reason to automatically assume that there is a complete separation between the niches of farming and herding activities, at least in terms of land use. Saving all differences, both boil down to requiring land surface and its biomass capacity, mainly limited by soil moisture in the growing seasons. The various forms of mobile livestock breeding in arid Eurasia, which array from daily local movements to long distance non–linear movements, have all one aspect in common: oases are benchmarks. Overall, herders have many incentives to visit and stay as long as possible in oases. Even if the quality and abundance of pastures in and around oases is inferior to the ones elsewhere (e.g. high altitude pastures during summer), the harsh seasonal cycle and the scarcity of denser vegetation will regularly impose oases as indispensable refuges to herds. These incentives are not only ecological (i.e. access to water, higher temperature in winter), but also social, political and economic, given that oases enable herders to congregate with their families, clans and tribes, plus having access to trade–and sometimes pillage— opportunities offered by the settled population. These latter incentives would of course be closely related to how much area can be covered with less than a day journey from the settlements of an oasis. In this sense, it should be also taken into account that herding leaves very little and dispersed archaeological footprints (Barnard and Wendrich 2008, Cribb 2004), though it was already proved to be directly assessable in marginal lands (e.g., Rosen 1993).

Enkimdu, the man of the canals, the ditches and the furrows, The farmer, what advantage has he over me? Let him give me his black garment, In return I will give him, the farmer, my black ewe; Let him give me his white garment, In return I will give him, the farmer, my white ewe; Let him pour for me his finest beer, In return I will pour for him, the farmer, my yellow milk; Let him pour for me his sweet beer, In return I will set before the farmer my curdled milk… After I have eaten, after I have drunk, I will leave for him the extra fat, I will leave for him the extra milk: The farmer, what advantage has he over me? Although Inanna ultimately chooses Dumuzi to marry, a moral point of the fable is the further reconciliation of the rivals, having Enkimdu the farmer also bringing his gifts to the goddess. Therefore, not only this fable indicates that there was indeed a clear divergence of interests between the stakeholders of these activities, but also that it was a significant social problem to be coped with. Being it true for Southern Mesopotamia, as early as the third millennium B.C., there is no reason not to expect it in other latter contexts of arid Eurasia, many of which entail cultural differences forming prior to proper interaction.

Excluding the waterlogged alluvium, we assume that the area that can be used to grow farms, including both the natural oasis and the adjacent (potentially irrigable) plains, are also interesting for stationing herds during a given season. Consequently, in order to both farming and herding to use the same patch of land, coordination is needed, specifically to adjust the timing and intensity of each activity. For instance, it would be necessary for herds to leave the fields during the growth of crops, while fields should be already harvested or left in fallow if animals where to graze upon them.

Probably the main cleavage between farming and herding is the one regarding the use of land. It is suggestive that the goddess Inanna embodied –among other things— both

4

E x pl o r i n g t h e O a s e s o f C e n t r a l As i a All the work we developed rest on the assumption that there is no a priori coordination between farming and herding, such that they can coexist in the same area within an oasis at exactly the same time. Even if both activities are carried out in the same social fabric, with no clear–cut distinction between farmers and herders, we expect to observe some incompatibility between their simultaneous materializations in the context of oases. Another way of understanding this assumption is that, if and when cultivated fields are openly available for grazing, the herds feeding on them would not actually change its land use (i.e. farmlands would remain as such).

Although choosing the aspects to be incorporated in a model is inevitably a very explorative and discursive procedure, to fit them together in a single coherent system binds us to better specify definitions, to make explicit most underlying intuitions, to sacrifice most of the complexity observed in reality and, sometimes, to acknowledge contra–intuitive implications of the mechanism represented. The present article collects the last and most mature version produced during this arduous process, though it should be still considered work in progress. To the date, we have developed and explored only the first “core version” in our approach, an Agent– Based model called Musical Chairs, which will be briefly explained in the next section.

The lack of coordination between farming and herding could be simply the by–product of social distancing and a general disregard for livelihoods different from oneself. However, given that both interests would be favored by increasing coordination, it is probably the case that people deciding on how to carry out these activities are just not good enough predictors. A great deal of unpredictable events can prevent a fully coordinated situation, especially if they happen when people engaged in farming and herding are not fluently communicated. Different crops have often different growing seasons and the ripening term can fluctuate, depending on unpredictable environmental factors; herds’ numbers can vary drastically between good and bad years and the spell of extreme climatic events (e.g. droughts, blizzards) can force herds to suddenly change their pace and route; fields can be extended to supply new demands or allocate new settlers; and, at any rate, socio–political events in the local and regional sphere will constantly affect in a rather complex manner the practice of both farming and herding.

Before properly entering the issue at hand, some important notes are made below, regarding what is a model and what aspects of our model are developed here. A model is a representation of a phenomenon, so the miniature of a railroad and Newton’s laws of mechanics are both perfectly valid models. Formally, models can be defined by the combination of relationships and the assumptions on which they depend on, which as a whole should represent a phenomenon. This is, of course, easier said than done, particularly regarding computational modeling, as several examples in the literature attest to how complicated and self–centered models can turn out to be. The most extended approach to explain computational models is by far the one focusing on the description of the elements and their relationships, as they are formally implemented in computer code (e.g. equations, objects, classes). Thus, the preoccupation is often to show the validity of the model as being computationally consistent. However, few authors put effort into exposing and evaluating the assumptions on which elements and relationships depend, so that the implications of the model are truly appreciated in the light of the conditions to be met. Any model can be valid if its assumptions are confirmed in at least one case; however, few models—if any—can be said to represent a phenomenon under any conditions.

Consequently, our research question narrows down to what mechanisms–if any—solves the reiterated dilemma between farming and herding regarding land use (i.e. which is to be carried out in a given patch of land), leaving aside for the time being the possibility of perfecting coordination. Moreover, we will be interest in understanding the implications of the proposed mechanisms for the configuration of different land use patterns among oases. Our approach In the framework of the SimulPast project, financed by the Spanish government, groups of different background came together to unravel the processes involved in archaeologically–inspired case studies. One of such case studies, centered in the interaction of sedentary and nomadic population of oases in Central Asia, shaped the work published here and elsewhere (Angourakis et al. 2014). We followed a bottom–up approach to address this question in the theoretical arena, meaning that we first reduce the problem to some elemental aspects (the model), and then increasingly add complexity to this core (versioning the model), while contrasting the implications of each version against archaeological and historical cases.

This other side of models is of great importance in science, being that it holds the meaning of the model itself and its relevance as a tool to facilitate knowledge in a specific field. It is by contrasting its assumptions with the context of real cases that the value of a scientific model can be accessed. Furthermore, once assumptions are made clear, it is much easier to modify the model by relaxing or removing them, or by adding entirely new ones. In this sense, a computational model with no logical structure other than the definitions found in computer code is bound to be only this, a computational model; though, to do them justice, such models can be very useful as templates to develop new models. This is the case, for instance, of a model of bouncing balls without any specification on the conditions in which these bounce (e.g. gravity, elasticity of materials, attrition).

5

ANDREAS ANGOURAKIS will adopt the properties of a particular land use variant, including its class (i.e. farming or herding). Land use changes whenever a player takes a free chair or steals one previously occupied by another.

The Musical Chairs model Overview Since a full description of our model can be consulted elsewhere (Angourakis et al. 2014), together with a detailed discussion of its results, this paper delves especially into its assumptions and the corresponding implications. Nonetheless, it is obligatory to introduce some brief notions on the elements and relationships of the model, and how they play out in it. Briefly said, the Musical Chairs model consists of two populations competing for positions inside a limited two– dimensional space. Its name comes from the classical children’s game, in which players move around a group of chairs accompanied by music and, once music stops, must find a free chair to sit. The difficulty of this game is that, each time the music starts, one chair is removed and so one player, unable to sit in the next turn, must leave the game. Despite similarities, our model differs from it in four essential aspects:

Finally, the two classes of players behave according to different rules. Players representing variants of herding land use must stand and move each time music is playing (i.e. herds need to leave seasonally), while those corresponding to farming land use stay in their chairs. While the music plays (i.e. when herds are elsewhere), new players, representing the increasing demand of farming land use, may take chairs previously used by players representing herding land use. Once music stops (i.e. herds arrive), all players representing the variants of herding land use, both old and new, must find a free chair in order to keep playing. At this point, if all chairs end up taken, each player still standing may try to displace a player of the other class, posing a dilemma to be solved by the game: which variant of land use will be realized and which will disappear (Fig. 1).

1. Players belong not to one, but to two different classes and they cannot take chairs from players of their own class. 2. The players of one class stay seated when music is playing, while those of the other can force them out once music stops. 3. Instead of having fewer chairs every turn, the challenge is determined by new players constantly entering the game. Consequently the game never ends. 4. Standing players can choose to leave the game if conditions are deemed unfavorable for them to stay.

Implementation and simulation of the Musical Chairs model were done using Netlogo (Wilensky 1999). The logical structure As remarked above, the logical structure behind the design of a model must be addressed in order to apply it in disciplinary fields other than Computer Science. In short, the logical structure of a model is the map of dependencies between the assumptions that must be true for the model to be valid. It is actually the outcome of reverse engineering a model, going from its highly simplified mechanism back to the observations that inspired it. Such a map must capture how more fundamental assumptions converge in more complex assumptions, utterly having all assumptions implied in the

While the chairs of our model represent land units, players are potential (when standing) and effective (when sitting) land use states of these. Thus, when a player successfully occupies a chair, it means that a land unit

Fig. 1: The phases of the model.

6

E x pl o r i n g t h e O a s e s o f C e n t r a l As i a mechanism described by the model. As any kind of reasoning, this structure can be always mounted by even more fundamental assumptions; e.g. here we assume that both farming and herding exist as human livelihoods, and so the model cannot be applied to the Paleolithic. The decision on where to stop splitting hairs is of course arbitrary, depending on pragmatism and plain common sense.

exclusive regarding land use, at least during one phase of the cyclic movement of herds. 3. On the categorization of land use. The use of a land unit can be approximated to either farming or herding activities. Such approximation does not imply that other activities cannot occur simultaneously in a given patch; it only means that the one stressed is the most widespread within the land unit. This assumption depends on assumptions 1 and 2. 4. On the mobility implied in land use classes. Farming is a sedentary activity and herding is a mobile activity. This assumption depends on assumption 3. 5. On the land tenancy regime implied in land use classes. Farms are private and pastures are communal. Even though farms can also be partially or fully considered communal, such form of commons is mostly nominative: i.e. while property lies with higher community authorities (e.g. villages, clans, temples, aristocratic elites and monarchs), the particular persons (and their families) are those responsible of exploiting it. On the other hand, a proper exception to this assumption is the existence of private pastures, which seems to abound especially when farming reaches a regional peak of development, and pastures are scarce in the proximity of settlements. As this phenomenon concurs also with substantial fodder cultivation and commercial livestock breeding, one straightforward interpretation is that pasture privatization is closely related to a general growth of market economy along the main trade lanes. For sake of simplicity, we consider livestock enclosures and fodder cultivation to be different forms of farming land use. This assumption depends on assumption 3. 6. On the variation of land use. Land use varies in intensity and independence of the land use of the alternative activity. Intensity stands for the concentration of productive factors—including people, livestock and crops—while independence expresses how much the local activity depends on productive factors also engaged in the alternative activity. For instance, a meadow punctuated with small farmsteads engaged in herding corresponds to a variant of herding land use with medium to low intensity and independence. Correspondingly, the pastures used to graze a king’s flock will have a high score of intensity (i.e., it involves many herders and animals) but probably a relatively low independence of farming (i.e., a king may also rely on the tribute of local farmers). This assumption depends on assumption 3. 7. On the integration of land use classes. The whole of the land used for the same kind of activity (i.e. all farmlands, all pastures) has a particular degree of integration, accordingly to which the productive factors involved in each patch depend on each other. Dependency between variants of the same class of land use is a proxy of the strength of the social fabric shared by people relying on the same livelihood. Such degree of integration can be given by numerous factors, ranging from plain logistic requirements, such as the cooperation among farmers for the maintenance of dams and canals (e.g., Mukhamedjanov 1994), to

Fig 2: The dependencies between the assumptions of the model, as referenced in the text. Arrows point to the dependent assumption.

The list bellow enumerates all 16 assumptions deemed as necessary to the full understanding of the model’s implications. Even though some ordering of assumptions was performed from the more fundamental to the more complex, numbers may be regarded only as references to Fig. 2. 1. On the extension of food–producing activities. Farming and herding involve far larger extents of land than any other field of activity, such as those concerning hunting, fishing, gathering, storing, manufacturing, trading or mining. Housing, stabling and the various types of farming facilities are accounted as inherent components of the respective activities. 2. On the exclusiveness between food–producing activities. Farming and herding activities are mutually

7

ANDREAS ANGOURAKIS complex group identity reinforcers, such as tribal obligations and ceremonial reunions (e.g., Barth 1964). This assumption depends on assumption 3. 8. On the growth of land use demand. There is generalized growth of the economy, implying population growth, wealth accumulation and, last but not least, an increasing demand for land. 9. Ceteris paribus or on the immutability of other constraints. There are constant environmental and technological constraints. The change in either of them would change the area represented in the model and the maximum intensity allowed for each type of land use. Although it seems a rather extreme assumption, following a modeling rationale, it simply states that neither environmental nor technological constraints depend on the outcome of the model—i.e., the land use pattern. Considering that this assumption may be still too unrealistic, further relaxation could be done by partially internalizing these factors, for instance, representing two feedback mechanisms: one where innovations allow further intensification and a higher intensity increases the chance of developing further innovations; another where intensification decreases productivity on the long–run, through salinization or overgrazing, and so restricts intensity itself. 10. On the welfare imperative. The production generated by the respective activities is the main criterion of decision–makers to promote or assign a given land use. 11. On the extensive imperative. Given the assumption 9, increasing the intensity of activity will increase the production per land unit, but will reduce the productivity per person or resource invested. Therefore, decision–makers, if producers or representative of producers (e.g. tribal chiefs, elder councils), will prefer to extend their activity rather than intensify it. A direct implication is that these decision–makers will restraint growth (i.e. redirecting people and resources to other activities), if extending their land use is not possible. Even if such decisions are made at the household level, we assume that people prefer to migrate or even change their livelihood rather than to lead a poorer lifestyle. This not implies that high–level decision–makers (e.g. aristocratic elite) would not perceive intensification as interesting, as it does increase the aggregate production and thus their own wealth. However, this latter possibility was not explored in the present model. This assumption depends on assumptions 9 and 10. 12. On the impossibility of competition within land use classes. The variants of a class of land use do not compete for the same patch of land. Although particular individuals, families, groups or organizations engaged in the same activities may compete for land, we assume that the land use pattern –under the terms we express it— does not depend on the outcome of this dynamics. The direct implication of this assumption (and those on which it depends) is that any trend observed in land use pattern should

respond to the dynamics of farming and herding competing for land against each other. Note that this assumption must be discarded if, against assumption 11, we assume that land use change is promoted by decision–makers interested in intensification. If this were the case, less intensive land use variants would be replaced by more intense variants of the same class, without any intervention of the alternative class of land use. This assumption depends on assumptions 3 and 11. 13. On the possibility of land use change. The main land use within a land unit can change from farming to herding, and vice versa, within the passing of the cyclic movement of herds. This assumption holds the possibility that land formerly used by one activity can be used by the other, without any interim period of abandonment longer than a cycle. It can be more easily accepted if considered that, while land is in fact used, it may not be initially as productive as those with well–established infrastructures (e.g., irrigation systems may take time to function properly). This assumption depends on assumption 3. 14. On the existence of land use dilemmas. Economic growth beyond the available land units poses the dilemma of whether a land unit will be used mainly for farming or herding. This assumption depends on assumption 8, 12 and 13. 15. On the factors involved in land use change. The shift of main land use in a given land unit is caused by a complex chain of events, approachable as a stochastic process (i.e., involving random variables). However, it also significantly depends on four factors acting in three levels: I. Extension. What opportunities there are of further extending a class of land use? The less saturated land use class is favored. II. Independence. How much competing variants are independent of those of the alternative class? Given that there are sufficient opportunities (level I), the variants with greater independence are favored. III. Ratio of intensities. Given that the two competing variants are sufficiently independent (level II), what land use has the greater number of productive factors involved? It breaks down into two factors: i. Intensity. How intense are the competing variants of land use? The variants with greater intensity are favored. ii. Integration. How well each class of land use is integrated? The most integrated class is favored. This assumption depends on assumption 6, 7 and 14. 16. On the mechanism of land use change. Farming and herding activities have different constraints regarding the use of land. To extend farming, new immovable facilities (e.g. fields, canals, stores, barns, houses) need to be set when herds are elsewhere, so deciding whether or not to actually set them depends on the presence/absence of herding (extension), not in the specific characteristics of herds and herders (ratio of intensities). In contrast, herds will need the local pastures only once farming activities are already

8

E x pl o r i n g t h e O a s e s o f C e n t r a l As i a underway; hence the decisions taken by herding stakeholders (e.g. of whether or not to maintain the route or the current size of the flock) are made with a good estimation of the whole situation. This assumption depends on assumption 4, 5 and 15.

difference between overall intensities is relatively small (up to a ratio of one–against–two favoring herding). To compete or to cooperate: the selection of independence. Any restraint during a competitive situation will be relevant only to the extent that farming is somewhat dependent on herding, and vice versa. Therefore, a relatively independent variant will have a greater chance of persisting and extending itself under any conditions. However, the persistence of a variant is checked on the long–run by the resolution of dilemma events, which in turn depend solely on the aggregated intensity backing up the variants involved. Overall, independence has two opposing implications: it increases the chances of a class of land use persisting to the detriment of the other, but it also exposes new investments in land use to the risk of future unfavorable land use changes.

Considering the number and the interdependencies of the assumptions made, their implications could only be assessed by implementing and simulating a computational model. Again, both design details and simulation results of our model are presented elsewhere (Angourakis et al. 2014). Implications Leviathan: intensity and integration. As represented in the Musical Chairs model, integration is a costless relationship of reciprocity between unequal elements, so all land use variants will always have a better chance of persisting when associated with other variants within its class, whatever their intensity. The aggregated intensity will always increase with the number of associated variants. Whenever there is some integration among variants of each class, the change in intensity of productive factors in a given time is neither factor nor effect of the existing land use pattern (i.e., no intensification is causing or caused by land use competition). If we assume that the concentration of productive factors is correlated to the frequency of archaeological remains, this implication yields an archaeological parallel: the inequality of land use intensities within an integrated territory correspond to an observable inequality of site presence and size among land units at a local scale.

Attractors The Musical Chairs model is a dynamical system—i.e., it evolves with time— and as such may converge towards particular states, called attractors. Attractors are by definition stable, meaning that small punctual or noise– like perturbations will only temporally move a system away from it. Although a single system can have several kinds of attractors, the Musical Chairs model displays only one: the stationary point. Stationary points are steady states of a system, meaning that all variables are maintained in—or very near—particular values. This feature allows a very straight–forward exploration, since our interest will rely only on comparing conditions to states, and not on the system’s trajectories towards such states.

Bidding for land use: overall intensity. Even though the variance of intensity among land use variants is not causally related to land use pattern, an inequality between the overall intensities of farming and herding will strongly condition the land use pattern at equilibrium (see Attractors). Similarly to the socioeconomic status of buyers during bidding, the overall intensity of each class of land use dictates the probability that a farming variant overcomes a herding variant, and vice versa. In this sense, the activity with the greatest overall intensity will have more chances of extending itself at the expense of the other. Fortune favors the bold: timing and information bias. As stated in assumption 16, the decisions on extending farming and herding are made at different times and with different information. Farming is extended independently from the intensity and integration of herding, assuming that herding is on average as intense and integrated as farming. Therefore people pressing for extending farming are blind to the true nature of the situation; they are unable to perceive the stable land use pattern (i.e., attractor) and may insist even when conditions are unfavorable to farming. While this aspect entail a higher risk to farming stakeholders—e.g., immovable investments may be utterly lost—, it will in fact facilitate the extension of farming on the long–run, whenever the

Fig 3: The percentage of farming at equilibrium versus the ratio between overall intensities of land use. Each point represents data from a simulation with randomized parameters (see randomized experiments in Angourakis et al. 2014). The line

9

ANDREAS ANGOURAKIS technology. However, if assumed that farming and herding are equally feasible—and so have similar overall intensities—, the model predicts oases to be more often ‘big’ than ‘small’ (Fig. 3).

and grey area represent a nonlinear regression curve (GAM method) and its standard error, respectively.

Another important property of dynamical systems is how consistent attractors are throughout variations of conditions (i.e., setting of parameters). If an attractor remains qualitatively the same when conditions change, we may call it structurally stable. However, the Musical Chairs model presents qualitative differences between attractors, depending on the conditions. We summarized this diversity in three classes of attractor, separated by range of land use pattern, i.e., the proportion of land units used for farming (Fig. 3, 4, 5 and 6).

Overcoming the farming advantage. Given the systematic bias backing farming land use, change favoring herding must be caused by three kinds of transformation: 1. Greater herding intensity, lower farming intensity or both. Such effects would be achieved, for instance, by soil salinization, which causes farming to be less productive while herding is unaffected. 2. More herding integration, less farming integration or both. Integration is a proxy of the strength of the social institutions pressing for each activity. Therefore, in order to have unbalanced conditions, historical contingencies (i.e., conquests, political upheavals, migrations, etc.) should be affecting farming and herding stakeholders in a different manner. A recurrent pattern found in Eurasian history is to have land use change following the rise and decay of broader political powers, depending on their particular agenda. A clear example of this is how expanding empires—such as the Mongol (Hall 1991)—conquered some flourishing farming territories, by allying themselves with herding tribes, which then could impose their own interests regarding land use. 3. More herding independence, less farming independence or both. Change in dependencies among farming and herding must also be observable as changes in local politics and socio–economical relationships. For instance, an increase of herding independence could be reflecting the replacement of local agro–pastoral stakeholders by nomadic alliances engaged solely in pastoralism.

Big oasis. Big oases are attractors in which farming is carry out on most of the land units (> 80%), implying that it is an artificially augmented oasis. Big oases display the lowest counts of land use dilemmas and changes per year and have the least selective pressure over the independence of land use variants. Accordingly to this model, such a state is expected when farming is generally more intense and integrated than herding. Furthermore, this state can only be truly an attractor if the external pressure (e.g., regional land use demand) is kept at low levels, otherwise the system may shift into another class of attractor. Small oasis. Conversely, small oases are those attractors in which the major land use is herding and farming is a marginal activity (< 20%). Such scenarios would correspond to small riverine farming plots surrounded by the steppe–grassland landscape. Small oases correspond also to relatively low frequencies of land use dilemmas and changes, while the selection of independence between land use variants is the strongest. Similarly to big oases, this state is expected when the major land use is generally more intense and integrated than the alternative, though in this case herding is the one to be widespread. Also in the case of big oases, true stability do not exists if the external pressure is too great. Intermediate oasis. Finally, intermediate oases are attractors in which there is no clearly predominant land use (20–80%). Intermediate oases present high frequencies of land use dilemmas and changes, and an intermediate strength of selection over independence. Such states tend to be stable—i.e., be an attractor— when farming and herding are roughly similar in intensity and integration. However, unlikely the other two classes of attractor, intermediate oases seem to be sustained by relatively high levels of external pressures. Discussion Insights The model yields four expectations regarding real “oases”: Inherent viability and systematic bias. A land use is more likely to spread around an oasis when it is favored by the plurality of factors affecting its overall intensity, from soil characteristics to labor organization and available

Fig 4: The percentage of farming at equilibrium versus the overall external pressure (i.e., the extrinsic land use demand). Each point represents data from a simulation with randomized

10

E x pl o r i n g t h e O a s e s o f C e n t r a l As i a To assess the value of a model it is not enough to identify its implications and attractors. Models results must also be read and evaluated in the light of real case scenarios. Consider the complexity of land use pattern trajectories throughout the arid Eurasia (Stride 2005, in Southern Uzbekistan; Nesbitt and O’Hara 2000, in Southern Turkmenistan; Alizadeh and Ur 2007, in North–Western Iran; Abdi 2003, in Central Zagros; Newson 2000, in South–Eastern Syria). Each territory has specific contingencies, both cultural and environmental, which also may depend on contingencies of other territories. Still, oases may display some important common trends, for they clearly have some key common properties, such as being seasonal natural heavens for virtually any life form. As middle ground between relativism and universalism, a synthesis is possible by combining a geography and a history of land use in Eurasian oases.

parameters (see randomized experiments in Angourakis et al. 2014).

Sensibility to external pressure. Oases under great external pressure will tend to display intermediate land use pattern, whenever other conditions do not change (Fig. 4). Though enduring for centuries, they may be interpreted as transitional states on the long run, since it is a relatively unstable state (Fig. 5) and external pressure and other conditions are expected to change. The pastoral ‘character’. The prevalence of herding in oases is positively correlated to the benefits of independence between farming and herding (Fig. 6). Accordingly to this model, ‘small’ oases are as farming islands within an ocean of autonomous pastoralism, so there would be little or no involvement of farmers in herding and of herders in farming. On the other hand, ‘big’ oases are characterized by the presence of a herding minority that may be aligned with local farming. However, it should be noted that independence—as it is conceptualized here—is not constrained by any other aspect apart from those considered in this model (i.e., there is no upper or lower limit to independence). If such scenarios are to be interpreted on the light of real cases, we must assume that all known forms of dependence between herding and farming—e.g., nutritional (Khazanov 1994)—are actually surmountable at a higher scale. For instance, a pastoral group camping in an oasis could rely on staple food produced in a neighboring fully– sedentary territory, thus having no restraint in pressing against local farming.

Fig 6: The percentage of farming at equilibrium versus the correspondent mean independence of variants (0 equals no independence). Each point represents data from a simulation with randomized parameters (see randomized experiments in Angourakis et al. 2014).

From attractors to complex trajectories of land use pattern As a geography, a very clear spatial pattern can be identified: there are predominantly farming regions and there are predominantly herding regions (Adas 2001). There can be some problems defining a threshold between these two categories, especially if the units considered are too big (e.g., Mesopotamia); however, for most of arid Eurasia, such distinctions can be easily done (e.g., compare Sumer with Bactria). To explain that this geography is consistent throughout the different historical contexts, it must be conditioned by roughly constant environmental factors. Though aridity and temperature are sound candidates for explaining this pattern (Bendrey 2011, Bonte 1981), simple topology could also have an important role (e.g., the relative position of oases; Algaze 2005). The bimodal nature of our results—the fact that ‘big’ and ‘small’ oases are more frequent than

Fig 5: The percentage of farming at equilibrium versus the correspondent count of land use dilemma events. Each point represents data from a simulation with randomized parameters (see randomized experiments in Angourakis et al. 2014). The line and grey area represent a nonlinear regression curve (GAM method) and its standard error, respectively.

11

ANDREAS ANGOURAKIS intermediate settings— is in accordance with this broad spatial trend. Furthermore, also agreeing with the common geographical explanation, our model suggests that the state of an oasis strongly depends on the overall intensity levels of herding and farming, which in turn strongly depend on steady environmental factors.

There are numerous cases of this pattern also in later historical periods, such as the breakdown of the Greco– Bactrian Kingdom in the second century B.C. (Stride 2005, Enoki et al. 1994), the decline of Khwarizm cities and irrigation systems between the fourth and the sixth centuries A.D. (Nerazik 1996), the abandonment of Sasanian agricultural settlements of northwestern Iran between the seventh and the ninth centuries A.D. (Alizadeh and Ur 2007), and the drop of sedentary agriculture in the wadis of the southern Levant in the dawn of islamization (Haiman 1995, Hill 2004, Parker 1987).

Despite this relatively stable geographical pattern, a somewhat general pattern can also be observed diachronically: a long–term tidal wave pattern, in which expanding farmlands alternate with expanding pastures, given an overall predominance of one or another land use. Throughout the History and late Prehistory of the arid Eurasia, this pattern repeatedly occurs at different scales, from the local land use pattern of oases up to continental cultural borders. It starts with the protracted—but relatively unobstructed—process of farming growth, covering up to three thousand years of paced development. Once farming land use reaches its techno– environmental maximum, stakeholders involved are increasingly incentivized to intensify, pressed by the inertia of former demographic and economic growth.

The former examples can be illuminating as well regarding what may cause individual oases to vary from this general temporal pattern. The general geography sketched above—and so, its dependence on environmental factors—should be placed as an important element to explain such variation. For instance, farming growth will hardly happen in an oasis surrounded by very productive spring/summer pastures, hence no oscillation can occur. Nevertheless, there are some human–related contingencies that cannot be reduced to environmental causation, even though may be connected to it.

As many modern governments are now coming to realize, intensification in arid environments is not sustainable in the long run. Particularly at the farthest ranges of alluvium, uninterrupted cultivation can overwhelm the capacity of soils to regenerate, turning salty much of previously productive farmlands (Bucknall 2003, Geist and Lambin 2004, Geist 2005, Laity 2009, Zhao et al. 2005). However, salty steppe and deserts can be still used for seasonal animal grazing. In fact, there are strong botanical (Kapustina et al. 2001) and archaeological (Abdi 2003, Kohl 2007) arguments supporting that these bumps in farming growth were also booms of herding. Furthermore, since more livestock could be stationed near oases during the less productive months—i.e., typically winter—, more and farther pastures needed to be exploited during the more productive ones, progressively increasing herders mobility.

It seems evident that migrations play a very important role in land use change. In fact, the periodic decay or collapse of local sedentary settlements is often interpreted within the context of incoming migrations, particularly of people with more mobile lifestyles (e.g., Early–Semitic, Indo–Iranians, Scythians, Arabs, Turks, etc.). Conversely, farming seems to prosper and growth whenever oases frequently receives new settlers, e.g., when it is well communicated with other oases. However, migrations alone might not change a relatively stable land use pattern. For land use change to happen, the institutional framework in which land use decisions are made should also change. Again, there are much of the previous examples that suggest that changes in the socio– political context (e.g., the decadence of local political elite, a sudden subjugation to a foreign power, etc.) may similarly trigger changes favoring one or another land use. Among them, there are very illustrative examples of how this can be determinant, such as the use of the Mughan Steppe as winter quarters by Mongol Ilkhans and Timurids (Alizadeh and Ur 2007), the roman military policing the movements of nomads in the Transjordan frontier (Parker 1987) or the forced sedentarization of Central Asian nomads under Soviet rule (Luong 2004, Rahimon 2012).

In prehistoric and proto–historical times, this general pattern can be illustrated by the peak–and–collapse pattern of Late Chalcolithic/Bronze Age cultures of western arid Eurasia. Within this comprehensive period, roughly ranging from the forth to the second millennium B.C., there are several cases of waning and abandonment of prosperous urban or proto–urban settlements. Outstanding examples are the Cucuteni–Tripol’ye giant settlements in Southeastern Europe (Kohl 2007), the Uruk–type in Mesopotamia (Algaze 1993), the Bactria– Margiana Archaeological Complex in Southern Central Asia (Kohl 2007, Masson 1992), and the Harappan culture in the Indus Valley (Dani and Thapar 1992). Overall, the collapse of these cultures entailed less permanent settlements, more seasonal/ephemeral sites, or simply an escalation of settlement diversification at both local and regional scales (Abdi 2003, Nesbitt and O’Hara 2000, Potts 1999), all pointing towards greater commitment of the population with herding.

Concerning temporal variations, the complexity of real land use change is still outside the scope of our results. In the Musical Chairs model, the factors that affect land use change are conceptualized as parameters (overall intensities, intrinsic and extrinsic growths of land use demand, integration), and so the mechanisms changing them are not modeled. There is no explicit reference to the environmental, economic or socio–political context in which an oasis is situated. In this sense, as presented in

12

E x pl o r i n g t h e O a s e s o f C e n t r a l As i a assumption 9, we have chosen not to explore what may be the backbone of land use change in oases on larger temporal scales, namely the loss or gain of potential farmland (e.g., by salinization and irrigation, respectively).

For instance, it could specify how slope, distance to the nearest waterbody and soil erosion limit the intensity of farming. Within these lines, by specifying minimums for each environmental factor, it could be possible to also represent variation on the total number of land units available for farming. Secondly, variation of the intensity of specific land use variants could be implemented as the product of decisions aiming to sustain or maximize returns. Additionally, the relationships between variants could also play a role in these decisions, by rewarding farming–herding coordination; e.g., herders could sustain more animals in winter, if farmers are willing to provide them with fodder or fallow land. Performing these changes would require us to reform assumptions 6 and 15, but also to revisit assumptions 9 and 11.

This model allows us to explore towards what state, given specific conditions, may the competition for land use drive an oasis, and how changes in such conditions may alter a this state. However, it cannot illuminate anything on why and how these conditions may change. Providing there is no mechanism describing feedbacks between land use and those conditions, the Musical Chairs model displays only point attractors, whilst real land use trajectories may be better represented by oscillations, such as the tidal wave pattern.

Relational properties. Integration and independence are modeled as two constants and qualitative–distinct aspects—as two parameters and a variant trait, respectively. However, conceptually, they are representative of the same phenomenon, namely the degree of convergence between stakeholders involved in two different land use variants, whatever their activity. Also, during competitive situations, there is no reason to assume that stakeholders will be a priori aligned with only one land use. In fact, people may press for extending a given land use based on completely different terms (e.g., kinship, ethnicity, territoriality, political agenda, market prices, etc.). This could be corrected by explicitly representing bonds between variants, and by allowing farming to support herding and vice versa. From exploring models with these characteristics, it would be possible to better represent the role of group–level interests as drivers of land use change. Again, reforming assumption 6 and 15 would be needed.

Open routes Without taking away its virtues, there are several aspects in which the model misfits what can be observed in real cases, and so where it could be improved: Boundary constraint. In the Musical Chairs model, all land units are considered to be equally in the range of each other’s influence. We are assuming the dynamics to occur in a local setting, where people can circulate in a daily basis and so are indifferent to the specific location of their activities. However, real oases show that land use grows contiguously; for instance, a pasture adjacent to farmlands could easily become a farmland, contrasting with pastures that are surrounded by other pastures. This would modify assumption 15, by adding further constraint to land use change. Variable land units and land use variants. In order to represent more complex dynamics of land use change, both the maximum intensity per land use and the actual intensity of land units could be set as variable aspects of the model. Firstly, spatiotemporal variation of the maximum intensity would be possible by introducing a submodel to connect environmental factors and land use.

13

ANDREAS ANGOURAKIS Bibliography ABDI, K. 2003, “The Early Development of Pastoralism in of civilizations of Central Asia, Volume I, The dawn of the Central Zagros Mountains”, Journal of World civilization: earliest times to 700 B.C,. Paris, 283–318. Prehistory 17(4), 395–448. ENOKI, K.; KOSHELENKO, G. A.; HAIDARY, Z. 1994, “The ADAS, M. 2001, Agricultural and pastoral societies in Yüeh–chih and their migrations”, in J. HARMATTA (ed.), ancient and classical history, Philadelphia. History of civilizations of Central Asia, Volume II, The development of sedentary and nomadic civilizations: 700 ALGAZE, G. 1993, The Uruk World System, Chicago. B.C. to A.D. 250, Paris, 171–189. – 2005, “The Sumerian Takeoff”, Structure and Dynamics 1(1), 2.

GEIST, H. 2005, The Causes and Progression of Desertification, Burlington, Ashgate.

ALIZADEH, K.; UR, J. A. 2007, “Formation and Destruction of Pastoral and Irrigation Landscapes on the Mughan Steppe, North–Western Iran”, Antiquity 81(311), 148–160.

GEIST, H.; LAMBIN, E. F. 2004, “Dynamic Causal Patterns of Desertification”, BioScience 54 (9), 817–829. HAIMAN, M. 1995, “Agriculture and Nomad–State Relations in the Negev Desert in the Byzantine and Early Islamic Periods”, Bulletin of the American Schools of Oriental Research 297, 29–53.

ANGOURAKIS, A.; RONDELLI, B.; STRIDE, S.; RUBIO– CAMPANILLO, X.; BALBO, A. L.; TORRANO, A.; MARTÍNEZ, V.; MADELLA, M.; GURT, J. M. 2014, “Land Use Patterns in Central Asia. Step 1: The Musical Chairs Model”, Journal of Archaeological Method and Theory 21, 405–425: http://dx.doi.org/10.1007/s10816–013– 9197–0.

HALL, T. D. 1991, “Civilizational change: The role of nomads”, Comparative Civilization Review 24, 34–57. HILL, J. B. 2004, “Land Use and an Archaeological Perspective on Socio–Natural Studies in the Wadi Al– Hasa, West–Central Jordan”, American Antiquity 69 (3), 389–412.

BARNARD, H.; WENDRICH, W. 2008, The archaeology of mobility, Cotsen Institute of Archaeology, University of California.

KAPUSTINA, L. A.; TORRELL, M.; VALLÈS, J. 2001, “Artemisia Communities in Arid Zones of Uzbekistan (Central Asia)”, in E. D. MCARTHUR & D.J. FAIRBANKS (ed.), Shrubland ecosystem genetics and biodiversity: proceedings; 2000 June 13–15; Provo, UT. Proc. RMRS– P–21, Ogden, 104–112.

BARTH, F. 1964, Nomads of South Persia. The Basseri Tribe of the Khamseh Confederacy, Oslo. BENDREY, R. 2011, “Some like it hot: environmental determinism and the pastoral economies of the later prehistoric Eurasian steppe”, Pastoralism: Research, Policy and Practice 1, 8.

KHAZANOV, A. M. 1994, Nomads and the Outside World. Madison.

BONTE, P. 1981, “Ecological and Economical Factors in the Determination of Pastoral Specialization”, in J. G. GALAXY & C. SALZMAN (ed.), Change and Development in Nomadic and Pastoral Societies, Leiden, 33–49.

KOHL, P. L. 2007, The Making of Bronze Eurasia, Cambridge. LAITY, J. J. 2009, Deserts and desert environments, Hoboken.

BUCKNALL, J.; KLYTCHNIKOVA, I.; LAMPIETTI, J.; LUNDELL, M.; SCATASTA, M;. THURMAN, M. 2003, Irrigation in Central Asia. Social, Economic and Environmental Considerations. Europe and Central Asia Region—Environmentally and Socially Sustainable Development, report, World Bank.

LUONG, P. J. 2004, The transformation of Central Asia: States and societies from Soviet rule to independence, Cornell. MARTIN, L. 2000, “Gazelle (Gazella spp.) behavioral ecology: predicting animal behaviour for prehistoric environments in south–west Asia”, Journal of Zoology 250 (1),13–30.

CHULUUN, T.; OJIMA, D. 2002, “Land use change and carbon cycle in arid and semi–arid lands of East and Central Asia”, Science in China series C Life Sciences– English Edition 45(SUPP), 48–54.

MASSON, V. M. 1992, “The Bronze Age in Khorasan and Transoxiana”, in A. H. DANI & V. M. MASSON (ed.), History of civilizations of Central Asia, Volume I, The dawn of civilization: earliest times to 700 B.C., Paris, 225–245.

CRIBB, R. 2004, Nomads in archaeology, Cambridge. DANI, A.H.; THAPAR, B.K. 1992, “The Indus Civilization”, in A.H. Dani, V.M. Masson (ed.), History

14

E x pl o r i n g t h e O a s e s o f C e n t r a l As i a MOSCATI, S. 1960, The Face of the Ancient Orient. Near Eastern Civilization in Pre–Classical Times, London.

PRITCHARD, J. B. 1955, The Ancient Near East: An Anthology of Texts and Pictures. Princeton.

MUKHAMEDJANOV, A. R. 1994, “Economy and social system in Central Asia in the Kushan Age”, in J. HARMATTA (ed.), History of civilizations of Central Asia, Volume II, The development of sedentary and nomadic civilizations: 700 B.C. to A.D. 250, Paris, 265–290.

RAHIMON, R. M. 2012, “Evolution of Land Use in Pastoral Culture in Central Asia with Special Reference to Kyrgyzstan and Kazakhstan”, in V. SQUIRES (ed.), Rangelands Stewardship in Central Asia: Balancing Improved Livelihoods, Biodiversity Conservation and Land Protection, Springer, 51–67.

NERAZIK, E. E.; BULGAKOV, P. G. 1996, “Khwarizm”, in B. A. Litvinsky (ed.), History of civilizations of Central Asia, Volume III, The crossroads of civilizations:A.D. 250 to 750, Paris207–231.

ROSEN, S. A. 1993, “A Roman–Period Pastoral Tent Camp in the Negev, Israel”, Journal of Field Archaology 20 (4), 441–451.

NESBITT, M.; O’HARA, S. 2000, “Irrigation agriculture in Central Asia: a long–term perspective from Turkmenistan”, in G. BARKER & D. GILBERTSON (eds.), The Archaeology of Drylands. Living at the margin, Oxon, 103–22.

ROSENBERG, M. 1998, “Cheating at Musical Chairs. Territoriality and Sedentism in an Evolutionary Context”, Current Anthropology 39 (5), 653–681. STRIDE, S.; RONDELLI, B.; MANTELLINI, S. 2009, “Canals versus horses: political power in the oasis of Samarkand”, World Archaeology 41 (1), 73–87.

NEWSON, P. 2000, “Differing strategies for water supply and farming in the Syrian Black Desert”, in G. BARKER, D. GILBERTSON (eds.), The Archaeology of Drylands. Living at the margin, Oxon, 86–102.

STRIDE, S. 2005, Géographie archéologique de la province du Surkhan Darya (Ouzbékistan du sud / Bactriane du nord), [Diss], Université Paris I Panthéon– Sorbonne.

OBERHÄNSLI, H.; NOVOTNÁ, K.; PÍŠKOVÁ, A.; CHABRILLAT, S.; NOURGALIEV, D. K.; KURBANIYAZOV, A. K.; GRYGAR, T. M. 2012, “Variability in precipitation, temperature and river runoff in W Central Asia during the past ~2000 yrs”, Global and Planetary Change 76, 95– 104.

WILENSKY, U. 1999, NetLogo. http://ccl.northwestern.edu/netlogo/. Center for Connected Learning and Computer–Based Modeling, Northwestern University. Evanston, IL.

PARKER, T. 1987, “Peasants, Pastoralists, and the ‘Pax ZHAO, W. Z.; XIAO, H. L.; LIU, Z. M.; LI, J. 2005, “Soil Romana’: A Different View”, Bulletin of the American degradation and restoration as affected by land use change Schools of Oriental Research 265, 35–51. in the semiarid Bashang area, northern China”, Catena 59, 173–186. POTTS, D. T. 1999, The Archaeology of Elam. Formation and Transformation of an Ancient Iranian State, Cambridge.

15

Aryan warriors* The Invented Heroes in the Ancient Near East JORDI VIDAL Universitat Autònoma de Barcelona ABSTRACT The aim of this contribution is to analyze the historiographical myth concerning the supposed warrior/heroic aspect of the ancient Aryans. Thus, the study of the Ancient Near Eastern texts clearly demonstrates that those supreme warriors coming from the Central Asia never existed in that region; they just were a literary product of romantic and racist historians, who invented them in order to support their particular recreation of the past. KEYWORDS Semites – Aryans – maryannu – Ugaritic texts – Heroic tales The positive view of violence, or even its veneration as an effective tool to achieve, maintain and/or defend a position of power has been common in certain historical, political and many other classes of discourse throughout history. The holding up of aggressive behaviour and the positive assessment of war are constants that have occurred in an institutionalised manner in some discourses since the emergence of the early states in the Near East until now. 

In this context, concepts such as heroism, pain, death, courage, etc., become the backbones of both ideas and patterns of behaviour, and are often used by totalitarian regimes as a way to legitimise their power. Of course, this culture of violence, in addition to its own doctrinal developments, also has a powerful set of icons and myths that enable rapid dissemination and easy assimilation of the message. In this paper we propose analysing historically one of those myths, whose perverse consequences in contemporary history are more than well known: the Aryans, the legendary warriors that originate in the mountains of Central Asia1 began a historic voyage that many thought (and still consider) as the foundational epic of Western culture.

prevalent is the one that refers to the existence of a superior Aryan race, native to the mountains of Central Asia. Some historians liked to especially highlight these first Indo–Europeans' warrior skills, their worship of weapons, their taste for heroic deeds and their high ideals associated with honour in combat, etc. Of course, often this image of legendary warriors is juxtaposed to the cowardly and effeminate Semitic peoples, absolutely incapable of war and, therefore, a far cry from any possible heroism. One of the more explicit images of Aryan heroism quite opposite to the Semitic military incapacity appears in La Histoire romaine by the French romantic historian Jules Michelet. There Michelet spoke of the Punic Wars as the culminating moment in which Aryan heroes (Rome) were faced against Semite merchants (Carthage) to the point of death: “Ce n’est point sans raison que le souvenir des guerres puniques est resté si populaire et si vif dans la mémoire des hommes. Cette lutte ne devait pas seulement décider du sort de deux villes ou de deux empires; il s’agissait de savoir â laquelle des deux races, indo–germanique ou sémitique, appartiendrait la domination du monde (…) D’un côté, le génie héroïque, celui de l’art et de la législation; de l’autre, l’esprit d’industrie, de navigation, de commerce (…) les héros combattent sans cesse leurs industrieux et perfides voisins. Ceux–ci sont artisans, forgerons, mineurs, enchanteurs. Ils aiment l’or, la sang, le plaisir. Ils élèvent des tours d’une ambition titanique, des jardins aériens, des palais magiques, que l‘épée des guerriers dissipe et efface de la terre”2.

1. Racist myth of the Aryan Warriors The racist and romantic historiography of the 19th century is directly responsible for the creation of multiple historical myths, many of which are still widespread today, especially in popular culture. One of the most 

This paper has been produced in the context of the research Project HAR2011–23572 (Ministerio de Economía y Competitividad), and within the Research Group HICOAN – 2014 SGR1111 of the Generalitat de Catalunya. 1 See Chapoutot 2008, 21ss. for a discussion on the origins of the Indo–Europeans and the Nazi attempt to locate that origin not in Central Asia but in Northern Europe.

And, it is thus, as pointed out by Christian Bunsen, for the Romantic historiography of the 19th century “If the

2

Michelet 1833 [1830], 245f.

JORDI VIDAL Hebrew Semites are the priests of humanity the Helleno– Roman Aryans are, and ever will be, its heroes”3.

heroísmo, la fuerza, la destreza en el manejo de las armas, el honor en el combate frente al enemigo. Y la arqueología del I milenio a.C. en Europa pone de manifiesto ajuares dotados de armas de varios tipos en las tumbas de los príncipes, lo que transluce una sociedad jerarquizada y dominada por una nobleza guerrera”7.

In addition, Otto Schrader in his famous synthesis of Indo–European origins, reflected on this much loved image of some original Aryans as absolute masters of war: “...whether we have to picture the original Indo– Europeans to ourselves spreading as warriors armed with spears, swords, shields, helmets of bronze, bearing the gifts of a higher civilisation, and bringing the non–Aryan peoples with their stone weapons easily to the ground”4.

As seen in the two previous citations by Villar, it is clear that the early Aryans are still described as military heroes, supposedly in order to indicate the ancient epics and certain archaeological finds. However, where those great Aryan warriors actually exist are not found so much in historical sources but rather in the imagination of some historians determined to build the heroic image of their own ancestors in Western civilisation.

However, the historiographical emphasis on highlighting the link between Aryans, war and heroism is in no way an element of the past that is limited to the writings of the 19th century. So, Winn in 1995 continued to insist, albeit with a radically different tone, that Aryan heroism “is reflected in the warlike ideology of Indo–European myth, with heroic warriors and raiders supported by a war/thunder god. Evidence from myth, language and archaeology directly relates the Indo–European warrior tradition to their pastoral lifestyle: as livestock herders, it was imperative to protect their herds or be decimated by other raiders. This led to a cultural emphasis on males and to their higher status. Bravery and heroism were, of course, valued, nevertheless Often the warrior's actions were excessive and improper. As a need was felt to justify aggressive ventures, the legacy of a “just cause” was handed down in order to consecrate war. The “warrior culture” is among the most easily identified features of the Indo–European legacy: war is exciting and grabs our attention – peace is boring. No wonder the warrior legacy has had an indelible effect on Indo–European (and world) history”5.

In fact, if there were the will to do this, historians would also be able of creating, for example, the image of some heroic Semites, obsessed with war, conquest and epic tales of their victories. Thus, when the Akkadians appear in the history of Mesopotamia during the second half of the third millennium BC they do so as carriers of a very marked warrior ideology (and practice), which contrasts sharply with the previous period. We have evidence of this in iconography, where steles of those such as Sargon or Naram–Sin8 presented the monarch as a great warrior who leads his armies in an unbroken succession of glorious military victories. Moreover, Akkadian royal titles also incorporate numerous epithets which highlight the heroism and military capacity of the Akkadian monarchs: “conqueror” ( sa g. g i š.ra) 9, “powerful” (dannum)10, “leader of the troops” (muttarrî er i n 2 ) 11, and so on. Finally, the historical memory of the Akkad dynasty gave rise to a rich tradition of epic literature focused on the development of military deeds (real or invented) of Sargon and Naram–Sin. Some examples of this are found in stories like “The Conquering Hero”, “King of Battle”, “The Cuthean Legend”, etc12.

Also Francisco Villar, in a study of the Indo–European world, continued to point to these ideas:

Thus, following the argument of Villar but changing the characters, we may well define the Akkadians as members of a hierarchical society dominated by heroic values and led by a warrior monarchy. And if we continue with this process, finally we get a lengthy catalogue of ancient heroic warrior societies that ends up completely diluting the racist myth of unique Aryan warlike heroes.

“Es un tópico ampliamente difundido y rara vez discutido que los indoeuropeos eran un pueblo belicoso, que se extendió por Eurasia gracias a sus habilidades y técnicas guerreras, idea que, a la vista de los acontecimientos históricos, difícilmente podría ser puesta en duda. Y, sin embargo, llama la atención la notable penuria de términos específicos para la guerra y su técnica que existe en el ámbito del vocabulario común”6.

2. Aryan heroes in the Ancient Near East The expression of the Aryan military hero quickly outgrew the confines of Indo–European studies, to spread geographically and historically to other areas nearby, like Assyriology. At the beginning of the 20th century one of the subjects most studied by assyriologists was the

“En resumen, en sus primeros pasos por la historia y la protohistoria, los indoeuropeos se revelan como un pueblo fuerte para la guerra, que utiliza el caballo, bien montándolo, bien haciéndole arrastrar sus carros de guerra. La épica india, griega, germánica y celta refleja una sociedad dominada por los valores del

7

Villar 1996 [1991], 115. Börker–Klähn 1982, 127ff. 9 E.g. RIME 2, E2.1.1.8, l. 4. 10 E.g. RIME 2, E2.1.4.1, col. ii l. 2´. 11 E.g. RIME 2, E2.1.4.1, col. ii ll. 10´–11´. 12 See Goodnick Westenholz 1997. 8

3

Bunsen 1868–1870, 210. Schrader 1890, 190. 5 Winn 1995, 4f. 6 Villar 1996 [1991], 109. 4

18

A r ya n W a r r i o r s supposed Aryan influence on the societies of the Near East13. The data that had led to identify this influence were (1) the existence of Aryan names as well as words related to chariots and horses in the cuneiform documentation14, (2) subsequent references to gods of the Indian pantheon (Mitra, Varuna, Indra and Naššatiya) acting as divine witnesses to the treaty between Suppiluliuma of Hatti and Sattiwaza of Mittani 15, and (3) the frequent mention in the documentation of an alleged Aryan military elite specialised in the use of the war chariot: the maryannu16. At that time there was some consensus in explaining the presence of all these elements as the result of an Aryan invasion in the Near East that would have occurred in the 18th century BC. This invasion would have reached Mesopotamia, the Levant and Egypt and have culminated in the founding of the kingdom of Mittani in Upper Mesopotamia.

successive racial mixtures having again “semitised” the Assyrians. On being an essentially Semitic people again, their Aryan blood having drained away, the Assyrians were unable to maintain a military empire and they crumbled helplessly in a very short space of time. Fortunately, these types of racist theories no longer form part of contemporary Assyriological literature. During the 1960s and 1970s Assyriologists completely reformulated the question of Aryan influence in the region. Several researchers showed, contrary to what some of their colleagues had pointed out in previous decades, that linguistic evidence of Indo–Aryan origin in cuneiform sources was actually very limited and its importance had been clearly exaggerated20. In addition, the quantitatively insignificant linguistic and onomastic material cannot be regarded as an indicator of an imaginary political dominance by an Aryan nobility over the various kingdoms of the Near East. Actually, it was probably fossilised relics in the Hurrian language, provided by ancient Aryan groups that were culturally and linguistically fully assimilated with the Hurrians in that historical period. Therefore, the extent of these remains throughout the region would be the result of the creation and expansion of the Hurrian kingdom of Mittani 21.

As regards social structures, the arrival of the Aryans would have led to the imposition of a form of organisation until then unknown in the Near East: feudalism. Thus, as noted by O'Callaghan17, Mittani would have been a feudal kingdom articulated around the maryannu, who were defined as a military caste that offered their services to the king in return for land grants. One of the Assyriologists more interested in the Aryan question was Wolfram Von Soden. During the period of the Third Reich Von Soden clearly positioned himself alongside the Nazis18. His ideology went beyond the strictly personal sphere and is perfectly expressed in his work on the Assyrian Empire published in 1937 19. Von Soden's aim was to explain the political rise and subsequent fall of the Assyrian Empire. According to strict racist and anti–Semitic guidelines, Von Soden pointed out that during the second half of the second millennium BC the Assyrians left behind their traditional trading vocation to pursue war with a vigour and strength hitherto unknown. In his opinion, the change was due to strictly racial issues. Aryan populations of the disappeared kingdom of Mittani were integrated into the Assyrian territory and enriched that people, especially enabling it for war and giving it supremacy in the region. However, by the end of the 7th century BC the ancient Aryan influence was practically exhausted, annihilated by

One of the most obvious examples of desaryanisation of Ancient Near Eastern history is found in studies of the maryannu–group. In this regard, Von Dassow recently stated that the existence of this group in no way can be used as evidence of Aryan presence 22. Previously Speiser had noted that the term maryannu actually only denoted a social category, without any ethnic connotation23. Von Dassow continues the same argument noting that the ethnicity of the members of a social group or category is unrelated to the linguistic origin of the word used to designate that group or category. This can be seen very well, for example in Ugarit. There, text RS 11.715+ (KTU 4.69) contains a long list of people grouped according to their professions. In the section devoted to the maryannu it can be clearly seen that the vast majority of the individuals mentioned bear Hurrian names or, specifically Semitic ones, but not Aryan names24. The above table (table 1) serves as a perfect example to show that the idea of the existence of a dominant group of Aryan knights, in this case in Ugarit, lacks any compelling textual evidence to support it25. In fact, the vast majority of those legendary “Aryan” knights were in fact Semites26.

13

See, among others, Winckler 1907 and 1910; Meyer 1954 [1913], 671ff. and 1953 [1928], 33ff.; Childe 1988 [1926]; Albright 1930–31; Unger 1935 and 1936; Von Soden 1939. 14 See Hess 1993, 198 for an analysis of the Indo–Aryan names attested in the Amarna letters. See Watson 1999, 130f. and 2007, 146f. on the Indo–Aryan loanwords in Ugaritic. See Mayrhofer 1966, 13ff. for a compilation of Indo–Aryan words attested in cuneiform texts. 15 Beckman 1999, 42ff. nrs. 6A and 6B. 16 Composite word: Indo–Aryan márya (“man”) + Hurrian suffix –nni. See von Dassow 2008, 96f. for a recent discussion on the etymology of maryannu. See Wilhelm 1987–1990 and von Dassow 2008, 77 n. 188. for a discussion on the attestations of maryannu in the Ancient Near Eastern written sources. 17 O’Callaghan 1948, 64ff. 18 See, for example, Flygare 2006 and Vidal / Garcia, in press. 19 Von Soden 1937.

20

Mayrhofer 1966 and 1974; Kammenhuber 1968 and 1976– 1980; Diakonoff 1972. 21 Friedrich 1932, 148; Kammenhuber 1988, 41; Houwink Ten Cate 1995, 267. 22 Von Dassow 2008, 79. 23 Speiser 1954, 21. 24 See Vidal 2010. 25 See Van Soldt 2003, 688f. on the proper names of the maryannu in Ugarit. 26 Indo–Aryan proper names attested in Ugarit are just 1/3% (Gröndahl 1967, 298f.; Hess 1999, 509).

19

JORDI VIDAL As regards the scope of the Ancient Near East, it is now easy to identify in the literature an almost absolute rejection of the existence of any Aryan influence, whether social or cultural, in the historical development of the region. Moreover, this rejection is not the result of a prior ideological stance, but is based on a strict analysis of the sources. Philological studies certainly demonstrate the existence of Indo–Aryan linguistic and onomastic materials in cuneiform documents. However, this material is quantitatively limited and its presence can only be explained by being the result not of an invasion, not attested in any of our sources, but as fossilised remains present in the Hurrian language and disseminated in the region by these same Hurrians. Thus, the Aryan warriors, not only had no influence on historical societies of the Ancient Near East, but they did not even exist there.

4. Conclusions The historiography related to the studies of the origins of the Indo–European peoples, since the 19th century, has wanted to highlight their heroic warrior component. Of course, these characters are based on a particular reading of written sources and archaeology. However, the emphasis on this warrior/heroic aspect of the Aryans is also based on the willingness of many of the researchers to endow values of supremacy to that people. In fact, as noted earlier, it is not difficult to give such heroic military status to many other ancient peoples, among them the Semites. Hence the final image of the Aryan warrior is more a product of romanticism (and racism) than it is a relevant historical reality.

Table 1 ––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––– Linguistic origin of the maryannu cited in RS 11.715+ ––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––– Semitic Hurrian Anatolian Uncertain Etimology ––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––– nrn bn tlmyn bn arwṯ bn bly (Semitic?)27 bn rmyy bn adlḏn bn brsn (Indo–Aryan?, Anatolian?)28 bn ḥrm bn pṭḏn bn śbl (Semitic?)29 c ! bn ṯb nq bn ḫnzr (Semitic?, Hurrian?)30 bn ḥgbn bn nbdg (Semitic?, Anatolian?)31 bn ršpy bn tmr (Semitic?, Hurrian?)32 c bdḥgb bn prsn (Semitic?, Anatolian?)33 klby špšyn c bdmlk –––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––

27

Gröndahl 1967, 20, 117, 288; Van Soldt 2003, 689 n. 63. Gröndahl 1967, 298f.; Van Soldt 2003, 689 n. 65. 29 Gröndahl 1967, 29; Segert 1983, 213; Pardee 1989–1990, 425; Watson 1995, 226; Van Soldt 2003, 689 n. 66. 30 Ward 1961, 34f.; Gröndahl 1967, 22, 28, 140; Pardee 1989–1990, 399; Watson 1990, 119; 2007, 167; Van Soldt 2003, 689 n. 63. 31 Gröndahl 1967, 206, 243; Van Soldt 2003, 689 n. 65. 32 Noth 1942, 151 n. 1; Gröndahl 1967, 260; Pardee 1989–1990, 424; Watson 1995, 228; 1996, 105; Van Soldt 2003, 689 n. 64. 33 Noth 1942, 151 n. 1; Gröndahl 1967, 175, 268, 289; Dietrich / Loretz 1979, 193; Pardee 1989 / 1990, 419; Van Soldt 2003, 689 n. 63. 28

20

A r ya n W a r r i o r s Bibliography ALBRIGHT, W. F. 1930–1931, “Mitannian maryannu, “chariot–warrior”, and the Canaanite and Egyptian Equivalents”, AfO 6, 217–221.

– 1988, “On Hittites, Mitanni–Hurrians, Indo–Aryans and Horse Tablets in the Iind Millennium BC.”, in T. MIKASA (ed.), Essays on Anatolian Studies in the Second Millennium BC., Wiesbaden, 35–46.

BECKMAN, G. 1999, Hittite Diplomatic Texts (second edition), Atlanta.

MAYRHOFER, M. 1966, Die Indo–Arier im alten Vorderasien, Wiesbaden.

BERNABÉ, A.; ÁLVAREZ–PEDROSA, J. A. 2004, Historia y Leyes de los Hititas. Textos del Reino Medio y del Imperio Nuevo, Madrid.

– 1974. Die Arier im Vorderen Orient – Ein Mythos?, Wien.

BÖRKER–KLÄHN, J. 1982, Altvorderasiatische Bildstelen und vergleichbare Felsreliefs, Mainz am Rhein.

MEYER, E. 1953 [1928], Geschichte des Altertums, II/1: Die Zeit der ägyptischen Grossmacht, Basel.

BUNSEN, C. 1868–1870, God in History, or the Progress of Man’s Faith in the Moral Order of the World, London.

– 1954 [1913], Geschichte des Altertums, I/2: Die ältesten geschicghtlichen Völker und Kulturen bis zum Sechzehnten Jahrhundert, Basel.

CHILDE, V. G. 1988 [1926], The Aryans, New York. MICHELET, J. 1833 [1830], Histoire romaine. Première partie: République, Paris.

DIETRICH, M.; LORETZ, O. 1979, “Einzelfragen zu Wörtern aus den ugaritischen Mythen und Wirtschaftstexten“, UF 11, 189–198.

NOTH, M. 1942, “Die syrisch–palästinische Bevölkerung des zweiten Jahrtausends v.Chr. im Lichte neuer Quellen”, ZDPV 65, 9–67.

DIAKONOFF, I. M. 1972, “Die Arier im Vorderen Orient: Ende eines Mythos”, Orientalia 41, 91–120.

O’CALLAGHAN, R. T. 1948, Aram Naharaim. A Contribution to the History of Upper Mesopotamia in the Second Millennium BC, Roma.

FLYGARE, J. 2006, “Assyriology under Nazism: A contextual analysis of three texts by Wolfram von Soden from 1936–38”, Journal of Associated Graduates in Near Eastern Studies 11, 3–42.

PARDEE, D. 1989–1990, “Ugaritic Proper Nouns”, AfO 36 / 37, 390–513.

FRIEDRICH, J. 1932, “Arier in Syrien und Mesopotamien”, RlA 1, 144–148.

SCHRADER, O. 1890, Prehistoric Antiquities of the Aryan peoples. A manual of comparative philology and the earliest culture, New York.

GOODNICK WESTENHOLZ, J. 1997, Legends of the Kings of Akkade, Winona Lake.

SEGERT, S. 1983, “The Last Sign of the Ugaritic Alphabet”, UF 15, 201–218.

GRÖNDAHL, F. 1967, Die Personennamen der Texte aus Ugarit, Rome.

SPEISER, E. A. 1954 “The Alalakh Tablets”, JAOS 74, 18– 25.

HESS, R. S. 1993, Amarna Personal Names, Winona Lake.

UNGER, E. 1935, “Unsemitisches und Indogermanisches in der altorientalischen Kunst”, Forschungen und Fortschritte 11, 329–331.

– 1999, “The Onomastics of Ugarit”, in W. G. E. WATSON & N. WYATT (eds.), Handbook of Ugaritic Studies, Leiden/Boston/Köln, 499–528.

– 1936, “Zur Entwicklung des sumerischen Hakenkreuz– Symbols”, Forschungen und Fortschritte 12, 153–155.

HOUWINK TEN CATE, P. H. J. 1995, “Ethnic Diversity and Population Movement in Anatolia”, in J. M. SASSON (ed.), Civilizations of the Ancient Near East, New York, 259–270.

VAN SOLDT, W. 2003, “The Use of Hurrian Names at Ugarit”, UF 35, 681–707.

KAMMENHUBER, A. 1968, Die Arier im Vorderen Orient, Heidelberg.

VIDAL, J. 2010, “La cuestión indo–aria en los estudios ugaríticos”, Gerión 28, 27–39.

– 1976–1980, “Indogermanen”, RlA 5, 90–96.

VILLAR, F. 1996 [1991], Los indoeuropeos y los orígenes de Europa. Lenguaje e Historia, Madrid.

21

JORDI VIDAL VON DASSOW, E. 2008, State and Society in the Late Bronze Age. Alalaḫ under the Mittani Empire, Bethesda.

– 2007, Lexical Studies in Ugaritic, Sabadell. WILHELM, G. 1987–1990, “Marijannu”, RlA 7, 419–421.

VON SODEN, W. 1937, Der Aufstieg des Assyrerreichs als geschichtliches Problem, Leipzig. – 1939, “Die Indogermanen im Alten Forschungen und Fortschritte 15, 41–43.

WINCKLER, H. 1907, “Vorlaufige nachrichten uber die Ausgrabungen in Boghaz–koi im Sommer 1907: Die Tontafelfunde”, MDOG 35, 1–59.

Orient”,

– 1910, “Die Arier in den Urkunden von Boghaz–köi”, OLZ 13, 289–301.

WARD, W. A. 1961, “Comparative Studies in Egyptian and Ugaritic”, JNES 20, 31–40.

WINN, S. M. M. 1995, Heaven, Heroes, and Happiness. The Indo–European Roots of Western Ideology, Lanham.

Watson, W. G. E. 1990, “Ugaritic Onomastics (1)”, AuOr 8, 113–127. – 1995, “Ugaritic Onomastics (4)”, AuOr 13, 217–229. – 1999, “Ugaritic Lexicography”, in W. G. E. WATSON & N. WYATT (eds.), Handbook of Ugaritic Studies, Leiden/Boston/Köln, 122–133.

22

First notes on a treasure from Mes Aynak BARBARA FATICONI ABSTRACT This article is a first attempt to catalogue a small treasure found in 2011 at Mes Aynak site, in the Logar province (Afghanistan). Considered second largest copper mine in the world, Mes Aynak is also a huge Buddhist site, which was for many centuries active in close relationship with the mine. In 2007 the China Metallurgical Group Corporation obtained a concession for mining from the Afghan government. As a consequence, in 2009 started a series of archaeological operations to save, at least through an accurate documentation, the memory of the site. In 2012, during one of these operations, a small hoard of object in gold, silver, bronze and precious stones was found in a zone that can be identified as an area of metalworking. The state of preservation of the objects at the time of the discovery, as well as its non–homogeneous content, does not provided a precise date but some of the objects allow us to advance new hypotheses about the role played Mes Aynak throughout the centuries under the rule of the different populations that succeeded in the area. KEYWORDS Afghanistan – Buddhist site – stūpas – Workshop – Treasure – Jewels – Seal – Alkhan coinage –Mines The site of Mes Aynak, located about 40 km south of Kabul in the Logar province, is considered the second– largest copper reserve in the world to date. In 2007, the Afghan government awarded a concession to the China Metallurgical Group Corporation for mining the site. However, besides its wealth of mineral deposits, Mes Aynak is also one of the biggest and most important archaeological sites discovered in recent decades. This intriguing site consists of a series of Buddhist structures orbiting around Baba Wali Mountain (the “copper” mountain). For several centuries the mountain with its copper mining and the monasteries must have lived in a close relationship of osmosis and symbiosis. Due to the financial needs of the country, the Afghan Institute of Archaeology started archaeological rescue excavations in 2009, to document as much as possible of the site, in a nearly impossible race against time. In this challenging task the DAFA (Délégation Archéologique Française en Afghanistan) assists the Afghan archaeologists. Since the summer of 2011 an international team of archeologists collaborates with the Afghan archeological teams to save this part of Afghan heritage that is bound to be destroyed in the near future. Until now just a small group of experts have been able to get a glimpse of the material culture of the site, but their efforts are largely inadequate in light of the huge scale the project. In spite of the inadequacy of the means, the site in the last years the site as continued to deliver a considerable mass of artifacts that have not yet been adequately presented to a wider audience of scholars. A first attempt to collect and put order into all this material was made in May 2011, when on the occasion of the first exhibition of Mes Aynak held in the National Museum in Kabul, a small exhibition catalogue

was published1. The small hoard presented here was not included in the exhibition as it was found in October 20112. The Site

Fig. 1: Map of Afghanistan

Mes Aynak is southeast of the Gul Dara valley3. The site, with an estimated size of more than a thousand hectares, 1

Massoudi 2011. I was one of the first two Senior Archaeologists of the team from September 2011 to February 2012. The hoard was found in a zone under my supervision. I would like to thank the members of the MAAP (Mes Aynak Archaeological Project), the Ministry of the Mine and the Ministry of Information of the Islamic Republic of Afghanistan for the opportunity to work in Mes Aynak. 3 The first researchers to mention Mes Aynak were Fussman and 2

BARBARA FATICONI consists of several settlements, predominantly related to the exploitation of the mine, but also of different Buddhist structures. The excavations are still in progress, but some information about the topography of the site is available. On the north–east top of the Baba Wali mountain, where is the copper mine is located, there is a large fortress (site 005) overlooking the pass that gives access to the valley to the east. From the southern edge of the fortress, a sturdy stone wall stretches from north–east to south–west, crowning the crest of the mountain. It supports a number of other walls that descend to the east, where they intersect with other walls oriented in a N–S direction, presumably to create a system of terraces that covering the eastern slope of the copper mountain.

U–shaped underground space, found against the southern foundations of the building.

Fig. 2: Location of some of the sites east of Baba Wali Mountain.

Just below the fortress, further down the slope, towards the valley, a small agglomeration of structures is visible (sites 029–030). The organization of types of structures suggest a domestic quarter, perhaps including some sacred areas, as possibly confirmed by a small painted stele with the Dīpaṃkara Jātaka found in situ4. All around the mountain, a series of monasteries and places of worship have been identified. In 2009–2011 research focused on four sites: Kafiria Tepe, Gol Hamid, Shah Tepe and Shamar Tepe. All these complexes yielded stūpas, murals, sculptures and others materials, all waiting for study.

This underground space, built of large stones, was roofed by slab in two places. These slabs made it possible to cross this "channel" (“water place?”)5. In addition to this special architectural feature, the presence of bronze tools and many fragments of shells6 all over the site strengthened our conviction that this area comprised workshops. Following this intuition we decided to continue the excavation to the east of this site in the hope to find further structures related to those already excavated.

The zone in which the hoard was discovered comprises a series of buildings (sites 028–033) situated on the eastern side of the valley, east of the copper mountain with its system of terraces, and at the base of the high hill on which Shah Tepe (site 006) lies. The buildings of sites 028–033 contain features, such as large rooms and underground spaces, which can be identified as workspaces. The hoard’s context The discovery was made in site 033 E, east of Baba Wali, beyond the road (possibly old) that apparently divides the quarters more directly related to the mountain (sites 029– 030) from those sectors (sites 028–033) at some distance from the mountain, probably more independent. The 2010–2011 excavations carried out by the Afghan archaeologists in this area had brought to light an interesting complex (033 W) consisting of different units. A closer analysis carried out in 2011–2012 allowed us to recognize a series of large rooms, one of which probably vaulted, and a stone staircase giving access to an upper floor (FIG. 3). The most intriguing structure, however, is a

Fig. 3: Sketch plan of Site 33 W, not to scale.

We dug twenty squares (2x2 m each) at about two meters from the eastern boundary of the 2010–2011 excavation (trench A). The first structure appeared just below the

Le Berre who recognized the potential of the site in 1976 (Fussman and Le Berre 1976, 101–102). The site was later visited by Berthoud, Besenval, Cesbron, and Liszak–Hours in 1977 in order to check the extension of the mine. But the first archaeological excavation only took place in the 1980 thanks to Jean–Claude Gardin and Bertille Lyonnet. Finally Mes Aynak was also cited in the Archaeological gazetteer of Afghanistan published by Warwick Ball (Paris, 1982). 4 The small painted stele with the Dīpaṃkara Jātaka was on display at the 2011 exhibition in Kabul; Massoudi 2011, 42.

5

However, similar slabs may have covered the entire “channel”. This does not exclude its use as a "reservoir" for the water. 6 Unfortunately this material, still spread out on the surface near the trench, had not yet been documented and properly stored at the time of our arrival. In the absence of an appropriate recording system is impossible to retrieve the exact origin in the excavation.

24

F i r s t n o t e s o n a t r e as u r e f r o m M e s A yn a k modern surface. It was a stone wall of which 7,5 meters were exposed. The wall continued beyond the eastern limit of the trench in a southeastern direction (su 004). The area north–east of the wall revealed features related to a workshop. A “deposit” (su 005) of small stone fragments, rich in copper, situated against the north face of the wall, suggests a pile of raw material ready for processing. Around it a context (su 011) containing traces of fire and fireplaces (su 026 and su 027) has been detected. Most interesting is su 026: a small pit (20 cm in diameter and 30 cm deep) dug in su 011, the thick sand layer. At the bottom of the pit a thick deposit of ash, coal and copper has been found. The bottom and the sides of the pit were still coated with a thin layer of copper (Fig. 4).

piece of cloth. The latter has deteriorated through the centuries, but shreds of tissue were still attached to some objects. At the moment of the discovery, the hoard consisted of a heap of bronze and gold objects, as well as stones mixed with gray sand. Catalogue The hoard consists of twenty–five objects and lots which can be subdivided into three main classes: jewels, tools, and sculptures. The first two classes can be further divided in subclasses according to their materials. 1. Jewels 1.1. Gold 1.2. Bronze 1.3. Beads 2. Tools 2.1 Bronze 2.2 Silver 3. Sculptures 3.1 Bronze 1 – Jewels 1.1 Gold 1.1.1 Ear ring Provisional Inventory: Height: Width: Thickness: Weight:

N° 10 2.9 cm 2.7 cm

0.5 cm 5.0 gr Material: Gold, lapis lazuli and carnelian. Description: oval shape. In the centre an oval carnelian (1.1 x 0.9 cm) with chamfered edge surrounded by a plain border with thin raised edges (0.2 cm). All around the oval edge a row of gold beads. On both sides and at the bottom three small pendants, each made of a tiny lapis lazuli cylindrical bead and a micro pearl (diameter 0.3 cm), with a gold disk element in between. In the upper part the suspension ring is still present.

Fig. 4: SU 026.

Fig. 5: Localization of fireplaces ssuu 026–027 and the hoard su 049.

1.1.2 Seal ring Provisional Inventory: Height: Width: Thickness: Weight:

All evidence suggests that su 026 was a small fireplace used to process tiny amounts of metal and the space NE of wall 004 was a small workshop. The presence of the hoard, close to the pit, in the same layer of gray sand (su 011), seem to confirm this hypothesis. The objects in the hoard were tightly packed and originally wrapped in a

25

N° 11 2.6 cm 2.0 cm 0.6 cm 7.0 gr

BARBARA FATICONI Material: Gold and carnelian. Description: gold ring with central carnelian seal (1.3 x 1.6 cm), inserted into a shallow oval with chamfered edge. The seal shows a crouching lion facing left. Stout body, massive folded legs with large paws, claws and long tail snaking rising above the rear part of the figure. Thick mane and gaping jaws characterize the upright head.

1.1.5 Ring Provisional Inventory: N° 14 Height: 1.8 cm Width: 1.8 cm Thickness: 0.6 cm Weight: 2.0 gr Material: Gold and amethyst (?). Description: a simple golden circle on which a high bezel is grafted. The latter is composed of a flat collar connected to a bell–shaped element. A small cabochon (diameter 0.4 cm) amethyst (?) is set atop the bell.

1.1.3 Ear ring Provisional Inventory: N° 12 Height: 1.8 cm Width: 1.9 cm Thickness: 0.4 cm Weight: 1.0 gr Material: Gold, lapis lazuli, turquoise and pearls. Description: the earring shows a main element with a rich and complex design, below a simple "bow" closure. A central round object can be identified as a pomegranate fruit (this might be confirmed once the object is restored) surrounded by an arch decorated with granulation. Two symmetrical– composite decorations flank the arch, each consisting of a central round lapis lazuli element, with two small turquoises leaves at the top and a turquoise heart–shape element at the bottom. Below, a flat band embossed with five rings, each holding a small pendant composed of elongated golden elements within a granulated circle. Two micro pearls, attached laterally on the upper part of the earring, complete the jewel.

1.1.6 Lot of buttons Provisional Inventory: N° 17 Number of pieces: 18 Height: 1.7 cm Width: 1.7 cm Thickness: 0.1 cm Weight: 14 gr Material: Gold and iron. Description: 18 hemispherical buttons cast in moulds, decorated with an embossed rotating motif and a tiny globe at the centre. The back side is hollow and hosts a small iron loop. In all probability these were applied to clothes. The presence of the loop, instead of two simple holes to sew on to the fabric, distinguishes them from simple decorative plaques. 1.1.7 Ear ring Provisional Inventory: N° 18 Height: 2.3 cm Width: 1.9 cm Thickness: 0.7 cm Weight: 2.0 gr Material: Gold, turquoise and garnet (?). Description: round shape. A plain border with thin raised edges (0.2 cm) surrounds one large flat central cabochon garnet (diameter 0.7 cm). A row of 13 golden beads decorates the edge of the pendant. A gold tiny ring set with turquoise makes the suspension ring.

1.1.4 Ring Provisional Inventory: N° 13 Height: 2.4 cm Width: 2.5 cm Thickness: 0.7 cm Weight: 4.0 gr Material: Gold and emeralds (?).

1.1.8 Lot of gold beads Provisional Inventory: N° 21 Number of pieces: 36 Height: 0.5 cm Width: 0.5–0.6 cm Thickness: 0.5 cm Weight: – Material: Gold.

Description: gold ring with raised star–shaped bezel. Unfortunately the central stone is missing, while on the sides two green cabochon stones (emeralds?) are preserved (diameter 0.3 cm). Looking at the ring from the side, the volute decoration holding the two stones is clearly visible.

26

F i r s t n o t e s o n a t r e as u r e f r o m M e s A yn a k Description: 36 small hollow globes decorated with granulation. All globes consist of an extremely thin sheet of gold with little spheres welded all over the surface. Two opposite holes allow

deep bezels. The central oval (in one case almost round) contains a carnelian. At the four corners, a drop–shaped bezel contains either a turquoise or a lapis lazuli. On the two short sides four small holes form a system of attachment. The back is completely flat. In all probability the eight plaques were part of a necklace, a bracelet or a belt.

assembling. 1.2 1.1.9

Bronze

Lot of bell–shaped plaques

Provisional Inventory: Number of pieces: Height: Width: Thickness: Weight: Material: Gold.

1.2.1 Ring Provisional Inventory: N° 15 Height: 1.7 cm Width: 1.7 cm Thickness: 0.7 cm Weight: 1.0 gr Material: Bronze. Description: the high level of corrosion makes it almost impossible to know the exact shape of this object. The thin ring seems to be adorned by a small “rosette”. Any further description must await restoration.

N° 20 12 1.1–1.3 cm 0.9–1.2 cm 0.3 cm –

1.2.2 Ring Provisional Inventory: N° 16 Height: 2.2 cm Width: 2.4 cm Thickness: 0.3 cm Weight: 5.0 gr Material: Bronze. Description: slightly better preserved than 1.2.1. The ring has a thicker section and is marked by a series of parallel incisions along the entire length. The central motif is a “bud” with three round “petals” on the sides.

Description: 12 bell–shaped plaques, one slightly differs in shape from the rest. Eleven specimens have a wide bell–shaped silhouette. The bottom ends in three tips, each bearing a tiny golden globe. Other two globules decorate the body of the bell, vertically aligned with the central point. A small circular ring on top serves as loop for attachment. The back of the plaques is flat. The twelfth bell is slimmer and more elongated. The bottom edge is flat, and decorated with a waving “gimp”, which creates a sort of stylized clapper with its curvilinear movement. In this case, the suspension loop is flat and rather wide. The back of the plaque is flat.

1.2.3

1.1.10 Square plaques with semi–precious stones Provisional Inventory: N° 19 Number of pieces: 8 Height: 1.3–1.4 cm Width: 1.2–1.4 cm Thickness: 0.4 cm Weight: 4.0 gr Material: Gold, carnelian, lapis lazuli and turquoise.

Hairpin (?)

Provisional Inventory: N° 6 Height: 14 cm Width: 6.2 cm Thickness: 0.5 cm Weight: 42 gr Material: Bronze and gold.

Description: big bronze hairpin with pointed ends. The upper half of the straight segments and the curved segment are completely covered with thin gold foil. The latter is applied in transversal segments, giving the appearance of a “flexible pipe”. The gold foil ends on

Description: eight square plaques made of gold sheet. The slight differences between the various elements are the result of manual crafting. Each plaque has five

27

BARBARA FATICONI both sides with flat elements. Some threads and fragments of tissue, in all likelihood belonging to the cloth in which the hoard was wrapped, are still stuck to the bronze segment. 1.3

small dimensions, often so tiny that it is impossible to measure them without special tools.

Beads

1.3.1 Lot of coral beads Provisional Inventory: N° 23 Number of pieces: 4 Height: 1.4 cm Width: 0.4 cm Thickness: 0.3 cm Weight: – Material: Coral.

All beads are pierced, witnessing the advanced skills of the craftsmen who created them. 2 – Tools 2.1. Bronze

Description: elongated oval beads with oblong perforation. The four beads, made of light pink coral, slightly vary in shape and size.

1.3.2 Unidentified object Provisional Inventory: N° 1 Height: 9 cm Width: 0.2 cm Thickness: 0.2 cm Weight: 1.0 gr Material: Bronze.

1.3.2 Lot of mixed beads Provisional Inventory: N° 24 Number of pieces: 78 Height: – Width: – Weight: 30 gr Material: Carnelian, lapis lazuli, shell, bone, glass paste, agate.

Description: a thin shaft with one pointed end and with a small cup at the other end. The shaft is decorated with a series of incisions, forming a kind of ring–motif; the decoration stops at about one centimetre of the tip. The tip is smooth. The small cup is 0.5 cm in diameter. At the time of the discovery, tissue fragments still stuck to the surface did not allow any observations inside the cup. The presence of additional elements (e.g., a small hook) can only be conjectured. The pointed tip suggests that the object was used as a pin even though the general form recalls that of a small spoon. Maybe a tool for precision works with small quantities of substances (manufacture of jewellery)? Or maybe objects for toilette like palettes for the powder makeup or to clean ears. Restoration will hopefully provide more information.

Description: beads made of various materials: glass paste, carnelian, garnet, coral, agate, lapis lazuli, bone and shell. We also noted a wide range of shapes: round, oval, elongated oval, barrel, and cylindrical and many others with more complex and original shapes. This very interesting lot deserves a more accurate study of materials and shapes.

1.3.3 Unidentified object Provisional Inventory: N°2 Height: 6 cm Width: 0.2 cm Thickness: 0.2 cm Weight: 1.0 gr Material: Bronze.

1.3.3 Lot of pearl beads Provisional Inventory: N° 25 Number of pieces: – Height: – Width: – Weight: 25 gr Material: Pearl. Description: very high number of pearls varying in size and in shape. A significant part of them is of very

28

F i r s t n o t e s o n a t r e as u r e f r o m M e s A yn a k Description: similarities between this object and the previous are very strong. However fragmentation of the object does not allow us to establish whether it had a pointed end. Only the thin shaft is decorated in a different way with a spiral motif. A substantial fragment of tissue is stuck to the shaft. 1.3.4 Tablet with bowl Provisional Inventory: Height: Width: Thickness: Weight: Material: Bronze.

Description: the thin handle is decorated with a spiral motif and ends in a small globe. Near the scoop the handle is thicker and sturdier. The scoop is egg–shaped and shallow, almost flat, allowing the spoon to contain but small quantities. This feature, in addition to the fragility of the handle, makes this spoon hardly suitable for consumption of food. Most probably the spoon was destined for "symbolic" use.

N° 3 8.7 cm 8.6 cm 1.6 cm 52 gr

2.1.6 Unidentified object Provisional Inventory: N° 5 Height: 11.7 cm Width: 0.8 cm Thickness: – Weight: 16 gr Material: Bronze and silver.

Description: object with a pointed end and at the opposite end a small cup with a “collar” formed by a ring. The presence of additional elements in the cavity of the cup could not be observed prior to cleaning and conservation. The cup is connected to a wedge–shaped sturdy handle, encircled by a winding thread. The handle is divided into three segments by two parallel rings formed by the winding thread. A tapering silver “collar” (length 2.0 cm) connects the cup to the thicker end of the handle. This “collar” is decorated with a pattern of circles that gradually decrease in thickness towards to the opposite end to the handle. Despite the similarities with the previous objects, 2.1.1 and 2.1.2, we exclude the use as a pin due to the big size of its shaft and its weight. We can only conjecture about its use as a “tool” in the manufacture or repair of jewellery, or toiletries (ear cleaners?) and makeup.

Description: a tablet with small central bowl. The sheet is bent down on all four sides to create a stable support. In the centre, the metal has been worked to create a perfectly hemispherical cavity (diameter 4.0 cm). Fragments of tissues are still stuck to the surface. The object may have been part of a set of jewellery– manufacturing tools or toiletries. 1.3.5 Lot of flat fragments Provisional Inventory: N° 22 Number of pieces: Height: – Width: – Thickness: Weight: Material: Bronze.

12 – –

2.1.7 Plate Provisional Inventory: N° 8 Height: 15.5 cm Width: 8.7 cm Thickness: 0.1 cm Weight: 51 gr Material: Bronze and silver. Description: Twelve fragments of bronze sheet forming the support of the tablet with bowl (2.1.3). 1.3.6 Spoon Provisional Inventory: Height: Width: Thickness: Weight: Material: Bronze.

N° 4 9 cm 3.3 cm 0.1 cm 8 gr

29

BARBARA FATICONI Description: oval–shaped plate with a very shallow central depression surrounded by a large border and two flat polylobate handles. The central depression and the two handles are decorated in light relief. Corrosion as well as shreds of cloth sticking to the surface made it very difficult to obtain an integral view of the decoration. In the central image we observed a man sitting cross–legged, the torso rendered in a three– quarter turn and one of the two arms raised. Some details, crucial for identification, were still visible. The legs are crossed at the ankles and the raised arm holds a flower. The shape of the flower, not too clear, suggests a bud with three petals. Another noteworthy detail to note is that the head, although not plainly visible in outline, is clearly deformed into an elongated shape. The two images on the handles of the plate represent two busts that start at shoulder height. Unfortunately, only one of them could be analysed; it shows a male face, in the shape of an elongated oval in slightly three quarters view. Traces of ageminature seem to complete the decoration, with phytomorphic motifs below the two busts on the handles, and perhaps all around the edge of the plate. 2.2

2 – Sculpture 3.1 Bronze 3.1.1

Provisional Inventory: N° 9 Height: 4.7 cm Width: 2.9 cm Thickness: 1.0 cm Weight: 15 gr Material: Bronze, gold and carnelian. Description: bronze hand, most probability part of a sculpture. What remains is a tapering hand holding the base of a cup (although we cannot exclude the possibility that it held a flower). The thumb and index finger, which usually hold a flower’s stem7, in this case seem to support a cup. On the edges of the cup two birds are depicted face to face, separated by a round carnelian bead. Despite the high level of corrosion, we observed that the beaks are slightly open and that the two birds have long tails. The bead is perfectly spherical and polished. The bead is pierced, maybe evidence of re–use. A simple gold ring embellishes the ring finger of the hand. Its pill–shape stands out in size in comparison with the hand.

Silver

2.2.1 Spoon Provisional Inventory: Height: Width: Thickness: Weight: Material: Silver.

Hand with cup, fragment of sculpture

Discussion N° 7 13 cm

The first characteristic of the hoard, as the catalogue shows, is its heterogeneous nature. The objects are attributable to different cultural and chronological contexts, though some of them can be grouped into small units.

3.9 cm 0.1 cm 7 gr

The first and most striking group of objects are surely the golden jewels (1.1). A comparison with the gold of Tillia Tepe (1st century AD)8 is inevitable when we look at the metal work (the use of granulation9) and at the presence of decorative motifs like the famous heart motif. The Mes Aynak jewels can be compared also with older specimens or models that were in use for a very long time span. The lot of golden buttons (1.1.6) is a good example. This kind of decoration for cloths was in use among the nomadic populations of Central Asia for many centuries. Among

Description: a spoon which resembling the bronze spoon described above (2.1.5). The handle is decorated with a spiral motif. One end has a small globular element, whereas the scoop is egg–shaped, shallow, almost flat. The size of the scoop and the thinness of the handle are disproportionate. For this spoon we propose the same hypothesis (cf. 2.1.5) of a "symbolic" use.

7

Callieri 2002, 128. The six graves, in which more than 20.000 gold objects were discovered, are dated to the 1st century AD even if among the objects some can be dated to the end of 2 nd –1st centuries BC (see Zeymal 1999). 9 A technique already known long before the jewellery in the treasure of Tillia Tepe were created, as indicated by the jewels found in the Astrakhan province, dated to the 5th century BC; Anisimova et alii 2005, 84–85. 8

30

F i r s t n o t e s o n a t r e as u r e f r o m M e s A yn a k the earliest examples, comparable to the Mes Aynak buttons, are the “spiral” buttons found in Tasmola, in central Kazakhstan, dated to the 6th century BC (Fig. 6 A).10 The latter, however, have been interpreted as decorations for reins, whereas the Mes Aynak buttons more refined and elegant resemble the cloths ornaments of Tillia Tepe (Fig. 6 B)11.

This decorative motif already known since earlier periods – its “omnipresence manifeste”, in the words of V. Shiltz16 – becomes a characteristic trait of most of the gold jewels at Tillia Tepe. The Mes Aynak pendants have many direct analogies in the treasure of Tillia Tepe, notably in the choice of materials (carnelian, garnet, turquoise) and techniques. The difference seems to be in their use. The pendants from Mes Aynak seem to constitute single jewels whereas those from Tillia Tepe probably were parts of a composite jewel 17, like the eight plaques from Mes Aynak decorated with carnelian, lapis lazuli and turquoise (1.1.10). The holes present in two of the opposite sides of each plate suggest that they had to be linked to form a line, maybe of a necklace, a bracelet or even a belt. Surely, the manufacture and processing of gold foil, as well as the use of the bezel, are all part of a long tradition of local jewellery. The analysis and interpretation becomes more complex and uncertain for the seal ring (1.1.2). The form is well attested in Tillia Tepe18, but the seal itself allows for a much wider interpretation. The motif of the lion was used for a long period in a wide area. Comparing the lion on the Mes Aynak seal with those depicted on the seals studied by P. Callieri in his exhaustive catalogue 19, we observe similarities with various groups. The manes of the Mes Aynak lion fit into the tradition of the so–called Greco–Persian seals dated to the 4th–2nd cent. BC (Fig.8A)20. The posture of the animal and the position of the tail recall seals of Callieri’s Class XIX21, dated to the 9th–11th cent. AD (Fig. 7B). Also the area of origin cannot be easily identified. The way in which muzzle and nose are engraved seems to belong to a local tradition, more precisely from the “North–West of the Indian Sub– Continent”, as suggested by Callieri22(Fig. 8C).

Fig. 6 A–B: Buttons from Tosmola and plaques from Tillia Tepe.

The pieces from Tillia Tepe are not buttons, such as those of Mes Aynak, but plaques, provided with two holes for application on tissues. The Mes Aynak buttons and the Tillia plaques share the use of moulds 12 in the production process. In both case the results are luxury hemispherical decorations. Beside gold buttons, the hoard also included bell–shaped plaques (1.1.9), closely comparable with those of Tillia Tepe (Fig. 7)13.

The wide range of parallels in space and time for this seal do not allow to suggest a precise date, which also applies to the chronological horizon of the Mes Aynak hoard. The beads merit a more thorough analysis, which is, however, beyond the scope of these first notes. The golden beads with the granulated decoration (1.1.8) clearly belong to the gold production of Central Asia well represented in Tillia Tepe23. The lots with beads in various materials (1.3.2), coral (1.3.1) and pearls (1.3.3) deserve a more careful study of the individual pieces. The analysis of the types of beads as well as the materials they are made of could provide new information about the commercial routes and networks through which objects and raw materials reached Mes Aynak. A more thorough analysis would also allow for the identification of the

Fig. 7: Plaques bell–sheep from Tillia Tepe.

The comparisons between the two treasures 14 do not end here. The heart motif of the delicate Mes Aynak earring (1.1.3) occurs on many gold items from Tillia Tepe15. 10

Arbore Popescu et alii 1998, 141. Sarianidi 1985, Ill. 21, 81. 12 For a reconstruction of the production techniques see Anisimova et alii 2005, 46–47. 13 Sarianidi 1985, Cat. Inv.1.29, 3.42, 230. 14 We use here the term “treasure”, as it was classified at the time of discovery, even if probably “hoard” would be more appropriate for this set of objects of various nature, wrapped in a cloth, possibly waiting to be repaired. 15 Sarianidi 1985, n°33, 86; n°35, 88; n°48, 104, just to give some exemples. 11

16

Schiltz 2006, 76 interprets the hearts as stylized ivy petioles and underlines their frequent use as decoration in Tillia Tepe. 17 Sarianidi 1985, Cat. Inv.3.3, 236. 18 Sarianidi 1985, Cat. Inv. 6.6, 255; 3.61, 243. 19 Callieri 1997. 20 Sarianidi 1985, cat. N° 4.11, 4.12 or 4.15, Pl 13. 21 Sarianidi 1985, cat. N° 7.53 and 7.54, Pl 27. 22 Sarianidi 1985, cat. N° 7.51 and 7.52, Pl 26–27. 23 Sarianidi 1985, n° 61, 113; n° 63, 115.

31

BARBARA FATICONI various techniques applied. As we said above, the findings of remains of shells in the area adjacent to the find spot of the hoard are evidence of a local processing of these materials.

used to hold clothes and, if necessary, to take products stored in flacons29. An alternative hypothesis, especially in the light of another object found in the hoard (2.1.6), is the use of it as ear cleaners. This particular class of toiletries, very common also in present day Asia, probably reflects an ancient tradition, as suggested by a specimen found in the Kostobe necropolis (6th–8th AD) in Kazakhstan (Fig. 9)30.

Fig. 9: Ear cleaners from Kostobe.

This silver spoon also helps us to identify the bronze spoon with silver decorations found in Mes Aynak (2.1.6). The size and weight of the object found in Mes Aynak do not support a classification as pin. One of the possibilities is to consider the object as part of a luxury toilet set which could also be used like ear cleaners. Although direct parallels for the tablet with bowl (2.1.3 and 2.1.4) were not found, there is no doubt that they represent containers for the processing of small quantities of unknown substances. One might think of a toilet set to mix or prepare substances for makeup. The absence of metal residues at the bottom of the bowl excludes the use of the bowl as a container for small quantities of metal in fluid state, although cleaning and restoration of the bowl may reveal more information on the presence of residual substances.

Figs. 8 A, B, C : Seals.

The identification and dating of the bronze jewellery appears to be more difficult. The two rings (1.2.1 and 1.2.2) are corroded and the very common shape does not allow a precise classification. The last bronze jewel (1.2.3) is an object of uncertain interpretation. We have included it in the group of jewellery due to its general impression of luxury; its gold inlay suggests that this object was meant for a special use. Its form brings to mind a hair ornament, like a big hairpin. The size as well as the weight (42 gr) makes this hairpin uneasy to wear, but comparison with the “monumental” hair ornaments found in the graves of Tillia Tepe gives credibility to our interpretation24.

The Mes Aynak hoard includes two spoons, one in bronze (2.1.5) and one in silver (2.2.1). What strikes as odd is the disproportionate size of scoop and handle. This characteristic is not unknown and the function of these particular spoons has long been debated31. It has been suggested that the spoons were used in shamanic rituals as typical of nomadic cultures. For sure, the weakness of the handle does not support their use for preparation or consumption of food. The ritual use (aspersorium of sacred liquids?) therefore seems more likely. However the Mes Aynak spoons lack the zoomorphic decoration of the handle, a common feature of nomadic tradition 32. In fact, the handle with its spiral pattern and the connection between the handle and the scoop of the Mes Aynak spoons have closer parallels in the tradition of metalworking in Bactria. A spoon found at Surkh Kotal 33,

The tools are another important category, which includes some objects that are not easy to assign to a specific class, in some cases due to their bad state of preservation. For two of these objects, 2.1.1 and 2.1.2, the closest resemblances are with some Central Asian pins dating to as early as the Bronze Age25. Compared to other pins, however, the two items from Mes Aynak have a special feature: one of the terminals ends in a small cup. The Mes Aynak pins are not an isolated case, as confirmed by an example of pin attributable to the Dandibaj–Begazy culture (10th–8th BC)26, quite similar to our specimens. The presence of the small cup suggests utilization of these objects as small spoons for tiny quantities of products. Or maybe, as suggested by M.E. Pottier in her study 27 of the material from burials28, both uses are compatible: pins

29

Pottier 1984, 30–31. L'uomo d'oro 1998, Cat. n°146. 31 Dvornichenko 1993, 151–154, Zhdanovskyi 1990, 107–112, Kruglikova 1966, Kovpanenko 1986, 90–92, Arbore Popescu et alii 1998, 234. One of the proposed hypotheses was that these spoons were used to mix liquids. Although the extreme fragility of the handle argues against this hypothesis. 32 See the example Anisimova et alii 2005, n° 94, 139. 33 Fussman and Guillaume 1990, Pl. V, n° 21. 30

24

Cambon et alii 2006, n° 134, 206. Pittman 1984, 48–49 26 Arbore Popescu et alii 1998, Cat. n°43. 27 Pottier 1984. 28 In her study Pottier analyses a series of objects found on the Kabul market and considered to be funerary objects. 25

32

F i r s t n o t e s o n a t r e as u r e f r o m M e s A yn a k dated to the Kushan period, is very similar to the Mes Aynak spoons.

The closest comparison for the Mes Aynak figure with the flower in his upraised hand can be found in a seal studied by P. Callieri35 (FIG. 10) who attributed it, relying on the fundamental work of Göbl of 196736, to a late Kidarite or more probably early Alkhan production 37. De facto, “Kidarites”, “Alkhan”, “Nezak” and “Genuine Hephthalites”38 were the four political entities comprising the more generic "Hunnish wave” that from the second half of the 4th century AD began to reach the Indo–Iranian borderlands. Regardless of the differences of strictly ethnic nature, the numismatic study put order in the various productions, outlining complex political scenarios in which different groups frequently overlapped. In addition to the Alkhan coinage, the Mes Aynak plate might add a new element to the theory of the power of the Alkhans in the Kabul area as well as in North–West of the Indian Subcontinent from the end of the 4th to the 6th century AD.

One of the most characteristic objects of the Mes Aynak hoard is the small bronze plate with decoration in low relief (2.1.7). It belongs to a class of similar dishes whose function has not yet been ascertained. Although the decoration is difficult to read, one can clearly observe that the central figure is a cross–legged seated male. Some further details are still visible and help to describe this figure: the right arm is raised and holds up a flower, while the skull displays a marked deformation (Fig. 10). Even if this figure clearly belongs to the so–called “Huna” iconography, it is not easy to define its exact geographical origin and chronology. Still, this object could in a way contribute to the ongoing debate on the transmission of Sasanid elements to Huna iconography, which tried to put order in the genal definition of “Huns” and achieve a more comprehensive understanding of the various groups that made up the “Huns”.

Fig. 11: Alkhan seal.

The discovery of the Mes Aynak hoard represents an important clue to their location. If we accept that their monetary production started at the end of the 4 th century, when they took possession of the mint of the Sasanian kings Shapur II (309–379) and Shapur III (383–388) in the “Kabul area”39, we still have no information on the location of the mines they controlled. The mine of Mes Aynak would perfectly fit this role, at least for the copper production. The cranial deformation displayed by the main figure in the decoration is of great help to establish the date of the plate. Numismatic studies suggest that this feature started in the third phase of Alkhan coinage, around the beginning of the 5th century40 when a shift took place from the simple imitation of Sasanian models to the use of features clearly belonging to Huna culture, like the skull

Fig. 10: Detail of a central figure.

Without repeating all the issues perfectly summed up by M. Alram and M. Pfisterer in their article on the Alkhan and Hephthalite coinage34, we think it is possible to find a more precise chronological indication for the Mes Aynak plate.

35

Callieri 2002, 121–141. Göbl 1967. 37 Callieri 2002, 131. 38 Alram and. Pfisterer 2010, 13. 39 Alram and. Pfisterer 2010, 15. 40 Alram and. Pfisterer 2010, 17–18. 36

34

Alram and Pfisterer 2010, 13–38.

33

BARBARA FATICONI deformation (FIG 11).41 The Mes Aynak plate, therefore, is unlikely to date from a period earlier than this change.

Recent stylistic studies47 indicate that the bowl can hardly be dated later than the 5th century AD. Even if we cannot completely exclude that the hand of Mes Aynak holds a flower, in fact both the hand underpinning the stem of a flower and the hand supporting the foot of a cup refer to a Bactrian tradition documented from the 5 th century onward.

Although the cross–legged seated figure with its raised arm bearing a flower is a theme used for centuries 42, the cranial deformation represents a clear terminus post quem, which provides a date to the entire hoard. If gold jewellery could suggest a date around the first century AD or earlier, the presence of the plate clearly indicates that the hoard could not have been stored before the beginning of the fifth century AD.

To conclude, the analysis of the individual pieces does not provide a precise date. Jewels, tools, and ornaments seem to belong to a generic Central Asian production, spanning several centuries. For sure, the hoard cannot have been deposited before the 5th century AD. Therefore, it can be inferred that at that time the Mes Aynak mine was still productive. Furthermore, the heterogeneous items in the hoard and its context suggest the existence of a workshop, possible to repair precious objects. The hypothesis of a workshop could shed light on the spatial organization of the site, especially once the results of the excavations still in progress have been published in detail.

Fig. 12: Alkhan Third Minting Phase.

A small bronze hand belonging to a small sculpture (3.1.1) seems to confirm this date. The existence of metal sculpture of small proportions was already known in Afghanistan, as the beautiful pieces belonging to the famous treasure of Begram (1th–2th century AD)43 indicate. The sequence of Eros, Herakles, horsemen at Begram44 evidently belongs to the Greco–Roman art tradition. In the case of the fragmentary hand of Mes Aynak, the classical model has to be discarded in favour of a Central Asian model. The motif of a hand supporting a cup, like that of the Mes Aynak sculpture, is not unique. One of the men with crossed legs on the famous Storogonov silver bowl in the Hermitage is also represented holding a cup with the thumb and index finger (Fig. 12)45. Its provenance and date are still being debated, although a Central Asian origin seems clear 46.

Fig. 13: ‘Storogonov’ Bowl.

41

Alram and. Pfisterer 2010, 17. See the fine example decorating a silver bowl dated to the 6th– th 8 centuries, Arbore Popescu et alii 1998, Cat. n°545. 43 The dating of the Begram treasure is still under discussion: Cambon et alii 2006, 95–100. 44 Cambon et alii 2006, cat. n° 218, 255; n° 220 256; n° 222, 257. 45 Callieri 2002, fig. 22, pag. 140. 46 Staviskii 1960, 67–71, Trever 1960, Azarpay 1981, 103. 42

47

Also the dating of this bawl is still under discussion: Callieri 2002, 126–127.

34

F i r s t n o t e s o n a t r e as u r e f r o m M e s A yn a k Bibliography ALLCHIN, F. R.; HAMMOND, N. (eds.) 1978, The Archaeology of Afghanistan from earliest times to the Timurid period, Cambridge.

DVORNICHENKO, V. V.; FEDOROV–DAVYDOV, G. A. 1993, “Sarmatskoe pogrebenie skeptukha I veka n.e.u selo Kosika Astrakhanskoy oblasti”, Vestnik Drevney Istorii, 3, 141–179 (non vidi).

– 1985, L'Archéologie de la Bactriane ancienne: actes du Colloque franco–soviétique. Dushanbe (URSS), 27 octobre–3 novembre 1982, Paris.

FUSSMAN, G.; LE BERRE, M. 1976, Monuments Bouddhiques de la Région de Caboul, I, Le Monastère de Gul Dara,. Paris.

ALRAM, M.; PFISTERER, M. 2010, “Alkhan and Hephtalite Coinage”, in M. ALRAM et al. (eds.). Coins, Art and Chronology II, The First Millennium C.E. in The Indo– Iranian Borderlans, Wien, 13–38.

FUSSMAN, G.; GUILLAUME, O. 1990, “Surkh Kotal en Bactriane. 2. Les monnaies, les petits objets”, in D. SCHLUMBERGER, M. LE BERRE & G. FUSSMAN (eds.), Mémoires de la Délégation archéologique française en Afghanistan XXXII, Paris.

ANISIMOVA, L.; BONOGRA G. L.; FRANCHI, C.; KARAVAEVA, L. M.; PLAKHOV, V. V. (eds.) 2005, I Tesori della Steppa di Astrakhan, Milano.

FUSSMAN, G.; MURAD, B.; OLLIVER, É. 2008a, “Monuments Bouddhiques de la Région de Caboul. II/1. Inventaire et descriptions (comprehensive inventory of Buddhist monuments in Kabul city and province with excavation reports)”, Mémoires de la Délégation archéologique française en Afghanistan XXII, Paris.

ARBORE POPESCU, G.; SILVI ANTONINI, C.; BAIPAKOV, K. (eds.) 1998, L’uomo d’oro. La cultura delle steppe del Kazakhstan dall’età del Bronzo alle migrazioni, Mantova. AZARPAY, G. 1981, Sogdian Painting, the pictorial epic in oriental art, Berkeley – Los Angeles – London.

– 2008b, “Monuments Bouddhiques de la Région de Caboul. II/2. Planches, résumes, index”, Mémoires de la Délégation archéologique française en Afghanistan XXII, Paris.

BALL, W. 1982, The Archaeological Gazetteer of Afghanistan: Catalogue des sites archéologiques d’Afghanistan, 2 vols, Paris.

GIGNOUX, P.; GYSELEN, R. 1982, Sceaux sasanides de diverses collections privées, Leuven.

– 2008, The Monuments of Afghanistan, History, Archaeology and Architecture, London.

GÖBL, R. 1967, Dokumente zur Geschichte der Iranischen Hunnen in Baktrien und Indien, 4 vols., Wiesbaden.

BOPEARACHCHI, O.; BOUSSAC, M. F. (eds.) 2005, Afghanistan, ancien carrefour entre l'Est et l'Ouest, Turnhout.

KOVPANENKO, G. T. 1986, Sarmatskoe pogrebenie I v.n.e.na juzhnom Buge, Kiev (non vidi).

CALLIERI, P. 1997, Seals and Sealings from the North– West of the Indian Subcontinent and Afghanistan (4th century BC – 11th century AD). Local, Indian, Sasanian, Graeco–Persian, Sogdian, Roman, Naples.

KRUGLIKOVA, I. T. 1966, “Raskopki Gorgippii”, in Ktatkie Soobschenija Instituta Arkheologii, fasc.108 (non vidi).

– 1999, “Huns in Afghanistan and the North–West of the Indian Subcontinent: the Glyptic Evidence”, in M. ALRAM et al. (eds.), Coins, Art, and Chronology. Essays on the pre–Islamic History of the Indo–Iranian Borderlands, Vienna, 277–291.

MASSOUDI, O. K. (ed.) 2011, Mes Aynak. New Excavations in Afghanisthan, Kaboul. PITTMAN, H. 1984, Art of the Bronze Age: Southeastern Iran, Western Central Asia and the Indus Valley, Metropolitan Museum of Art.

– 2002, “The Bactrian Seal of Khingila”, Silk Road Art and Archaeology 8, 121–141. CAMBON, P.; JARRIGE, J.–F.; BERNARD, P.; SCHILTZ, V. (eds.) 2006, Afghanistan, les trésors retrouvés, Paris.

POPESCU, G. A.; SILVI ANTONINI, C.; BAIPAKOV, K. (eds.) 1998, L'uomo d'oro: la cultura delle steppe del Kazakhstan dall'età del bronzo alle grandi migrazioni, Milano.

DALTON, O. M. 1964, The treasure of the Oxus with other examples of early oriental metal–work, London.

POTTIER, M. E. 1984, Matèriel Funéraire de la Bactriane Méridionale de l’Age du Bronze, Paris.

DE MEYER, L.; HAERINCK, E. (eds.) 1989, Archaeologia Iranica et Orientalis, Miscellanea in honorem Louis Vanden Berghe, Gent.

RAPIN, C. 1992, Fouilles d’Aï Khanoum VIII: La trésorerie du Palais hellénistique d’Aï Khanoum, Paris.

35

BARBARA FATICONI SARIANIDI, V. 1985, Bactrian Gold from the Excavations of the Tillya–Tepe Necropolis in Northern Afghanistan, Leningrad.

ZHDANOVSKYI, A. M. 1990, Npvye pogrebenija kochevnikov sarmotskogo kruga iz Zakuban’ja, in Drevnie pamjatniki Kubani, Krasnodar (non vidi).

SCHILTZ, V. 1994, Les Schythes et les nomades des Steppes, Paris.

ZEYMAL E. 1999, “Tillya–Tepe within the context of the Kushan Chronology”, in M. ALRAMET al. (ed.), Coins, Art and Chronology, Wien, 239–244.

TREVER, K. 1960, Novoe “sasanidskoe” bljudče Ermitaža, in Issledovanija po Istorii Kul’tury Narodov Vostoka, Leningrad (non vidi).

36

The Second after the King and Achaemenid Bactria on Classical Sources MANEL GARCÍA SÁNCHEZ Universidad de Barcelona–CEIPAC ABSTRACT The government of the Achaemenid Satrapy of Bactria is frequently associated in Classical sources with the Second after the King. Although this relationship did not happen in all the cases of succession to the Achaemenid throne, there is no doubt that the Bactrian government considered it valuable and important both for the stability of the Empire and as a reward for the loser in the succession struggle to the Achaemenid throne. KEYWORDS Achaemenid succession – Achaemenid Bactria – Achaemenid Kingship – Crown Princes – porphyrogenesis Beyond the tradition that made of Zoroaster the King of the Bactrians, Rege Bactrianorum, qui primus dicitur artes magicas invenisse (Justinus 1.1.9), Classical sources sometimes relate the Satrapy of Bactria –the Persian Satrapy included Sogdiana as well (Briant 1984, 71; Briant 1996, 403 s.)– along with the princes of the Achaemenid royal family and especially with the ruled out prince in the succession, the second in line to the throne, sometimes appointed in the sources as “the second after the King” (ὁ δεύτερος μετὰ τὸν Βασιλέα; secundus a rege) (Volkmann 1937–8; Benveniste 1966, 51–65; García Sánchez 2005, 228 s.; 2009, 166 s.). The Greek and Roman authors, Herodotus, Pompeius Trogus–Justin, Arrian or Plutarch, pay special attention to the fratricidal struggle and harem royal intrigues among the successors to the throne in the Achaemenid Empire (Shahbazi 1993; García Sánchez 2005; 2009, 155–175). It is in this context where we might find some explicit references to the reward for the prince who lost the succession dispute: the offer of the government of Bactria as a compensation for the damage done after not having been chosen as a successor of the Great King (Sancisi–Weerdenburg 1980, 122–139; Briant 1984, 69–80). In this particular case and in the Achaemenid Bactria on the whole, our documentary dossier takes the sources from the Classical authors (Briant 1984, 10), whose mental picture of Bactria, natio antehac belatrix et potentissima, Persisque semper infesta, antequam circumsitos populos omnes ad dicionem gentilitatemque traheret nominis sui (Ammianus Marcellinus 23.6.56), is partly une vision égéocentrique and quasi fantasmatique (Briant 1984, 64 s.). On the one hand, archaeologists can not dare to date a satisfactory synthesis due to the dispersed character of the discoveries or to the lack of

them; and, on the other hand, these discoveries have not always been published or they are not easily accessible (Sancisi–Weerdenburg 1980, 131–134; Briant 1984, 102; Bactriane 1985, 241 ss.; Holt 1989, 27 ss., 33; Vogelsang 1992, 270–280; Genito 1996; Gardin 1998; Francfort 2005, 316). An alabastron of Bactria with the name Artaxerxes in cuneiform and hieroglyphic writing (Schmitt 2001) could be, as Francfort (2005, 324) points out, a royal gift for a local dignitary, a recycled object that may have arrived there accidentally or as part of booty. It could confirm the postulate of the Persian miracle about the Achaemenid administration (Briant 1984, 9) the archaeological remains of irrigation canals or qânats (PTF 1947: 59, 62; Briant 1982a, 486–488; 1984, 49–55, 92; 1996, 87 s.; Francfort & Lecomte 2002), that show again the Achaemenid interest in the farming exploitation of the Empire’s satrapies, in Bactria’s case for the growing of lucerne (medicago sativa), that was known in Greece like βοτάνη μηδική or μηδική πόα (Strabo 11.13.7), a fodder plant for the pastures with which to graze the large Bactrian equine population (Curtius 7.4.30) and the famous Bactrian camels (camelus bactrianus) (Gabrielli 2006, 25; Francfort 2005, 325; Tuplin 2010; Llewellyn–Jones 2013, 85 s.). The grain’s production, vineyards and fruit trees (Strabo 11.11.1), along with the pistachio nuts (pistacia vera), brought to the west from Bactria after Alexander’s conquest (Theophrastos, HP 4.4.7), proved the fertility of the land, a fact which was later stressed by Curtius –Bactrianae terrae multiplex et varia natura est (Curtius 7.4.26–27; Strabo 11.11.1: πολλὴ δ’ἐστὶ καὶ πάμφορος πλὴν ἐλαίου... διὰ τὴν ἀρετὴν τῆς χώρας... τῆς συμπάσης Ἀριανῆς πρόσχημα εἶναι τὴν Βακτριανήν; Ammianus Marcellinus 23.6.56: humi gignentium fertiles, et pecus quod illic per campestria loca vescitur et montana– next to the desert

MANEL GARCÍA SÁNCHEZ which was part of his territory –terrae steriles harenae (Curtius 7.5.1–16). We know about the Bactrian barbarian customs (Strabo 11.11.3), about Bactrian silver and gold (DSf § 36) like the treasure of the Oxus (Barnett 1968; Francfort 1988, 185; Briant 1996, 411), the lapis lazuli in Sogdiana (DSf § 37) (Tarn 2010, 103), or some archaeological remains of ancient royal Persian palaces or residences, sacred architectural framework and fortresses (Francfort 1985; Sarianidi 1985, 424; Gardin 1998, 143; Francfort 2005, 334–6), or pottery (Lyonnet 1997; Gardin 1998, 24; Francfort 2005, 318–322), although the more illuminating data about the possible dating of the Achaemenid Bactrian administration might be the Aramaic documents which are connected with the court of the Satrap of Bactria of the fourth century BC, Akhvamazda, and with its governor (ὕπαρχος) Bagavant, at least through part of the period of Artaxerxes III and edited by Naveh and Shaked (Shaked 2004; Naveh & Shaked 2012, 17). Finally, we have the depicted Bactrians on the staircases of the apadana and between the throne– bearers of the Throne Hall of Persepolis, on the Darius Statue from Susa and on royal Achaemenid tombs, but maybe the relationship between the iconographies and their ethnic historical reality was not accurate and perhaps only an idealization (Vogelsang 1992, 138 ss.; Genito 1996, 409 s.). However apart from some hypothesis that might allow us to formulate very carefully and provisionally the nature of these Aramaic documents of Bactria; we have little information, apart from the information coming from Classical sources, about the period that lasts between the connection of Bactria and the question of the Achaemenid succession.

Briant has expressed his doubts about the passage of Ctesias, especially because these territories were not in 522 BC under the same command and it was more likely that Tanyoxartes received the government of Bactria and some complementary revenues from outer regions to the Satrapy (Briant 1984, 71). Somebody (Boardman 1970, fig. 877) used to defend a gem showing three crowned Kings with the same type of crown the hypothesis of the possible existence of a triarchy made up by Cyrus, Cambyses, as Babylon’s viceroy, and Tanyoxartes of Bactria. Even someone pointed out that the sister of Atossa and daughter of Cyrus, Roxana (Ctesias FGrHist. 688, F 13, 14), bore the most famous royal name in Bactria’s history (Holt 1989, 40). In the case of Darius I some scholars have related their ancestors to the former Bactrian rulers who also were the early patrons of Zoroaster and they have even provided evidence to prove that this Satrapy did not rebel against the Achaemenid when he came to the throne or that the name of Bactria’s satrap given in the Behistun Inscription (DB § 38) was Dadaršiš, unknown in Classical sources (Sancisi–Weerdenburg 1980, 129; Vogelsang 1992, 126 s.) although some scholars have interpreted it as the same name of Darius or they have even conjectured that Darius’ father, Hystaspes, was a Bactrian because Vishtâspa is a throne name frequently connected with Bactriana (Holt 1989, 39–41; Vogelsang 1992, 125 s.). We also know that the Irdabanuš which appears in the Fortification Tablets of Persepolis (PFT 1287, 1555 Hallock), Satrap in Bactria 500/499 BC, may be the brother of Darius I and the uncle of Xerxes (Benveniste 1966, 107 s.; Lewis 1977, 19, n. 96; Sancisi– Weerdenburg 1980, 124).

The first time that this connection appears in our sources, beyond the legend of Semiramis (Briant 1984, 25–33) or Zoroaster (Ctesias FGrHist. 688, F 1; Diodorus Siculus 2.1–19; Herodotus 1.184; Strabo 15.2.5; 16.1.2; Xenophon, Cyr. 1.5.2; Justinus 1.1.9) (Tarn 2010, 114; Holt 1989, 34–37), it goes back to the rise of the Achaemenid Empire, the conquest of Bactria by Cyrus (Ctesias, FGrHist. 688, F 9, 2), the succession of Cyrus the Great by Cambyses and the compensation for Tanyoxartes (Bardiya in the Behistun Inscription) with the government of Bactria (σατράπης of Media, Armenia and Cadusians in Xenophon, Cyr. 8.7.9–11). A Ctesias’ passage (Ctesias, FGrHist. 688, F 9, 8) reports us that Tanyoxartes obtained the government of the Satrapy of Bactria (δεσπότης Βακτρίων), Chorasmia, Parthia and Carmania, duty–free (ἀτελεῖς ἔχειν), either while Cambyses was Babylon’s viceroy and was still living in Cyrus, or when Cambyses was crowned as the new Great King after his father’s death in the battlefield and after Cyrus had conquered most of modern Afghanistan and Central Asia, known nowadays as Uzbekistan, Turkmenistan and Tajikistan, and had fought against the Sakai Massagetae (Herodotus 1.201–216; Strabo 11.8.6), the Scythian people from the Empire’s frontier in the North of Bactria, Sogdiana and beyond the Jaxartes river (Syr Darya) (Vogelsang 1992, 181–189; Briant 1996, 49 s.) and while he attempted to extend the territory of his empire eastward beyond the Oxus river (Amu Darya). P.

Among the sons of Darius and Atossa we find again the association between the Satrapy of Bactria and the Crown Princes. Nevertheless, the case is more complex as in the struggle for the succession after Darius I’s death we find in Classical authors the names of some other important princes (García Sánchez 2005, 228 s.; 2009, 165–170): Artobazanes, Darius’ eldest son in Herodotus (πρεσβύτατος, Herodotus 7.2.3), Ariamenes in Plutarch (Plutarchus, Reg. et. imp. apophth. = Moralia 173B–C; πρεσβύτατος, Plutarchus, De amore fraterno 18 = Moralia 488D–F), and even not always the loser won the government of the same Satrapy. Our main source of the documentary dossier is Plutarch because he is the one who mentions Bactria as the Satrapy given to Ariamenes (Ξέρξες ἐκείνῳ τὴν δευτέραν μεθ’ἑαυτὸν ἔδωκε τάξιν, Plutarchus, Reg. et. imp. apophth. = Moralia 173B–C), although other sources assign him the government of Media (Plutarchus, De amore fraterno 18 = Moralia 488 D–F; cf. Themistius, Or. VI, 74c). To make things even more confusing Plutarch says that Ariamenes was appointed “the Second after the King” (μέγιστος, Plutarchus, Reg. et. imp. apophth. = Moralia 173 B–C; δεύτερος μεθ’ ἑαυτόν,.. τὰ δευτερεῖα, Plutarchus, De amore fraterno 18 = Moralia 488 D–F; cf. a. p. pasā tanūm maθišta, XPf § 4), although we cannot clarify if this title was given to the Crown Prince, the greatest after

38

T h e S e c o n d A f t e r t he K i n g a n d A c h a e m e n i d B a c t r i a o n C l a s s i c a l S o u r c e s the king, the successor to the throne and exceptionally the first son born in purple or porphyrogenesis (Wisehöfer 1994, 54), as it is revealed in the Xerxes’ Persepolis Inscription (XPf § 4) or to the Second after the King, a rank second or the second in line to the throne (García Sánchez 2005, 228; 2009, 165).

Persian Kings are crazy about women and the constraints of desire, typical anyway of the Classical sources (Sancisi–Weerdenburg 1980, 267; García Sánchez 2009, 177–203). Xerxes appointed his son Hystaspes as satrap of Bactria too (Diodorus Siculus 11.69.2; Artapanus in Ctesias, FGrHist. 688, F 14, 35) in the place of his murdered brother Masistes or Ariamenes. But after Xerxes’ death, Hystaspes disputed his right to the Persian throne to his brother Artaxerxes I. Maybe Hystaspes won the satrapy of Bactriana because he was not chosen as the Crown Prince instead of the firstborn Darius, since Justin (Justinus 3.1.5–9) is quite clear on this matter when he asserts that Darius was designated as the successor of (quo maturius regno potiretur). Hystaspes, probably, he was not very happy with his luck and, after the assassination of Darius, who was accused of treason, he rebelled against his brother Araxerxes I, a brother crowned as the new Great King but maybe in favor of Hystaspes, Xerxes’ third son, who had no necessarily inheritance rights over him (Briant 1996, 587; García Sánchez 2005, 232; 2009, 170). It is in this context that we find Ctesias’ passage about the insurrection of Artapanus, satrap of Bactria too (Ctesias, FGrHist. 688, F 14, 35) in 465 BC, but there might be nothing else but a divergent tradition (Briant 1984, 76). Maybe Ctesias only uses the name of Artapanus to make a comparison (ὁ σατράπης, ἄλλος Ἀρτάπανος) (Sancisi– Weerdenburg 1980, 127; Lenfant 2004, 266, n. 452), we do not know if it is either his relationship with the Achaemenids or simply the briefness of Ctesias’ note which does not allow us to go further away.

We find Bactria again related to the Crown Prince of the royal Achaemenid house with another Darius I’ son, Masistes, appointed as satrap of Bactria (Herodotus 9.113) and it is stated that he came with Xerxes during the invasion of Greece (Herodotus 7.82), although it was their brother Hystaspes, the commander of the Bactrians and Sacae (Herodotus 7.64) (Burn 1962, 333). Masistes started a rebellion from Bactria against his brother Xerxes, already crowned as the Great King (Dandamaev 1989, 85; Briant 1982b, 207), but to some scholars the name Masistes brings back the Old Persian word maθišta which appears in the Persepolis’ harem Inscription and when Xerxes stated how his father Darius I appointed him as his heir: Other sons of Darius were there, but unto Ahuramazda was the desire Darius my father made me the greatest after himself (maθišta) (XPf § 4. 28–32 KENT; cf. PF–NN 1657; Calmeyer 1976, 83; Llewellyn– Jones 2013, 18). This case is at once confusing and very interesting. For some scholars, Masistes’ story resembles a novella and should not be regarded as an historical reporting of events (Sancisi–Weerdenburg 1980, 48 ss.; 268; cf. Wiesehöfer 1994, 86; Llewellyn–Jones 2013, 139), and Masistes could be the same person or the double of the Plutarch’s Ariamenes and the story a popular rendering of a revolt against the throne by Xerxes’ full brother Masistes (Sancisi–Weerdenburg 1980, 126, 268). If this interpretation was correct and Ariamenes and Masistes were the same person we could complement the passage of Plutarch, the passage of Herodotus (Herodotus 7.82; 9.107) and Xerxes’ Persepolis Inscription: Ariamenes was appointed as “the Second after the King” (μέγιστος, Plutarchus, De amore fraterno 18 = Moralia 488 E–F) and Masistes could be the Greek translation of a title or noble rank by Herodotus or his sources, maθišta, meaning “the Second after the throne”, “the greatest after the King” or maybe a temporary vice–royalty (Sancisi–Weerdenburg 1980, 268; García Sánchez 2005, 229; 2009, 167 s.; Llewellyn–Jones 2013, 17). Nevertheless, the problem in this case is that in the Xerxes’ Persepolis Inscription maθišta is a term to name the Crown Prince but not the Second after the King as a brother of the Great King or the second in line to the throne and never as a satrap of Bactriana or viceroy (Briant 1996, 540; García Sánchez 2005, 228 f.; 2009, 164–168). Another not easy problem to answer is this: if Ariamenes and Masistes were really two different Xerxes’ brothers: when and why Ariamenes was replaced by Masistes as Bactrian satrap? Finally, we do not know the real motive that drove Masistes to rebel against his brother coming down from Bactria and the Herodotus’ Masistes story (Herodotus 9.108–113), but the fact that Xerxes was crazy for his brother’s wife and daughter and that this was the real cause of the revolt gives the story a commonplace of an oriental harem appearance where

Bactria disappears of the struggles of succession between Achaemenid princes until the end of the dynasty and the treachery of οἰκειότης (Arrianus, An. 3.21.5) Darius III, Bessus, the satrap of Bactriana (Diodorus Siculus 17.73.2, Arrianus, An. 3.21.5; Naved & Shaked 2012, C1), over Darius III, and that haud sane aequo animo in secundo se contiens gradu (Curtius 4.6.4). Finally defeated in Gaugamela, Darius III decided to find safety in Bactria (Diodorus Siculus 17.73.2), but he was murdered by Bessus before he could be helped by Bactrian, Sogdian and Scythian supporters (Arrianus, An. 3.21.4–5; 3.21.10; Diodorus Siculus 17.73.2; Curtius 6.6.16). Once again Bactria was the key satrapy for the ambitions of usurpers (Briant 1982b, 209). P. Briant also considers the use of the term οἰκειότης by Arrian (Arrianus, An. 3.21.5) we are not allowed to conclude with absolute certainty that Bessus was a close relative of the Achaemenid family (Briant 1984, 7; Briant 1996, 801). It is true that this term can be related to a friend or simply to a close person of Darius III, but its semantic field covers the relationships too (Chantraine s.v.) and then Bessus could have also perfectly been a close relative of Darius III who got the government of the Bactrian satrapy for his position in secundo (Curtius 4.6.4). We also have concurrently the case of another close to the Great King, the royal kinsman Hystaspes (propinquus, Curtius 6.2.7; Naved & Shaked 2012, C2, 1), κάρανος or an Achaemenid military official that maybe conducted operations against the warlike

39

MANEL GARCÍA SÁNCHEZ nomadic Scythian incursions on the margins of Bactria and Sogdiana (Hyland 2013).

we should have relativiser ce qu’on appele fréquentment l’irrédentisme bactrien: il paraït risqué –pour parler en litote– de conclure que la Bactriane était choisie à cet effet de préférence à d’autres satrapies (Briant 1984, 75) and the nomination of a King’s son as satrap of Bactria only it is due to avoid the inheritance struggle and à désamorcer le conflit qui s’annonce avec l’héritier choisi and it provides de disposer des moyens financiers propres à lui assurer un train de vie princier et surtout d’utiliser les forces militaires de sa satrapie pour entrer en lice dans la lutte dynastique (Briant 1984, 76). Really, P. Briant is right when he relativizes the Bactrian irredentism and of course no all satraps’ insurrections in Bactria respond to a secessionist will of the satrapy, but that il s’agit bien plutôt d’objectifs «achéménides» and neither the power of the Great King nor the unity of the Empire or the control of Central Asia were at risk par une révolte authentiquement «bactrienne» (Briant 1984, 77). Nevertheless, it is true too that Bactria was a prominent assignment to the ruled out Princes in the inheritance struggle and in any case the fact that several brothers of Great Kings ruled in this satrapy shows its great importance for their political career and for the security of the Empire too, among other things to defend the Eastern frontier and to stem the raids of the wa r l i k e S c y t h i a n no mad s (B ri a nt 1 9 8 2 b , 205; 1984, 71–74; 1996, 914), Bactriana was always a buffer between the sedentary and the nomadic world (Strabo 11.11.3; Francfort 1990; Gardin 1990; Genito 1996, 407), and if they had great dynastic aspirations for their political ambitions and supports too, but definitely and unlike the satrap Diodotus I in the second half of the third century BC, he formally proclaimed himself King of the Seleucid Bactria (Holt 1999), to be only crowned as the Great Kings of the Achaemenid Empire.

Alexander appointed the Persian Artabazus as satrap of Bactria and got married to Roxana, according to Arrian (Arrianus, An. 7.4.4) a Bactrian princess, the daughter of the ὕπαρχος Oxyartes, praetor Bactrianorum (Curtius 9.8.10; Strabo 11.11.4), a political marriage if we have in mind the hard campaigns of the conquest of Bactria and Sogdiana (Briant 1982b, 232). P. Briant is somewhat skeptical about the relation between the Bactrian’s satrap and the close relatives to the Great King and he considers utile de dissiper quelques illusions sur le caractère privilégié des rapports des satrapes bactriens avec le Grand Roi (Briant 1984, 75) and the author calls that phenomenon myopie bactrienne. For this French scholar the relation between Bactriana and the Second after the King is not very different from the government of the other satrapies of the Empire for the grande famille achéménide: the son of Darius I and the brother of Xerxes Achemenes in Egypt (Herodotus 7.7), the prince of the royal family Aršama in Egypt too (Grelot 1972, 300), the half–brother of Darius I Artaphernes in Sardes (Herodotus 5.25.30) or Cyrus the Younger, the brother of Artaxerxes II and satrap of Lydia and Caria and κάρανος of troops of Lydia, Phrygia and Cappadocia (Xenophon, An. 1.1. 2; Xenophon, HG 1.4.3) (Petit 1983), Artairos, brother of Artaxerxes I and as satrap of Babylon (Ctesias, FGrHist. 688, F 14, 41; Stolper 1974, 164), Hystaspes, Darius’ father, as satrap of Parthia or Hyrcania (DB § 3.36; Herodotus 1.209; 3. 70–71), Ochus, the future Great King Darius II as satrap of Hyrcania (Ctesias, FGrHist. 688, F 15, 47), Codoman, the future Great King Darius III (Justinus 10.3.4) in Armenia and finally Bessus, the member of Darius’ own household (οἰκειότης) (Diodorus Siculus 17.73.1; Arrianus, An. 3.21.5) in Bactria. The conclusion would be obvious and

40

T h e S e c o n d A f t e r t he K i n g a n d A c h a e m e n i d B a c t r i a o n C l a s s i c a l S o u r c e s Bibliography Bactriane 1985, L'archéologie de la Bactriane ancienne : actes du colloque franco–soviétique, Dushanbe (U.R.S.S.), 27 octobre–3 novembre 1982, Paris.

achéménide”, Annales Histoire Sciences sociales 57.3, 625–663. GABRIELLI, M. 2006, achéménide, Istanbul.

BARNETT, R. D. 1968, “The art of Bactria and the treasure of the Oxus”, Iranica Antiqua 8, 34–54.

Le

cheval

dans

l’empire

GARCÍA SÁNCHEZ, M. 2005, “La figura del sucesor del Gran Rey en la Persia aqueménida”, in V. ALONSO TRONCOSO (ed.), ΔΙΑΔΟΧΟΣ ΤΗΣ ΒΑΣΙΛΕΙΑΣ. La figura del sucesor en la realeza helenística, Madrid, 223–239.

BENVENISTE, E. 1966, Titres et noms propres en Iran ancien, Paris. BOARDMAN, J. 1970, Greek Gems and Finger Rings, London.

– 2009, El Gran Rey de Persia. Formas de representación de la alteridad persa en el imaginario griego, Barcelona.

BRIANT, P. 1982a, “Appareils d’état et développement des forces productives au Moyen–Orient ancien: le cas de l’Empire achéménide”, in Rois, tributs et paysans. Paris, 475–489.

GARDIN, J.–C. 1990, “Indicateurs archéologiques du nomadisme : études de cas de Bactriane”, in H.–P. FRANCFORT (ed.), Nomades et sédentaires en Asie Centrale : apports de l'archéologie et de l'ethnologie : Actes du colloque franco–soviétique, Alma Ata (Kazakhstan) 17–26 octobre 1987, Paris, 131–139.

– 1982b, État et pasteurs au Moyen–Orient ancien, Paris. – 1984, L’Asie centrale et les royaumes proche–orientaux du premier millénaire (c. VIIIe–IVe siècles avant notre ère), Paris.

– 1998, Prospections archéologiques en Bactriane orientale (1974–1978). Volume 3 Description des sites et notes de synthèse, Paris.

– 1996, Histoire de l'Empire Perse. De Cyrus à Alexandre, Paris.

GENITO, B. 1996, “The Iranian Empires and Central Asia: an Archaeological Perspective”, in Convegno internazionale sul tema: la Persia e l'Asia centrale da Alessandro al X Secolo (Roma, 9–12 novembre 1994), Roma.

BURN, A. R. 1962, Persia and the Greeks, Stanford. CALMEYER, P. 1976, “Zur Genese Altiranischer Motive, V: Synarchie”, Archäologische Mitteilungen aus Iran 9, 63–95.

GRELOT, P. 1972, Documents araméens d’Egypte, Paris.

DANDAMAEV, M. A. 1989, A Political History of the Achaemenid Empire, Leiden.

HALLOCK, R. T. 1969, The Persepolis Fortifications Tablets, Chicago.

FRANCFORT, H.–P. 1985, “Fortifications et sociétés en Asie centrale protohistorique”, in J.–L. HUOT, M. YON & Y. CALVET (eds), De l‘Indus aux Balkans. Recueil à la mémoire de Jean Deshayes, Paris, 379–388.

– 1972. The Evidence of the Persepolis Tablets, Cambridge. HOLT, F. L. 1989, Alexander the Great and Bactria, Leiden.

FRANCFORT, H.–P. 1998, “Central Asia and Eastern Iran”, in J. BOARDMAN, N. HAMMOND, D. LEWIS & M. OSTWALD (eds), The Cambrige Ancient History, 2ed., vol. IV, Cambridge/New York/New Rochelle/Melbourne/Sidney, 165–193.

– 1999. Thundering Zeus. The Making of Hellenistic Bactria, Berkeley/Los Angeles. HYLAND, J. 2013, “Vishtaspa krny: an Achaemenid military official in 4th–century Bactria”, Arta 2013.002, 1–7.

FRANCFORT, H.–P. 2005, “Asie centrale”, in P. BRIANT & R. BOUCHARLAT (eds), L’archéologie de l’empire achéménide: nouvelles recherches, Paris, 313–352.

LENFANT, D. 2004, Ctésias de Cnide. La Perse, l’Inde, autres fragments, Paris.

– (ed.) 1990. Nomades et sédentaires en Asie Centrale : apports de l'archéologie et de l'ethnologie : Actes du colloque franco–soviétique, Alma Ata (Kazakhstan) 17– 26 octobre 1987, Paris.

LEWIS, D. M. 1977, Sparta and Persia, Leiden. LLEWELLYN–JONES, LL. 2013, King and Court in Ancient Persia 559 to 331 BCE, Edinburgh.

FRANCFORT, H.–P.; LECOMTE, O. 2002, “Irrigation et société en Asie centrale des origines à l’époque

41

MANEL GARCÍA SÁNCHEZ Bactriane et son gouverneur. Documents araméens du IV e s. avant notre ère provenant de Bactriane, Paris.

LYONNET, B. 1997, Prospections archéologiques en Bactriane orientale (1974–1978). Volume 2 Céramique et peuplement du chalcolithique à la conquête arabe, Paris.

STOLPER, M. W. 1974, Management and politics in later Achaemenid Babylonia: new texts from the Murašū Archive, Michigan.

NAVEH, J.; SHAKED, S. 2012, Aramaic Documents from Ancient Bactria (Fourth Century BCE) from The Khalili Collections, London.

TARN, W. W. 2010 [19381], The Greeks in Bactria and India, Cambridge

PETIT, TH. 1983, “Káranos. Étude d’une fonction militaire sous la dynastie perse achéménide”, Études Classiques 51, 34–55.

TUPLIN C. 2010, “All the King’s Horses: In Search of Achaemenid Cavalary”, in M. TRUNDLE & G. FAGAN (eds), New Perspectives on Ancient Warfare, Leiden, 101–182.

SANCISI–WEERDENBURG, H. 1980, Yaunā en Persai. Grieken en Perzen in een ander Perspectief, Groningen. SARIANIDI, V. 1985, “Monumental Architecture of Bactria”, in J.–L. HUOT, M. YON & Y. CALVET (eds), De l‘Indus aux Balkans. Recueil à la mémoire de Jean Deshayes, Paris, 417–432.

VOGELSANG, W. J. 1992, The Rise and Organisation of the Achaemenid Empire. The Eastern Iranian Evidence, Leiden/New York/Cologne. VOLKMANN, H. 1937–8, “Der Zweite nach dem König”, Philologus 92, 285–316.

SCHMIDT, R. 2001, “Eine weitere alabster–Vase mit Artaxerxes –Inschrift”, Archäologische Mitteilungen aus Iran und Turan 33, 191–201.

WIESEHÖFER, J. Zürich/München

SHAHBAZI, A. SH. 1993, “Crown Prince”, Encyclopaedia Iranica 6, 430–432.SHAKED, S. 2004, Le satrape de

42

1994,

Das

antike

Persien,

Achaemenid Royal Women and Alexander CLÀUDIA ZARAGOZÀ SERRANO Universitat Autònoma de Barcelona ABSTRACT In his Asiatic campaign Alexander the Great came across a lot of countries and established relationship with their people, in most cases as their new ruler. However, his relationship with the Royal Persian Women was special. This group of captives enjoyed a different treatment from king Alexander, who took care of them and protected their royal status, even after Darius’ III death. In the present paper we want to see the role played by Achaemenid Royal Women in the court and during the campaign, where they were part of the expedition, and the relationship between Alexander and them, especially with Sisigambis, Darius’ III mother. KEYWORDS Alexander the Great – Sisigambis – Achaemenid Court – Persian Women – Ancient Macedonia Alexander the Great was the first western king who arrived to the eastern world confines, where nobody had arrived yet. In his travel, Alexander met a lot of people of the different tribes of Asia, but the most important people with whom Alexander established some kind of relationship was the Persian. It is due to the importance of this empire and the persecution that Alexander carried out toward the Great King, Darius III. The most curious thing without a doubt is the importance that sources give to a special group of Persian people: women. The Macedonian king established an important relationship with Persian women, above all, the Royal Family ones, women that turned into important persons in Alexander’s life. In the present paper, we want to show the relationship between Alexander the Great and the Persian women and evince the importance of that group of people, who sometimes are underestimate and obviated by the scholars. First, we will make a revision of the Royal Persian Women, focused in the Achaemenid Family, and the role they played in the monarchy to understand the paper that some of them took when Alexander appeared in their lives. Then, and being the largest part of this paper we will explain the episodes whose protagonists are the Persian Woman, the reaction of Alexander and the most important thing, how the Macedonian king acted towards them. We will see how they turned into the most important feminine figure in the life of Alexander the Great (except Olympias, the king’s mother) and staged some of the most curious episodes. The monarchies ruled by the Achaemenids were organized with a strong political structure, with counselors, advisers and officials under the King’s orders. The King’s relatives and friends and the Persian nobles usually held the most important positions, as King’s personal counselors and advisers. The Persian territory

was organized by satrapies, obviously under the King supervision. It is difficult to think that, in an Empire with a so complicated and organized structure women didn’t have their own structure and organization. The Persian women rank was as elevated as closer was his relationship with the King, so the Royal Persian women had a prominent paper. There are three titles that had the highest level: the King’s mother, wife and daughter. The Greek sources tell us about the Persian women. They give us two terms for the first figure we will talk about, the King’s mother, also called the mother queen. The terms are μήτηρ βασιλέωσ1 and βασίλεια2. It creates a problem: what of the two terms reflects the Persian way to refer to the queen mother? Macurdy attended it, and she concluded that the second term implies a political charge, because βασίλεια not only describes the relationship between that woman and the king but also a political paper played by the woman3. If this term was the Persian one, it implied the importance of the queen mother figure in the Persian court. Nevertheless, βασίλεια is a term rarely used by the Greek sources. The first mention appears in an inscription after 306 B.C. referring to Phila, Demetrius Poliorcetes’ wife4. In the same way Amestris, Oxiartes’ daughter and Darius III’s niece, who became Dionisus of Pontic Heraclea’s wife and Lisimacus’ wife later, appears in another inscription like ΑΜΑΣΤΡΙΟΣ ΒΑΣΙΛΙΣΣΗΣ. About this inscription, Carney says that in this case the use of the term would be related with the increment of the utilization by the Alexander’s successors5. What meant βασιλισσα and who carried it? 1

Aesch. Pers. 151. Aesch. Per. 152. 3 Macurdy 1982. 4 Plu. Demetr. 31, 5. 5 Carney 1991, 156. 2

CLÀUDIA ZARAGOZÀ Drew says that this term was used to refer to King’s mother, wife and daughter6.

but in these cases the proper name appears, like in the case of Arystone18. The last term is the one used for the king’s daughter. The elamite term is sunki pakri, and it would correspond to the babilonian term mārat ŝarri, used in neo–babilonian documents to refer to Xerxes’ daughter (486 BC). If the king’s daughter was married to a king, she appears in the texts with her proper name, like in Arystone’s, Parysatis’ or Stateiras’ cases.

However, Brosius says that βασιλισσα means “royal women” in general, and not only “queen” and all the connotations that the term implies. In the Persian case, Brosius opines that the term was used by the King’s mother and wife. The term μητηρ βασιλέωσ is usually used by the Greek Sources to refer to a biological relationship like in the Amestris case, Artaxerxes I’s mother7, but also in cases of non biological relationship, like in Sisigambis’ case, Darius III’s mother, or Ada of Caria8, two women who “adopted” Alexander and were called “mother” by the King. In these two cases, the use of this term implies not only the establishment of a positive political relationshipfor Alexander being adopted by the queen, but also the acknowledgment of the new ruler by the mother of the defeated ruler 9. We have other no Greek terms to refer to the king’s mother before the Achaemenid period that demonstrate that the position of the queen mother was recognized for a long time, like amma in ancient elamite or ummi sari in the Nabonidus Chronicle10. The Persian would have followed this ancient political structure giving the queen mother an special rank in the Royal Court, even higher than the king’s wife one11.

The Persian Court wasn’t only formed by the Royal Family’s women. The king’s concubines appear in many texts. The Greek sources refer to them with the term μαλλακαι. However, it’s difficult to compare this term and its significance in the Greek society and in the Persian one as well as the figure of the concubine itself. Plutarch says that Darius III’19 had 365 concubines, and Heraklides tell us their functions: go along with the king in the royal hunting, sleep with him and play and sing music for him. The importance of this women in the court has been studied by Olmstead20, and other scholars like Schmidt, who suggested that the called “Southern Building” of Xerxe’s Palace would be the concubine’s house 21, theory rejected by Cool Root22. Brosius suggests that the royal concubines were selected by the king among the best foreign women, that’s why they couldn’t be the king’s wife. She supports her argument with Herodotus, who tell us that Cambyses, in the Egypt’s invasion, requested the king Amases’ daughter as his wife, but the king cheated Cambyses sending to him another woman because he didn’t want to see his daughter turned into a concubine23. The Greeks described the Persian society as a monstrous society, were the men were dominated by the women. Because of that, the Greek sources pay attention to the Royal Women intervention in the Court and their influence. However, we have to consider that these descriptions came from the Greek point of view and its stereotyped vision of Persian society. In most cases, maybe due to the influence of the Greek vision of the Persian harem, we think that Persian women were confined into the palace. Their appearances would be limited to accompany the King.

To refer to the king’s wife, the Greek sources use the term γυνή του βασιλεωσ, or the king’s name followed by γυνή. Plutarch uses to refer to Atossa “married wife” or lawful wife”12 Stateira is called τησ Δαρειου γυναικοσ13. The Persepolis Tables give us three terms to refers to Royal Women: mutu, irti and dukŝiŝ. Mutu appears in nine texts to refer to the king’s mother14 and irti to refer to the king’s wife. In the neo–babilonian period appears the term ŝa ekalli, to refers in legal documents to “the palace’s women”15, but we don’t know if it refers to the King’s wife or mother. Nevertheless, the proper name of the king’s wife, like Parysatis, Darius II’s wife, usually appears in legal documents, without any title 16. Brosius concludes that it means that the use of the name of the king’s wife without any title in palace’s documents implies that it is a sufficient reference and this fact demonstrates the importance of this figure 17. Nevertheless, Brosius adds that we don’t have enough evidences to conclude that sa ekalli means “king’s wife”, it could mean simply “woman with properties”. We have tablets that indicate that the king’s wife had properties,

This image is accompanied by the absence of Persian feminine representations in Achaemenid’s art. But this fact is not an evidence of the Persian women reclusion: Herodotus tell us about a gilded statue of Aristone, Darius II’s wife, and we also have representations of anonymous women like a limestone plate of IV BC, the bronze Kohl tube from Syria, presently in Ashmoleon Museum of V– IV BC24 or the Ammam’s feminine figure 25. Apart from the artistic representations, we have seals with feminine figures. Brosius thinks that the variety of objects and the different scenes represented in them suggest that not only

6

Drews 1983, 114–115. Alc. 123c; Plu. Them. 24–26. 8 Plu. Ale. 24, 7; Curt. 3, 12, 17. 9 Brosius 1996, 22. 10 Brosius 1996, 23. 11 Brosius 1996, 24. 12 Plu. Art. 5, 5. 13 D.S. 17, 21, 3. 14 PF–NN 151(mutana), PF–NN 1431; PF NN 1734; P–NN 2112 (muti); PF–NN 2145 (muti); PF–NN 2405 (mutar) 15 Parpola 1988, 75–76. 16 Stolper 1985, 63. 17 Brosius 1996, 29. 7

18

PF 1835–7; PF–NN 761; PF–NN 2523. Plu. Alex. 21. 20 Olmstead 1948, 424 21 Schmidt 1953, 255–264. 22 Root 1979, 101. 23 Hdt. 3, 1, 2. 24 Spycket 1980, 43. 25 Khalil 1986, 104. 19

44

A c h a e m e n i d R o ya l W o m e n a nd A l ex a nd e r the noble women were represented, but also the anonymous women.

After this introduction about the Royal Persian Women in the Achaemenid Empire, we’re going to talk about the relation between Alexander the Great and the Royal Persian Women and the most important episodes where they had a special role. The most important figure of the Persian women was without a doubt Sisigambis, mother of Darius III. She became the leader and the defender of this group of women after Darius’ escape and the neglect of his family in Alexander’s hands. We cannot say that Sisigambis installed a kind of matriarchate; nevertheless, she took the reins of the group and took care about her family as possible.

We have information about the Royal Persian Women and travelling. We know that the women of Darius III’s family travelled with him in the campaign against Alexander the Great, but we have information of other women, like Aspasia, Cyrus the Young’s concubine 26 or Cassadane, Cyrus II’s wife who accompanied them27. But we know some of the travels that Royan Persian Women made alone, as Royal House’s representatives, without the King’s presence. Amytis, Amestris’ daughter, travelled to Syria to persuade her husband Megabyxos to return to the royal court after having staged a revolt against the king 28 or Parysatis, who travelled to Babylonia to ask for his son Cyrus the Young’s body29. The Fortification Texts testify that Royal Persian Women travelled alone with an official capacity by the fact of holding a Royal title30, usually in commercial transactions.

After the battle of Issos and Darius’ escape in 333 BC, he deserted the women who went with the Persian army. That group included the women of the Royal Family and the women of his relatives, friends and generals. Briant opines that, in fact, the State itself was on the march 38. Diodorus says that it was a Persian custom that women went in opulent gilded carts inside of which travelled their trousseau and rich furniture. All of them were stolen by some of Alexander’s men, in front of the a astonished and terrified women’s group, who using to dress very smart and always covering their whole bodies, were forced to leave their tents and carts to escape from the soldier’s violence with their clothes totally slanted and crying. Diodorus adds that some women were abused by the soldiers, who didn’t respect their nobility39.

Finally, we have to talk briefly about public appearances. Royal women appear in public celebrations, like the King’s birthday31 because these official celebrations required their presence. Herodotus says that the Persian legates to Amyntas’ court ask for the presence of their women in banquets and feasts, including wives and concubines32. Plutarch pays attention to this fact, so explains that royal Persian women assisted at banquets but left it when the drink arrived33. Since that moment, only concubines, dancers, and musicians were allowed to be present. Brosius says that the concept of concubine was different among Persians and Greeks, and the separation between wives and concubines in Persian case at banquets and feasts didn’t exist, but it existed of course in the daily life, giving us the example of Barsine and Roxana34. The private dinners included the women’s presence. Heracleides of Kyme says that wives joined their husbands to enjoy breakfast and the dinner together 35 but we haven’t information about their role in these celebrations. Greeks viewed this Persian custom like something extravagant, so the Alexander’s decision of introducing it it’s described by Curtius as:

Nevertheless, some other careful Macedonians including Alexander and his closer men, were affected seeing the terrible destiny of the women belonging to the Royal Court and then turned into prisoners. Alexander decided to respect the women and went with his general Hephaestion to visit Sisigambis, Darius’ III mother. The old woman thought that Hephaestion was Alexander and addressed to him. When she realized her misunderstanding, Alexander said, “Don’t worry, mother, cause he is also Alexander”. After this meeting, the king ordered that all their properties and their servants were returned to them, gave Darius’ daughters a generous dowry, educated Darius’ son like his own son and took care of Darius’ wife40. For all of this, the king won the affection of all Persian women.

“There was a general slipping into foreign habits”36. About the royal Persian women’s role in funerary practice we haven’t information enough. Herodotus says that Persian people have a mourning period for the death of a member of the Royal Family, but not much about this 37.

In the same way Curtius tells the first meeting between Alexander and Sisigambis, saying again that the old lady confused Hephaestion with Alexander41. He emphasizes Alexander’s clemency and magnanimity due to his behavior towards the group of Persian Women, especially the Royal ones, who were treated as the nobles they were. Sisigambis expressed her gratitude to the king saying these words:

26

Ael. Var Hist. 12, 1; Xen. An. 1, 10, 2. Hdt. 2, 1, 1. 28 Ctesias FGrH 688 F 14(39). 29 Ctesias FGrH F15 (59). 30 Brosius 1996, 92. 31 Hdt. 9, 10, 2. 32 Hdt. 5, 18, 2: ημιν νομος εστι τοîsi Περσεσι. 33 Plu. Mor. 613a; Art. 5, 5; 26, 6. 34 Brosius 1996, 94. 35 FGrH 689 F2; Ath. 4. 145c; Plu. Art. 5. 5. 36 Curt. 6, 2, 2: Omnia in externum lapsa morem. 37 Hdt. 2, 1, 1. 27

“Oh, king! We will ask in our prayers all things we asked for Darius in another time; you are worthy of them (...). You call me mother and queen, but I’m 38

Briant 2002, 187. D.S. 16, 35, 2–7. 40 D.S. 16, 38, 1–3. 41 Curt. 3, 12, 17. 39

45

CLÀUDIA ZARAGOZÀ your slave: only you can decide if your power over us testifies your clemency or your cruelty”42.

prisoners like Sisigambis, ran to told her the good news. The notice didn’t make happy the queen mother, who remained immobile without articulating any word.

One of the most curious things is that Alexander calls Sisigambis “mother”, and she calls the king “son”. Diodorus said that this fact was a demonstration of Alexander’s humanity to their prisoners and a way to calm down Sisigambis and the rest of royal women about their future43. Iustin doesn’t explain the first meeting between Alexander and the Persian women, but tells us that Alexander notified that Darius was still alive, trying to calm down his family. Alexander added that he would arrange the daughter’s marriages according to their nobility44. Plutarch also tells us about this encounter and Sisigambis’ mistake, emphasizing Alexander’s magnanimity towards Persian women.

“Seeing her, we don’t know what she preferred”46. The same episode is told by Diodorus, who tells us that facing with the news the rest of women exhorted Sisigambis to join the celebration, but she remained immobile again. The author says that Sisigambis didn’t believe that turnaround of the Fortune and didn’t want to betray Alexander’s generosity. Therefore, if we trust this episode, we can think that between Alexander and Sisigambis began a relationship of respect that stopped any joy or celebration for Darius’ victory, because it could betray Alexander’s magnanimity towards her and her family.

After this first meeting and established the relationship between Alexander and his special prisoners, Sisigambis will appear in most of the chapters as the representative of the Persian women. One of the most important events is the incident with de woolen clothes, which reveals the collision between two cultures that took place in Alexander’s Asian conquest. Curtius gives a detailed description of this episode: in some occasions, some dresses and purple clothing arrived from de Macedonian Court to make more dresses, and Alexander sent Sisigambis all of this and a couple of women who would teach Darius daughters how to weave. In this way, they could make Alexander’s dresses in the future. But this meant a terrible offense to Sisigambis, because weaving was the most shameful profession for Persian women. Only slaves used to weave. In front of this, Alexander apologized to Sisigambis, saying he didn’t know this Persian custom, and remembered the queen that he had tried to respect other Persian customs, like waiting for the queen to sit on the table to dinner or call her mother, a title that Alexander considered

The Persian people knew the special relationship between Alexander and Sisigambis and tried to benefit from it. In this way we have another incident associated with the Satrapa Medates, king of Uxian, land where Alexander arrived with his army during his Asian conquest. The king proposed to take the Uxian cidadel because Medates, who was married to one of Sisigambis’ sister according to Curtius, expressed his intention to stay loyal to Darius III. In front of the disadvantageous situation for Medates with the Macedonian army at the walls of the cidadel, he asked Sisigambis to take part in the conflict and negociate with Alexander for a good solution for the Uxians. The queen mother resisted for a while, because she didn’t want to offend Alexander with this request, but she finally gave up. She sent a letter from Susa, when she remained whit the other women, to Alexander asking him for the life of the Uxians and her own life for that imprudence. Alexander forgave Medates, brought freedom to uxian people and allowed them to cultivate their own land without paying any tribute:

“that belongs to my lovely Olympias” 45.

“If Darius had been the winner, he wouldn’t have obtained so much from his own mother”47.

This incident demonstrates the cultural differences between Persians and Macedonians with an anecdote starred by Alexander, who made an effort to adapt to the Persian costumes. More than an anecdote, this announces the progressive assumption of Eastern costumes by Alexander, what brought him a lot of problems with his soldiers.

In this incident we can see that Sisigambis assumed the role of intermediary in this occasion, an important role if we keep in mind the result. We don’t know if this role only took place in that occasion. Nevertheless, we have to consider that if we trust this episode, Sisigambis’ role became crucial, given that Alexander achieved Sisigambis’ requests. Requests made from a woman, and not any woman, but his enemy’s mother. Not even Olympias made a request as important as that which was accepted by her son, what gives us an idea of the special relationship between the Macedonian king and the queen mother.

Sisigambis demonstrates her gratitude to the king in a lot of episodes that sources tell us. We don’t forget that she “adopts” Alexander as a son (at least verbally) while Alexander was still persecuting Darius, her biological son. This fact brought Sisigambis in complicated situations, like an incident told by Curtius. When some news arrived at the camp telling about a first victory of Darius over Alexander, and most of the Persian people,

The last action Sisigambis did as another example of her relation with the king was his reaction in front of

42

Curt. 3, 12, 23–25. D.S. 16, 37, 5. 44 Iust. 11, 9, 12 45 Curt. 5, 2, 18–22. 43

46 47

46

Curt. 4, 15, 10–11. Curt. 5, 3, 12–15.

A c h a e m e n i d R o ya l W o m e n a nd A l ex a nd e r Alexander’s dead. Diodorus is brief in his relate of the fact:

Arrian tell us about queen Stateira’s beauty and the Alexander’s containment. He says that she was the most beautiful woman in whole Asia but the king didn’t lust after her, or if he did, he controlled it:

“After the king’s dead, Darius’ mother, Sisigambis, regretted for a long time Alexander’s death and her own solitude. And, because of her arriving at the limit of her life, she left herself to die after five days of fasting, leaving life with great sorrow, but not with shame”48.

“Being Alexander young as he was, and in the crowning moment of his career, when young people are more unyielding”51. Alexander didn’t lie with Darius’ wife and respected her, demonstrating his reputation again. Curtius reiterates that Alexander didn’t inflict any violence against Stateira, who had the first place of beauty among all women and the king ordered everyone to respect Stateira as he did.

She starved herself to die of hunger. When Darius’ death took place she decided to stay alive, because of her role as “matriarch” and her responsibility to keep her family safe. But after Alexander’s death, she and her family lost their protector and savior, so she considered that it wasn’t necessary to stay alive. Curtius explains with a lot of details Sisigambis’ reaction in front of the misfortune of losing no one, but two sons. Curtius hit the nail on the head: what would happen with her grandchildren, Darius’ daughters and son, after Alexander’s death? Sisigambis realized that without Alexander no one can assure her security and her family’s. Therefore, we suggest that in addition to the special relationship of affection between Alexander and the queen mother, Sisigambis was really concerned about the future of her family without the figure who kept them alive.

The sources tell about Alexander’s good intention towards Stateira, but also reflect Darius’ nervousness about Alexander’s and Stateira’s relationship. In some occasions he doubted about the news about Alexander’s behavior towards Stateira, Sisigambis and the rest of the Royal Family. Plutarch testifies Darius’ doubt. Stateira was pregnant when she was captured at Issos, and she died in the labor. Plutarch tells us how a eunuch, Tireus, who had also been captured by Alexander, gave Darieus the news. Tireus escaped from the Macedonian camp to tell Darius about her wife’s death. The Great King cried a lot for his wife’s destiny and her body, which rested without the appropriate burial, but the eunuch interrupted him telling that Alexander celebrated a luxurious burial for Stateira according to the Persian customs. Darius doubted about his wife’s loyalty:

We have seen that the relationship between Alexander and Sisigambis is probably the most important king’s relationship with a woman, apart from the relationship with his mother, Olympias. First, he called “mother” his enemy’s mother and decided to gave Sisigambis, Darius’ wife and the rest of the family, their status as a Persian Royal Family, establishing very special conditions for them as prisoners. We can understand Sisigambis’ gratitude toward Alexander, which made her to behave as we explained before, in complicated situations trying to avoid defining her own position. Nevertheless, we can say without a doubt that her decision to let herself to death was associated to Alexander’s death, which also killed all the hope for Sisigambis and the rest of the Royal Family.

“Which kind of honest relationship could take place between a young man and his enemy’s wife to have her in so high appreciation?”52. But the eunuch, interrupting the king again, threw himself at Darius foot, begged him to stop talking and asked him not to slander Stateira and Alexander, because they didn’t have any kind of relationship, and even told the Great King he should admire Alexander for the temperance that he had had with the Persian women.

The relationship between Alexander and Darius’ wife, Stateira has been another example used by the sources to demonstrate Alexander’s magnanimity and generosity. The sources tell us about a respectful relationship, but this one wasn’t as intense as Sisigambis’ one. Most of the sources emphasize Stateira’s beauty. Plutarch tells that Stateira was the most beautiful queen, as Darius was the most beautiful king, and their children were like their parents. But Alexander never touched one of them49. In this way, Alexander explained to Parmenion his respect to them:

Like in Sisigambis case, the respect towards Stateira is a reason to honor Alexander, who ordered an appropriate burial for her. Arrian tells us an episode in where an eunuch, who had escaped from the Macedonian camp after the battle of Issos , arrived at Darius camp. The Great King asked the eunuch about his mother, wife, and children. When he knew that all them were safe, the king asked if Alexander had respected Stateira and the eunuch answered he had. At that moment, Darius exclaimed that if at any time he lost his power over the Persian people, gods should only give it to Alexander.53 Iustin confirms that Alexander never had any relationship with Stateira, he explains us how Darius received the news about her wife’s death:

“Indeed, it won’t be found that I had seen Darius’ wife or would have wanted to see her, but not even listened to anybody who told me about her beauty”50. 48

51

49

52

D.S. 16, 118. 3. Plu. Alex. 20. 50 Plu. Alex. 22, 5.

Arr. An. 4, 19, 4–6. Plu. Alex. 30, 1–10. 53 Arr. An. 4, 20, 1–3.

47

CLÀUDIA ZARAGOZÀ light army with their bosses, responsible for taking the march in regular rhythm”57.

“In the way, he receives the news about her wife’s death because of a miscarriage, and that Alexander had cried her death, had followed cortège and had done this not for a loving feeling, but humanity; and that, in fact, Alexander had only visited Stateira once, while he had frequently visited Darius’ mother and daughters”54.

We see that Curtius also fixes the number of concubines in 365. Apart from the Royal Family women, he talks about a mass of women riding horses, without specifying their status or their number. Besides those women, who we supposed they were a large number, the concubines and the different wives, Curtius says that the infants were accompanied by their governesses, who travelled together in armamaxas, with the eunuchs, who we know that accompanied the queens and were usually messengers to Darius. This last group travelled in fifteen carts, a large number too.

Once again, we find the image of a respectful and kindhearted Alexander towards her enemy’s wife, what honored him and even Darius himself had to thank his enemy for this. In the next part of this paper we want to talk about a group of women usually forgotten by scholarships, maybe due to the poor information that sources report to us. We’re talking about all women without a name who went with the Royal Women when Alexander made them prisoners at Issos. The beauty of those women is reflected by the king himself. When Alexander met them first, the king was impressed by all their beauty, and Plutarch says that the king exclaimed:

The information about the members of this group is limited. Nevertheless, with this few lines about it we can make an image of the picture that Alexander found at Issos. Apart from the Royal Family, important prisoners, Alexander had in his hands other important hosts: the most important Persian men wives. We don’t know if the daughters and little sons not older enough to fight travelled with them due to the absence of words in the sources about them, but we suppose that this was very probable, because it has no sense that with men in combat and wives in the legation the daughters and sons were remaining at home alone. We have also a number of 365, that even might be exaggerated is the one given by sources. We have to add an undefined number of governesses and the women riding horses, and the result is a large number of women (and eunuchs, and children) in Alexander’s hands. We don’t know if the women apart from the Royal ones also had a special status like them or not. The sources don’t talk about it. It’s possible that the ones who belonged to the Royal Family service would go on doing their office, given that, as we said before, Alexander ordered that all properties and servants were returned to the Royal Family. The women riding horses are still unknown to us due to the poor information. About the other wives, we don’t know if because of their noble position they were treated in the same terms that Royal women were or not, as the sources kept silence again. However, we know that, no matter in which conditions, all of them turned into Alexander’s prisoner. And, if we trust sources, in special conditions that all other group of captives.

“Persian women are a torment to our eyes”55. Nevertheless, demonstrating his temperance and control, he passed over them as if they were lifeless statues. Diodorus reaffirms this aspect. The author says that Darius had a court of concubines in a number no less than the days of the year, of extraordinary beauty as they had been chosen among the most beautiful Asian women, and every night they walked in front of the Great King so that he could decide with who he would spend the night with56. Therefore, from of this passage we have new information: at least, 365 women travelled with the Royal Family. But we have further information about who travelled with Sisigambis and her family. Curtius give us more details about Persian legation. In one passage the author explain us that, with Darius’ family women traveled with him: “At a stadium of distance, a cart transported Sisigambis, Darius mother, and another his wife, the queens, who carried as bodyguards a mass of women riding horses. Behind them there was fifteen carts called armamaxas. In them were travelling the infants, and with them their governesses, and a crowd of eunuchs, who were quite regarded among that people. Following them went the 365 king’s concubines, also royally dressed in, and behind them, the royal treasure was transported in 600 mules and 300 camels, escorted by a mass of archers. Inmediately behind them went the relatives’ and friend’s wives, and banchs of cantiners and water and firewood bearers. Closing files went a troupe of

As we know, Darius had two daughters, Stateira and Dripetis and a son called Ocus. We also know that in Alexander’s and Darius’ times it was usually to make wedding alliances even with your worst enemy, and Darius had no one, but two daughters to use in that term. We had explained that after the battle of Issos’ Persian Royal women’s had been captured by Alexander and became his special prisoners. When Darius knew it, he sent to Alexander a letter with a marriage proposal to her daughter Stateira. Plutarch tell us that the Great King sent another letter to Alexander asking for the liberation of the prisoners in return for ten thousand talents, the hole

54

Iust. 11, 12, 6. Plu. Ale. 21, 5. 56 D.S. 16, 77, 6–7. 55

57

48

Curt. 3, 3, 22–25.

A c h a e m e n i d R o ya l W o m e n a nd A l ex a nd e r territory at west Euphrates and one of his daughters in marriage58. In the same words says Diodorus, who tell us that Darius sent a legation to Alexander to express his gratitude for the manner over his mother, wife and children and the Great King offered him the amount of three thousand talents, the territory at west Euphrates and one of his daughters, with the intention of signing a peace treaty. Becoming in his son–in–law, Alexander could work as his own son, ruling together 59. Arrian explains practically the same, giving us the sum of one thousand talents and adding one petition: Alexander had to liberate Darius’ family60. Nevertheless, Arrian wants to make clear that Alexander rejected it: if he had wanted to marry Darius’ daughter, he could done it without Darius’ consent. Iustin says the following:

other Persian Women, like a common action of reconciliation, due to the dissatisfaction of the Macedonian soldiers because of Alexander’s attitude toward the Persians65. It is Arrian who gives us more details about the marriages and new different information from Plutarch and Iustin: “The king celebrated in Susa his own marriage and his companions’ ones. Took as wife Dariu’s old daughter, Barsine, and as Aristobulous says, also Oco’s young daughter, Parisatis, although he was married to Roxana, Oxiarte’s daughter”66. We have some problems with that version: first of all, Darius’ old daughter is called here Barsine, like Memnon of Rode’s wife and Alexander’s lover too, and secondly we find here a new Alexander’s wife, Parisatis, Oco’s daughter. But right after, Arrian gives us a list of Macedonian names and his respective wives, who married at Susa:

“Some time, is sent to Alexander another letter from Darius, in which the Great King offers his daughter in marriage to Alexander and a part of his kingdom. But Alexander responded that Darius was trying to give Alexander something that was already his own propriety. He ordered Darius’ to came to him imploring and leave Alexander to make his own decisions about his kingdom”61.

“The king ascribed to Hephaestion Dariu’s young daughter Dripetis, his wife’s sister, so that Hephaestion’s children became part of the Macedonian Royal Family, being his nephews; Craterus married Amastrines, Oxiarte’s daughter, who was Dariu’s brother; Perdicas married Atropate’s daughter, satrapa of Media; Ptolomeus, his personal guard, married Artacama and Eumenes, the Royal secretary, Artonis, Artabazo’s daughters; Nearchus married Barsine and Mentor’s daughter; Seleucus married Spitamenes the Bactrian’s daughter; and in that way the rest of the Macedonians married the most noble Persian’s daughters in number of eighty”67.

Alexander didn’t accept any of Darius’ petitions. Not while Darius stayed alive. After Darius’ death, Alexander considered that a marriage to some of the Great King’s daughters was practically a necessity, as that marriage could permit a strong alliance, with the possibility of a common heir to Macedonian and Persian people. It could avoid tensions with the Persians. Tensions that, without a doubt, had increased since the news about the marriage of Alexander and an unknown girl of the Sogdian mountains called Roxane. Nevertheless, the union between Alexander and Stateira is not the same in all the sources. The most extended version is the Susa’s Marriages in 324 BC, very problematic for the scholarships due to the confusion, repetition or even disappearance of some of the personages in the different versions62. Plutarch says that Alexander celebrated the marriages between his most important men and western women at Susa, where he himself took Stateira as his wife and ascribed the best women to his men, and the number of guests reached the sum of nine thousand63. Diodorus says that Alexander married Stateira at Susa too, and in addition, Hephaestion married Dripetis, Darius’ other daughter: “[Alexander] persuaded his closer friends to being married and ascribed them the most distinguished young women among the Persian nobility”64.

As we see, all women were related to Darius or they were noble’s daughters. We want to emphasize the marriage between Nearcus and Barsine’s daughter, the same Barsine who was Alexander’s lover, who was in fact Heracles’ sister, Alexander and Barsine’s son. But Arrian didn’t say anything about this. In addition, Arrian gives us a description of the marriages: how Alexander received the wives’ dowry and inscribed all Macedonian men who being married in Asia (over ten thousand) to give them their respective marriage present68. The fact that Arrian called Darius’ daughter Barsine it’s not a big problem, because we usually find names in the sources that are confused or repeated. But Parisatis only appears in this source. It’s not reliable information, because only Arrian tells something about her. Arrian tells, like the rest of sources, that Alexander married Stateira–Barsine, Darius’ old daughter, after being married to Roxane, and Hephaestion married Dripetis, Darius’ young daughter.There is no proof in any sources

Iustin doesn’t give us any significant information, either simply employs two lines to mention that Alexander married Stateira and other Macedonians did the same to 58

Plu. Alex. 29. 7. D.S. 17, 54, 2. 60 Arr. An. 2, 25, 1. 61 Iust. 11, 12, 3. 62 Van Oppen 2013, 2–7. 63 Plu. Alex. 70, 3. 64 D.S. 17, 107, 6. 59

65

Iust. 12, 10, 9 Arr. An. 7, 4. 67 Arr. An. 7, 4, 5. 68 Arr. An. 7, 4, 7–8. 66

49

CLÀUDIA ZARAGOZÀ about the existence of children from those marriages, as Hephaestion died after Susa Marriages, in autumn 324 a.C. and Alexander died the following year. The two women’s destiny after their husband’s death is explained by Plutarch: he says that Roxane, pregnant by Alexander, was jealous towards Stateira and after Alexander’s death she cheated her sending a letter inviting Stateira and Dripetis to go to hers. When the sisters arrived, Roxane murdered them and threw the bodies into a well, with Perdicas’ help69.

Nevertheless, we consider that although the most accepted option is the one about the massive repudiation, we have to keep in mind the option suggested by Van Oppen for future investigations. We have seen throughout this paper the importance that Royal Persian Women have in the Alexander’s the Great Asiatic campaign. Obviously, their names and acts are less important than others staged by men for most of the scholars, but our intention in the present paper is to demonstrate how a group of foreign women became major figures of a lot of episodes during the campaign years. This fact shows three things: first, the role that Sisigambis took in a dangerous situation for her and her family, second, an image of Alexander different from his image as a conqueror and at last, the cultural differences between Persian and Macedonian. Sisigambis’ paper in this story isn’t understandable without the study of Royal Persian Women in the Achaemenid Empire, that we explain in the first lines of this paper. However, we have seen how the queen mother assumed her paper as a leader, taking part in conflicts like the Uxian one as an intermediary between Alexander and Medates, due to the special relationship established between the queen mother and the king. But we also know a king’s unknown side: his generosity, magnanimity and compassion toward these special captives. All the sources make an effort to show this Alexander’s facet, in contrast to the cruelty that the king demonstrated towards other people and tribes. We have also seen the two different cultures thanks to some of the episodes, like the clothes incident or the women’s presence or absence in banquets and feasts. With this paper we want to bring to light these women’s stories, that in fact are bound with Alexander the Great’s story since without him they might not be remembered or mentioned by ancient authors and would remain forgotten.

We have seen in the case of the Susa Marriages a use of the marriages as a political weapon. Being married to those Persian women meant to make alliances between the Macedonians and the Persian and Indian Nobility what reported political and economical earning to both sides. Nevertheless, a question to discuss is the fact that after Alexander’s death, Macedonian and Greeks being married with Persian and Indian women repudiated them, in opinion of most of the scholarships 70. We think that this is accepted because, after Alexander’s death, we haven’t any information about those women. We don’t know the destiny of any of the women that Arrian enumerates in the marriages list. Ogden considers that Persian women were forced by Alexander to marry and after the king’s death they “escaped” from their Macedonian husbands71; Cary thinks that, despite of Seleucus, the rest of Macedonian officers got rid of their Persian and Asiatic wives72, and this implies that they were forced to marry them by the king; Bosworth, for his part, speculates about an idea: the fact of being married to Asiatic women could be so humiliating to Macedonians that the unions didn’t continue after Alexander’s death 73. Nevertheless, other scholars like Van Oppen 74, adjusting to the axioma “lack of evidence doesn’t evidence the lack”, suggests that if we haven’t any information about them after the king’s death (except notices from Apame, Seleucus’ wife that starts Seleucide dynasty), neither about the supposed massive repudiation, we cannot assume this like reliable. Van Oppen add some arguments that could prove that in most of cases those women didn’t survive longer after Alexander’s death, that’s why the lack of information. It’s not our intention here to take part of a discussion about a so complicated matter.

69

Plu. Alex. 77, 6–7. Brosius 1996, 78; Ogden 1999, 69; Hölbl 2001. 71 Ogden 1999, 6. 72 Cary 1951, 250. 73 Bosworth 1988, 157. 74 Van Oppen 2013, 15–16. 70

50

A c h a e m e n i d R o ya l W o m e n a nd A l ex a nd e r Bibliography BRIANT, P. 2002, From Cyrus to Alexander. A History of the Persian Empire, Eisenbraun.

OGDEN, D. 1999, Polygamy, Prostitutes and Death: The Hellenistic Dynasties, London/New York.

BROSIUS, M. 1996, Women in Ancient Persia, (559–331 BC), Oxford.

OLMSTEAD, C. 1948, “A Greek Lady from Persepolis”, AJA 54, 10–18.

CARY, M. 1951, A History of the Greek World from 323 to 146 B.C, London.

PARPOLA, S. 1988, “The Neo–Assyrian Word for ‘queen’”, SAAB 2, 73–76.

CARNEY, E. 1991, “What’s in a Name? The Emergence of a Title for Royal Women in the Hellenistic Period”, in S. POMEROY (ed.), Women’s History and Ancient History, London, 154–172.

ROOT, M. C. 1979, The King and Kingship in Achaemenid Art: Essays on the Creation of an Iconography of Empire, Leiden. SCHMIDT, E. F. 1953, Persepolis, I–III, Chicago.

DREWS, R. 1983, Basileus: The Evidence for kinship in Geometric Greece, London.

SPYCKET, A. 1980, “Women in Persian Art”, in D. SCHMANDT–BESSERAT (ed.) Ancient Persia, Malibu, 43– 45.

HÖLBL, G. 2001, A History of the Ptolemaic Empire, London/New York.

STOLPER, M. W. 1985, Enterpreneurs and Empire: The Murašu Archive, The Murašu Firm, and Persian Rule in Babilonia, Leiden.

KHALIL, L. F. 1986, “A Bronze Caryatid Censer from Amman”, Levant 18, 103–110. MACURDY, G. H. 1928, “Basilinna and Basilissa: The Alleged Title of the ‘Quee–Archon’ in Athens”, AJP 49.3, 276–82.

VAN OPPEN, B. 2013, “The Susa Marriages. A Historiographical Note”, AncSoc 44, 2–7.

51

When Greece came to India Herodotus and the Eastern boundaries of the world* César Sierra Martín Universitat Autònoma de Barcelona ABSTRACT The main purpose of the following lines is to analyze the contacts between Greece and India in the period between the end of the Archaic and the beginning of the Classical Ages. Through the careful read and study of the extant fragments of the works of Scylax and Hecateus and the books written by Herodotus, we can appreciate how India became configured in the Greek mentalities as a dual space, with an unclear difference between the geographic and ethnographic description, by one hand, and the paradoxographic, by the other. The absence of a continuous, direct and fluid contact between India and Greece surely was the main cause for this division in the concept the Greeks had about India, and even so, for the maintenance of this picture unaltered along the centuries in the Graeco–Roman Culture. KEYWORDS Greek Paradoxography – Ancient Geography – Indian Gold India and Greece: first contacts.  Any essay about the Greek knowledge about India needs to begin with the reference to Scylax of Caryanda, mariner in the late VI century BC. As it is frequently assumed, Scylax was the first Greek who travelled to India in a journey of exploration organized by Darius I, who was considering invading the country (Hdt. IV. 44)1. This information gains some veracity if we take into account that we have no other mention of India in the extant Greek sources of the Archaic Greece, and it is just after Scylax’s trip that the early references to India begin to appear in the authors of the Classical Greece2. Modern historiography agrees to consider the tour of Scylax as the 

Universitat Autònoma de Barcelona. Àrea d’Història Antiga (Proyecto HAR 2011–23572; external member of the Research Group HICOAN – 2014 SGR1111 of the Generalitat de Catalunya). 1 Herodotus’ passage about the travel of Scylax is maybe the main starting point for any kind of work about the relationships between the Greek world and India: Cft. Karttunen 1989: 65–68; Casevitz 1995: 9; Gil 1995: 26–32; Gómez–Espelosín 2000: 86– 104 and Albadalejo 2005: 15–20. This travel rememebers in some way to others made under the patronate of kings, as that one organized by the Pharao Necho (609–594 BC) , who has sent a Phoenician expedition to explore and circumnavigate Africa (Hdt. 4, 42, 2). 2 See for example the opinion Albadalejo 2005: 15, and our consult of Thesaurus Linguae Graeca confirms it (Stephanus 1954: 606–607).

main source that inspired the logographs’ later accounts about India, as Hecateus’s and lately, Herodotus’s himself3. Nevertheless, we just have scarcely information about Scylax and this has frequently come indirectly. It seems that Scylax had been renowned somehow as mariner and naturalist, and these facts came to knowledge of the Great King of Persia Darius, who had been considering sending an expedition in order to recognize and to explore the Indus River to the Sindh country4. Therefore, Darius wanted for this mission that an expert mariner could write a detailed report about the routes adding also all the recordable data about geography, ethnography and habits of the people living in the surrounding area. According to Herodotus, the expedition started at the city of Caspatyrus, in the region of Pactiyc, places that are hardly localized by modern researchers, who doubt intensely about the authentic reliability of this information, finding them probably in the area of Gandhāra5. In the northwest of India, the expedition

3

Gómez–Espelosín 2000: 94 ff. Strabo 12, 4, 8 recoded the activities of Scylax in Bitinia (Black Sea), but we do not know if they were after or before his travel across the Indus. 5 We follow here the opinion of the recent historical commentary of Asheri/Lloyd/Corcella 2007: 613, although we actually know the controversy produced by the read in Hecateus of the place name πασπάπυρος/paspápyros (FGrHist 1 F 295). The development of the different approaches can be found in How/Wells 1967: 319; Gisinger 1972: 620–623 and Gil 1995: 4

CÉSAR SIERRA MARTÍN travelled across of the Indus to the Ocean continuing to reach the Persian Gulf, in a trip similar to that guided by Nearchus in Alexander’s last months.

write a tour nowadays, we guess that the book delivered to the Persian King would did not contain information about fantastic or imaginary peoples or any other paradoxographic data. In this sense, closely to the purposes of Darius should surely be the references to a kind of prickly plant which supposedly grew in the banks of the Indus: kynára/κυνάρα, as is recorded later by Hecateus10. About that, Aristotle (Pol. 1332b) records a fragment of Scylax’ account about the relationships between subjects and kings, but the passage is absolutely out of context. Nevertheles, all these scarce fragments about Scylax’s work sheds some light on the nature of a book that probably contained very different types of information, from naturalist to political, with very interesting data for the Persian King.

The short extant evidences about the account of Scylax are also involved in a high paradoxography fashion. As far as we know according to Philostratus (Vita Apollonius III. 47), Scylax wrote a series of treatises where he described the different peoples that lived in India, detecting the pygmyes, located at the East of the Ganges; the sciapods and macrocephals, all of them mythical populations with fantastic costumes. Similarly, John Tzetzes (Chil. VII 629) wrote an essay about India, including the sciapods, a people whose members just had one huge foot useful to move themselves fastly, and even useful to protect them from the sun using the foot as an umbrella. Besides, we must add the otolicnos, owning big ears, the monophtalmoi who had only one eye, the enotiktontes, described as enormous monstrous6. There are also many other references to Scylax’s geographical and etnographical description, such as Harpocration, who assures that Scylax located in India the troglodytes, people who lived underground in caves (FGrHist 709 F6). As a result of all these kind of examples, and many other that we can mention, scholars have usually considered Scylax’s account as a tour of an intense paradoxographical content, which was well accepted by Greek audience7. In our opinion, it seems difficult to venture the nature of the work of Scylax taking into account the current information. According to the information recorded, mainly in Herodotus and Strabo, Scylax was a famous mariner, popular by his descriptions of the Pontic region, but especially of India. We can also link Scylax to the incipient Jonian ethnography, highly described later by Hecateus and Herodotus8.

We must not forget that between India and Greece, the Persian Empire stood as a geopolitical frontier that probably seemed impassable for most the Greeks. Threfore, paradoxographical data recorded in some late accounts related with Scylax’ work can not be contrasted during the times of the Archaic and Classical ages, so this is the reason why the fantastic vision about India prevailed long time. It seems that the Persian political leaders had a kind of information about India that was very different compared to the information handled by the Greek logographists of the VI century BC, so they have had order a exploration in order to obtain and corroborate their strategic information. Nevertheless, the impact of Scylax’ travel in the Greek audience was very significant, and we consider of interest his contribution to the ancient Greek cultural conception of the world. From then on, the Greek wises began to show their interest about India in an ethnographic sense. A few years after Scylax’s adventure, the famous logographer Hecateus of Miletus wrote a treaty named περιήγησις/Periégesis (description) that compiled the geographical and the ethnographical information that Greeks have had accumulated during decades. Unfortunately, the extant sources we have about Hecateus’ work do not allow us to rebuild in detail the level of Greek knowledge about India at the end of the archaic age. However, some data linked the information of Hecateus about India with Scylax’s travel, as the mention of the kynára in both cases. Nevertheless, Hecateus recorded toponims that Scylax’s account did not list (at least in our extant fragments), as for example the already quoted πασπάπυρος/paspápyros, a city that Hecateus located in the region of Gandara (FGrHist 1 F295), or Ἀργάντη/Argánte (FGrHist 1 F297); and also the details about ethnonims like those of the Καλατίαι/calatíai (FGrHist 1 F298), ᾿Ωπίαι/Opías, people who lived in a strongehold (FGrHist 1 F299); and the Γανδάρας/Gandáras, who fought against Dionysos (FGrHist 1 F294). If Hecateus and Herodotus used the work of Scylax as a source to write their accounts about India, as the most part of the scholars suggest, it seems

Firstly, we have information about the book that Scylax supposedly wrote for the Great King Darius from the reference of scholia in Pseudo–Scylax, which mentions that Scylax had dedicated to the king a work describing a tour of the oikouméne (FGrHist 709 T4)9. Moreover, Marciano of Heracles, a Greek geographer who lived during the Late Roman Empire (IV century BC), assures that Scylax wrote a route, but he does not give us extra information (FGrHist 709 T6). Whether Scylax would 28 n 11. In Herodotus, the city of Caspatyrus and the Pactiyc region are also mentioned in Hdt. 3, 102. 6 Cft. Tzetzes’s account in Gisinger 1972: 627–628; Jacoby, FGrHist 709 7. 7 This is the explanation followed by Gómez–Espelosín 2000: 93, although Albadalejo 2005: 16 assures that Scylax could write in the beginning a ethnographic and naturalistic work, but later he could extend this kind of content with the aim of gain more attention from the Greek audience. Karttunen 1989: 66 shows himself a more cautious about this question. 8 Karttunen 1989: 68; Albadalejo 2005: 16–17. 9 Information from the mss. Parisinus graecus 443, ascribed initially to Scylax but now dated in the first part of the IV century BC (Gómez–Espelosín 2000: 88.). The term ‘περίπλοος/periplus’ arises in the context of the maritime explorations of the archaic age, due to the need to compilate the geographical and ethnographical information: Dueck 2012: 6.

10

Ath. 2, 82=FGrHist 709 F3 (Scylax) and 2, 70= FGrHist 1 F291 (Hecateus).

54

W h e n G r e e c e c a m e t o In d i a realistic that Scylax’s work was probably focused more in this kind of reliable information that in the description of uncanny or mythical peoples and other fantastic peculiarities. In fact, we also have a fragment of Hecateus (FGrHist 1 F327) recording the existence of the sciapods as an imaginary people who lived in Ethiopia and therefore we do not consider Hecateus a paradoxographic writer, since we have many more fragments from his work that allow us to know the real concept and contents of his work11. Something similar could happen in the case of Scylax.

Herodotus and India. The lógos about India of Herodotus is the first account of considerable dimensions that have reached to us about the perception of India in the Classical Greek culture (Hdt. 3, 98–106)14. In fact, the main historical commentary about the work of Herodotus defines the excursus about India as chaotic, centered in the explanation of how the Indians extracted the gold and paid the tribute to the Persian King15. Other scholars maintained that this digression about India is based on Hecateus account who took the references from Scylax16. As far as we are concern, all this is part of the historiographical tendency centered in the search of the sources of Herodotus, the Quellenforshung, which has in the digression about India one of the main focus of the scholar’s attention17. We think difficult to know which were the authentic sources of Herodotus to write his Indian lógos, as far as we have not a clear idea of his antecedents. We can understand that one reason for Herodotus to write his account about India could be the will to explain the causes of its rich golden wealthy, although this can be surely not the unique motive. In fact, the themes developed in his account are various, with predominance of the ethnography. For Herodotus, the habitants of India were located in the antipodes of the Greek lifestyle (δίαιτα/díaita), and this served to him not only to stress the differences in a pejorative way, but to show the power of customs (νόμος/nómos) around the world. A paradigmatic case of this can be found in the famous comparision of the funerary costumes between Greeks and Indians callatiae (Hdt. 3, 38, 3)18. In Herodotus’ opinion, the situation is originated in a kind of social experiment organized by Darius I, who wanted to show that each individual loves his customs as the best ones, more suitable and logical.Therefore, Darius reunited in his court some Greeks and he asked them if they would be ready to eat the corpse of his parents if the King payed them a big amount of money. The Greeks, of course, showed them shocked by the proposal, but then Darius called to the Indians callatiae, to which he asked them to burn their parents’ corpses by a big amount of money. The Indians shout the king not to blaspheme. In our opinion, the essence of this episode is the will to show the power of custom and the cultural distance between Greece and other regions in the world, and this can be also considered the second motive which dominates the Indian lógos. As a consequence, although at the start of the account about India, Herodotus tries to explain how the Indians obtained these fabulous amounts of gold, the truth is that

The ancient Greek culture received with interest and with pleasure the news of this new region of the extreme orient, India. However, the knowledge that the writings of other Greek intellectuals detached from periplography and geography is very mixed–up. As an example, Aeschilus in his Supplicants (vv. 284–286), where he includes a dialogue between the chorus of Danaids, native from Egypt, and the King of Argos, who rejects his Greek roots and defends to have a kind of likeness with the women of Libia and Egypt. The king says that he knows about the existence of a group of Indian nomad women who mounted in chamels and lived in India, locating it as a neighboor land of Ethiopia12. The geographical confusion is clear, but the inclusion of India as a remote region was an effect of the visions of Scylax and Hecateus. On the other Sophocles seems to be the first to talk about the gold sources of India. This is a stereotype that will have a great reception among the Greek authors. In Antigone (vv. 1037–1039) Creon reproached Tiresias his avaricious behaviour, recommending him to make his business for the white gold of Memphis or for the gold of India. A similar approach can be found in Xenophon (Cyr. 3, 2, 28), where Cyrus tries to gain the money of the Indian king13. Following our impression, the fabulous fortune of India is a paradoxographical issue because of the impossibility of having a real contact between Greece and India.

11

The identification of the work of Hecateus with the historiography through his Genealogies, and with the geography–ethnography of his Description is not currently accepted without discussion: Alganza 2012. 12 There are different readings of this passage: in a version of the text, very extended among the editors and translators, the women belonged to an Indian ethnic: Ἰνδάς (Ferrari 1995: 250), Ἰνδούς (Mazon 1976: 23); while in the Loeb edition, H. W. Smyth points out τοίας (Smyth 1973: 28). As it seems, the read implied in the Indian ethnic resulted from the edition of F. C. Tucker Brooke of the beginning of the XX century. On the other hand, about this geographical confusion about the proximity between Ethiopia and India, see Karttunen 1989: 134–138 and Casevitz 1995: 11–13. We prefer not to read “India” when the text says “Ethiopia” as a result of a mistake between these regions. 13 Xenophon calls “the Indian” (τοῦ Ἰνδοῦ) to the Indian King, in the same way as he talks about the King of Armenia or Asiria. In the most passages of Cyropaedia when the Indian King is related, he appears linked to his great richness: Cyr. II. 4. 28; VI. 1. 2.

14

Karttunen 1989: 73; Albadalejo 2005: 27. Asheri/Lloyd/Corcella 2007: 496. 16 Karttunen 1989: 68; Gómez–Espelosín 2000: 95; Albadalejo 2005: 16–17. 17 About this historiographical line in Herodotus, see Hornblower 2002 and the most extensive and notorious work of Fehling 1989. 18 The passage is used to justify that Cambises’ outrages in Egypt constitued a clue of his madness (Baragwanath 2008: 116), but generally the lesson from the episode rules the whole work of Herodotus. 15

55

CÉSAR SIERRA MARTÍN Lybia (Hdt. 4, 185)22. After this geographical precision, Herodotus follows describing the people that inhabited this region, stressing in those who he suposed that can attract attention of Greek audience. As an example of this it can be found the customs of the Indians padaei, who did not take care of the sick and eat them in order to prevent to spoil their meat (Hdt. 3, 99)23; nevertheless, other people lived in a peculiar way as far as they did not kill any living being, did not have houses, were vegetarians, and when they become illness they move away to a lonely place without taking any other measure (Hdt. 3, 100). After that, Herodotus describes the common physical characteristics of all Indians, as the black colour of their skin and their sperm (Hdt. 3, 101; 7, 70, 1)24. All this is inserted in a line of knowledge that search to visualize two cultural opposite extremes with the aim to caused impact in a Greek audience 25.Thus, the more remarkable customs are the cannibalism (Hdt. 3, 38, 3; 3, 99); the consumption of crude meat and the public copulation (Hdt. 3, 98, 3; 3, 101)26. In addition, Herodotus does not forget to mention the material culture, always with the clear intention to obtain the deepest impact in the reader. In this sense, it can be considered the news about certain kind of Indian boats, built from a unique cane, or the information about the clothing, made with rushes (Hdt. 3, 98, 3–4), the use of cotton (Hdt 3, 106, 3; 7, 65), the equipment of their infantry and cavalry (Hdt. 7, 70, 2; 7, 86, 1).

etnographical thematic includes the following chapters: Hdt. 3, 98, 2 – 102, 1, both inclusive, whereas the explanation of the origins of the Indian gold is developed in the chapters 3, 102, 2 – 106. Therefore, the book dedicates the same part to the geographical and ethnographical information (people living in India, exotic lifestyles, naturalistic data) than to the golden question and the paradoxographical perspective. From our point of view, the ethnographic technique used by Herodotus in his Indian lógos cannot be considered as different to the main lines of the rest of his work, although the space that the description of India occuped in the book is short. Therefore, Herodotus describes India in a geographical way as an extreme region belonging to the Persian Empire and that constituted the eastern satrapy. The description of India deserves full attention: ‘ἔστι τῆς Ἰνδικῆς χώρης τὸ πρὸς ἥλιον ἀνίσχοντα ψάμμος: τῶν γὰρ ἡμεῖς ἴδμεν, τῶν καὶ πέρι ἀτρεκές τι λέγεται, πρῶτοι πρὸς ἠῶ καὶ ἡλίου ἀνατολὰς οἰκέουσι ἀνθρώπων τῶν ἐν τῇ Ἀσίῃ Ἰνδοί: Ἰνδῶν γὰρ τὸ πρὸς τὴν ἠῶ ἐρημίη ἐστὶ διὰ τὴν ψάμμον. ’ “To the east of the Indian country is sand. Of all the people of Asia whom we know –even those about whom something is said with precision– the Indians dwell nearest to the dawn and the rising sun; for on the eastern side of India all is desolate because of the sand” (Hdt. 3, 98, 2)19.

It can be argued that all this data were compilated by Herodotus from other sources, as far as he did not travel to India. By this reason, scholars argued that Herodotus compiles or sums up to Scylax and, even more, Hecateus, but this argument cannot be validated as more than a mere speculation, as we have already stressed before. In our own perspective, it seems plausible and adequate to think that Herodotus used other sources. In this sense, Herodotus could add information with accounts that circulated during his times, like a Persian oral testimony, maybe of merchants or mercenaries. Then, scholarship must be as cautious as it can be in this question, and links Herodotus’ work just with his own demonstrate author, Herodotus’ himself, while we have no more evidences to allow us to consider other explanatory options.

It should be noticed in the passage the efforts made by Herodotus in relation to the geographical location of India. For example, an explanation of this kind seems not to be needed in order to establish the location of other regions described in the book, as for example Egypt, as far as the country of the Nile is a famous land. Thus, these facts situate India in the category of the regions of the oikouméne. The arguments that we have exposed reinforce the idea that we have explained previously about the importance of the travel of Scylax for the inclusion of India in the Greek geographical knowledge. On the other hand, the concept used by Herodotus to refer to the oriental area of Indus, the desert (ἐρημίη/eremíe)20 is remarkable. As it is stressed in the modern scholarship, the Greek geographical descriptions frequently describe the extreme regions surrounded of deserts, as a result of the proximity of the sun21. However, the word ἐρημίη/eremíe can also refer an uninhabited regions but not exactly a desert, as we can see in the description of Herodotus of certain regions of Scythia (Hdt 4, 17) and

22

See the excellent commentary of Romm 1992: 35–36, where the autor analyzes the shades of Herodotus between the ἐρεμός (desert) and ἐρεμός αληθός (real desert). 23 See our commentary in Sierra 2012c, focused in the impact of this kind of answer towards the illness in the Greek culture. 24 This is a well known passage, reviewed later by Aristotle (Gen. 736a 10). However, we do not think that this passage can be understood in a restrictive sense: see Sierra 2012b: 55–58, with detailed arguments. 25 A usual feature in Herodotus’ work: Jacob 2008: 85. 26 Costumes that remember us the passage of Hesiod (Op. 275– 280), where he remembers to his brother Perses that Zeus had instaured the laws to give order to humanity, and that eating each other is a characteristic of the beasts and vermins: Sierra 2012c n 41.

19

Greek text and translation from Godley 1920. We must remember the information also recorded by Hecateus (FGrHist 1 F299), and later confirmed again by Herodotus himself (IV. 40). 21 Following the geographical notion of the earth as a plain and that the world keeps a symmetrical structure. See Edelman 1970; Gómez–Espelosín 2000: 97; Jacob 2008: 69 y Dueck 2012: 74–75. 20

56

W h e n G r e e c e c a m e t o In d i a confines of the world there was accumulations of the wealth desired by the Greek audiences (Hdt. 3, 116, 3)30. This idealize wealth consisted of precious minerals or in mild climate conditions that produced unlimited fertility, has a traditional link with the Archaic Greek perspective of the world. We can think in the Garden of the Hesperids, mentioned by Hesiod (Th. 517–520) which produced golden apples that awarded with inmortality, or also Scheria, the island inhabited by the phaeacians in Odissey (4, 1), a land of rare fertility. In both cases, the descriptions showed extreme places associated with fantastic natural and benign conditions for the human beings31. Since the renewed Ionian ethnography and the extension of the geographical knowledge, all these regions were transfered in the ancient Greek imaginary, to very far places, like the India, Ethiopia, and the country of the Hyperboreans32. The result is the combination of real and unreal or fantastic data.

Herodotus and the other India. At the middle of Herodotus’ account on India (Hdt. 3, 102), he begins to explain the amazing way for extracting the Indian gold: some Indians, neighbors of the Indians who inhabited the Pactiyc region, were in charge of the extraction. The desert located beyond Indus was inhabited by some kind of marvellous ants, bigger than a fox but smaller than a dog. These uncanny animals excavated tunnels in the same way as the Greek ants did, forming piles of sand in the surface of the tunnels. Nevertheless, fortunately for the Indians, this sand was rich in gold, so they just have to seize the sand piles of the surface and turn back with this wealth offered by nature. Unfortunately, the was two main problems: the expedition could lost the orientation during the crossing of the desert back home, and the ants were very hostile towards of the human presence. As a consequence, the Indians tried to resolve this situation organizing a caravan of three camels, two male and one female, which have recently had his young. The female was in the middle of the caravan. They marched trying to arrive to the land of the big ants in the hottest hour of the day27, when the ants were less active and looked for shelter from the sun in the tunnels. Once the gold was picked up in a hurry, the caravan began the travel back home, trusting in the speed and guidance of the female camel, guided by her maternal instinct, while the fast ants were held up while devouring the male camels that served as baits (Hdt. 3, 102–105). Herodotus does not defend the veracity of this account but he refers to information about the existence of captured specimens of this kind of ants in the court of the Persian King, information that surely was not contrasted by Herodotus. However, in general no attempt to reinforce the credibility of this episode can be traced in Herodotus’ account.

Despite the fantasy of the account, it seems clear that the “Ants’ gold” has had a kind of connection with the Indian literature. In the Mahabharata (2. 48), writen by Vyasa in the III century BC, there are mentions of gold extracted by ants in a region next to the Himalaya. In this passage, the heir of the Pandava, Yudhistira, obtained gifts from the different regions, where the peculiar gold is included33. Besides, in the Artha–Sastra, a work of contents that goes from economic to political and military information, also written during the III century BC, a reference to the ants’ gold is also found 34. Was this related to the travel of Scylax or was maybe a result of a compilation of information made by Herodotus? We have no clear answer to this, but we can surely notice that the question of the ants’ gold is not only just a paradoxographical curiosity. In our opinion, the account could have had an echo in the Indian culture or became a deformation of an Indian traditional tale. The question has attracted the attention of many scholars, who had made a lot of proposals and associations in order to identify these ants with real animals: marmots, scaly anteaters, Tibetan miners…, but the truth is that this question always seems to be in an impasse35.

Nevertheless, this tale has eclipsed the whole Indian lógos of Herodotus, and it has been easily linked with the paradoxographic nature attributed to the work of Scylax 28. In fact, Herodotus probably may relate this episode from an oral source or maybe following other writing source 29. However, this passage can be considered as an excellent example of the thematic wealth of Herodotus’ work, which, in our opinion, can be understood as a kind of link between the Archaic and Classical Greek thinking. Our position is based on the idea that this kind of ethnography is typical of ancient stages and, moreover it is related with the famous Greek approach which consider that in the

Moreover, the account of this big ants and the extraction of the Indian gold had an extraordinary reception in the ancient Greek literature, surviving during a long time. Strabo records the impressions of Nearchus the Cretan (15, 1, 44), who assures that the skins of these miner ants were similar to those of the leopards. In the same passage he also mentions one version of early Seleucid times

27

This hour did not coincide with the Greek one because of the geographical position of India. See n. 22 (infra) about the geographical conception of Herodotus. This ‘naturalistic’ digressions are also very frequents in his work. 28 Albadalejo 2005: 38. 29 Not only the Quellenforshung has insisted in the analysis of the sources that Herodotus could have used, but there are also studies about the relationship of Herodotus with his contemporaries, like Helanicus of Lesbos, Charon of Lampsacus, etc… See Fowler 1996. Also, a synthesis of the different scholars’ perspectives can be found in Alganza 2012: 31ff.

30

See the commentaries of Karttunen 1989: 122ff. and Albadalejo 2005: 29. 31 This kind of ethnography had a great success among the Greek audience, as can be seen in Diodorus’ work: Sierra 2012a. 32 See Romm 1992: 45ff. 33 An interesting comparative reading between the epic poem Mahabharata and the Greek Archaic epic (Iliad) can be found in Lins–Brandáo 1995 and Wulff 2008. 34 See Karttunen 1989: 172. 35 Karttunen 1989: 171ff.; Albadalejo 2005: 38; Jennison 2005: 190.

57

CÉSAR SIERRA MARTÍN about Megasthenes, who defended that a great Indian tribe lived in a mountainous region at the east of India, wealthy in gold, obtained the gold thanks to some strange ants, bigger than a fox and very aggressive which connects geographically with the data included in the Mahabharata. This Indian tribesmen secretly pick up the gold of the piles of sand accumulated in the entrances of the tunnels. An almost identical version to Strabo’s one is recorded by Arrian (An. 8, 15, 4), although Arrian separates himself a little from the version of Megasthenes and Nearchus (who says that he did not see anyone of these ants but he can observe their skins, as far as someone had brought them to the Macedonian camp). Let’s read Arrian’s opinion about the matter:

audience, promoting that other authors included India in their works. This was actually the case of Hecateus of Miletus, and, although we also have just a few fragments of his work, these seem to show a real ethnographical and geographical content. In this way, if Hecateus could not visit India and used Scylax’s account, we can argue that the information recorded by Scylax was probably similar to the kind usually attributed to Hecateus by the scholars. Then, the view of two Indias arises: one India approached from the Jonian ἰστορίη/historíe with the will of describing with certain degree of detail its geography and ethnography; and other different India approached with a mix of ethnographical and mythical information usually near to fantasy. Thanks to Herodotus and his Indian lógos, we can take into account the expression of this two Indias. In the first part of this lógos, Herodotus shows the information he had considered worth or relevant to be known about India, with data like the geography, the peoples living there or their material culture, and in order to do it he selected the details that can produce a deepest impact in a Greek audience; while in the second part of the lógos, he turns back to the India of the fantasy and the wonders, with references to the large resources of gold. In this context, we have stressed that the possibility that a tale like the one related to the ants’ gold could have had a real echo in the Indian epic literature. In this sense, we consider that despite the adoption of an utopian genre, maybe this is not completely resulted from Herodotus’s imagination and creativity, as far as this kind of ethnography corresponds to the Greek (even, Ancient) understand of the edges of the world, as it is shown in Herodotus’ work, with examples of similar approaches like the cases of Arabia (Hdt. 3, 107) and Ethiopia (Hdt. 3, 114). Thus, following the popular thinking that the extremes of the world were actually in touch, the ethnographic techniques reflected in the description of the confines of the known world (oikouméne) we find a characteristic association of fantastic and ethnographic data, a mix between the archaic descriptions of mythical places and the new, more logical, descriptions purposed by the Jonic ethnography.

ἀλλὰ Μεγασθένης τε ἀκοὴν ἀπηγέεται, καὶ ἐγὼ ὅτι οὐδὲν τούτου ἀτρεκέστερον ἀναγράψαι ἔχω, ἀπίημι ἑκὼν τὸν ὑπὲρ τῶν μυρμήκων λόγον. “Megasthenes, however, merely quotes hearsay, and as I have no certainty to write on the subject, I readily dismiss this subject of ants” (Arr. An. 8, 15, 7)36. It can be seen that, in spite of the pass of the ages, the account recorded by both Strabo (I century BC) and Arrian (II century BC) is extremely similar. Despite the distance of Arrian with the reliability of the information, we guess that both Megasthenes (III century BC) and Nearchus (IV century BC) have been in direct contact with India, something that the most of the Roman authors did not have. Also, this fact serve for keeping alive the Greek ideal about the topics of India, like that of the gold of the miner ants. Thus, the authors who wrote about India were determined to tackle and comment this information. However, as it is well known, the different authors that we have been quoting supra recorded ethnographical data, like Herodotus, showing that the combination of paradoxographical and ethnographical data has a great success during antiquity. Greece and the two Indias.

Finally, we can attribute this dual perspective about India in the Greek thinking to the absence of real, direct and fluent contact between both cultures. Two centuries passed from the travel of Scylax until Alexander’s campaign. At this moment, the utopian India had gained a firm place in the Greek mentalities, and both Strabo and Arrian, who recorded the fragments of Nearchus and Megasthenes, continued to make reference to the same kind of paradoxographical data explained by Herodotus’ account, but including just some geographical precisions. Therefore, we can state that these decades since the end of the IV century BC and the beginning ones of the V century BC were of the main importance to configure the idea about India that remained valid in the Greek thinking for a long, long time.

To sum it up, and in order to make our final remark, we would like to stress that the incorporation of India, to the History and the Greek thinking and culture became with the travel of Scylax across the Indus. Unfortunately, we do not know so much about what Scylax wrote or said, but the limited extant fragments seems to point out that his kind of account about India was strongly paradoxographical. In the preceding lines, we have tried to go away from as far as we can from this topical perspective, because in our opinion the fragments with a clear paradoxographical perspective from Scylax are just a few, and the authors who collected them lived in the Roman Empire, so far from the times of Scylax. Despite these considerations, we must not reject the impact that Scylax’s travel and account could have in the Greek 36

Greek text from Hercher & Eberhard 1885; translation by Robson 1966.

58

W h e n G r e e c e c a m e t o In d i a Bibliography ALBADALEJO, M. 2005, La India en la literatura griega. Un estudio etnográfico, Madrid.

GÓMEZ ESPELOSÍN, F. J. 2000, El descubrimiento del mundo. Geografía y viajeros en la antigua Grecia, Madrid.

ALGANZA, M. 2012, “Hecateo de Mileto, «historiador» y «mitógrafo»”, Flor. Il. 23: 23–44.

HERCHER, R.; EBERHARD, A. 1885, Nicomediensis Scripta Minora, Leipzig.

ASHERI, D.; LLOYD, A.; CORCELLA, A. 2007, A commentary on Herodotus Books I–IV, New York.

Arriani

HORNBLOWER, S. 2002, “Herodotus and His Sources of Information”, in E. J. BAKKER, I. J. F. DE JONG & H. VAN WEES (eds.), Brill’s Companion to Herodotus, Leiden, 373–386.

BARAGWANATH, E. 2008, Motivation and Narrative in Herodotus, New York. CASEVITZ, M. 1995, “L’Inde des poètes grecs, d’Eschyle à Nonnos”, in J. C. CARRIERE, E. GENY, M. M. MACTOUX & F. PAUL–LEVY (eds.), Inde, Grèce Ancienne. Regards croisés en anthropologie de l’espace, Paris, 9–19.

JACOB, CH. 2008 [1991], Geografía y etnografía en la Grecia Antigua, Barcelona.

DUECK, D. 2012, Geography in Classical Antiquity, New York. EDELMAN, H. 1970, “Eremos und eremie bei Herodot”, Klio 52, 79–86.

LINS BRANDAO, J. 1995, “Le mode narrative d’Homère et du Mahabharata”, in J. C. CARRIERE, E. GENY, M. M. MACTOUX & F. PAUL–LEVY (eds.), Inde, Grèce Ancienne. Regards croisés en anthropologie de l’espace, Paris, 139–151.

FEHLING D. 1989 [1971], Herodotus and His ‘Sources’. Citation, Invention and Narrative Art, Leeds.

MAZON, P. 1976, Eschyle. Les Suppliantes, Les Perses, Les Sept contre Thèbes, Prométhée Enchainé, Paris.

FERRARI, F. 1995, Eschilo. Persiani, Sette contro tebe, Supplici, Milano.

ROMM, J. S. 1992, The Edges of the Earth in Ancient Thought, Princeton.

FOWLER, R. L. 1996, “Herodotos Contemporaries”, JHS 116, 62–87.

SIERRA, C. 2012α, “ἌΝΘΡΩΠΟΣ ΚΑΙ ΚΟΣΜΟΣ: Diodoro Sículo y la etnografía clásica”, Maia 64/3, 445– 458.

and

JENNISON, G. 2005, Animals for show and Pleasure in Ancient Rome, Philadelphia.

His

HOW, W. W.; WELLS, J. 1967, A Commentary on Herodotus, vol. 1, Oxford.

– 2012b, “Diferentes cuerpos, diferentes pueblos: algunos ejemplos en las fuentes históricas”, Habis 43, 47–62.

KARTTUNEN, K. 1989, India in Early Greek Literature, Helsinki.

– 2012c, “El Heródoto nosológico”, REA 114/2, 387–404.

GIL, J. 1995, La India y el Catay. Textos de la Antigüedad clásica y del Medievo occidental, Madrid.

SMYTH, H. W. 1973, Aeschylus. Suppliant maidens, Persians, Prometheus, Seven against Thebes. Cambridge (Mass.).

GISINGER, F. 1972, “Skylax”, RE V2, col. 619–646.

STEPHANUS, H. 1954, Thesaurus Graecae Linguae. vol. 5, Graz.

GODLEY, A. D. 1920, Herodotus, with an English Translation, Cambridge (Mass.)

WULFF. F. 2008, Grecia en la India: el repertorio griego del Mahabharata, Madrid.

59

Ancient vardāna Results of the last archaeological investigations SILVIA POZZI ‘L'Orientale’ – University of Naples ABSTRACT Vardāna is an early medieval settlement located in the northern frontier of the Bukhāra oasis (Uzbekistan), which is currently under archaeological investigation by a Swiss–Uzbek expedition. Describing this settlement, the Arab chronicles report that its ruler held a relevant role especially at the time of the Arab conquest. The report, now supported by archaeological evidence, seems to suggest strict political and cultural contacts between Vardāna and the nomadic populations living in the northern steppe. This paper presents the architectural remains recovered during the last campaigns, in light of the main historical events occurred at the time of the Arab conquest. KEYWORDS Sogdiana – Early Medieval Period – Vardān Khudāh – Türks – Arab Conquest – Architecture At present time, most of the ancient settlements forming the actual landscape of the Bukhāra oasis (Uzbekistan) consist in ruins buried by sand, which often bear modern appellations due to the physical features of the remains 1. The archaeological site locally known as Vardanzeh (or Kurgan Vardanzeh) is one of the few to emerge from the anonymity and to transmit in the current toponym traces of the name used in the Islamic chronicles, at least since the 9th century AD. In fact, this place has been identified with the ancient Vardāna, one of the many villages of the Bukhara oasis inhabited during the early medieval period2 (Fig. 1).

Vardāna should have been rather important because it is one of the few sites for which Naršakhī3, compiler of the Ta’rīkh–i Bukhārā (‘The History of Bukhāra’), has provided the name, or at least the title, of a pre–Islamic ruler. There are several mentions of the involvement of Vardāna’s ruler, named or titled Vardān Khudāh, in the tumultuous years of conquest of the Bukhāra oasis by the Arabs. Naršakhī reports that the founder of Vardāna4 was Šāpūr, a Sasanian prince fled from his homeland and welcomed at the Bukhār Khudāh court. Indeed, the toponym seems to confirm the Iranian origin. M. Compareti 5 proposes several interpretations based on comparisons with the ancient Iranian languages in his etymological study. The meaning is consistently an enclosed space, like a garden or a village6. These interpretations certainly fit well with the foundation legend of this site, which is described as the wild and uncultivated land where Šāpūr used to hunt. However, other interpretations can be suggested. From the Avestan var comes the meaning “castle”7, which would be very appropriate for this site, dominated by a huge manor. Moreover, strict similarities are found with the name Wardan8, of Iranian origin, attributed to at least 3

Fig. 1: Map of Bukhara oasis (Lo Muzio 2009).

1

Nil’sen 1966, 111–112. For an historical introduction on Vardāna, see Adylov and Mirzaachmedov 2001. 2

Frye 1954, 12-19. Frye 1954, 16, 31–32. 5 Compareti 2003. 6 Compareti 2003; see also Kent 1950, 207, § vardana. 7 Bartholomae 1904, 1363, 5var–. I thank for this suggestion and bibliographic help of Prof. Agustí Alemany (Universitat Autònoma de Barcelona). 8 Justi 1895, 351–353, § Wardān. 4

SILVIA POZZI two Parthian kings9, as well as to the Mamikonean prince of Armenia, who was martyred by the Sasanian king Yazdgerd in the 5th century AD 10. The personal name Vardanes also appears on some Sasanian seals11.

could result rather realistic if ‘Turkistan’ (the land inhabited by Türks) can be considered a close territory23.

We know from Naršakhī12 that Šāpūr built a canal, named Šāpūrkām, some villages and a palace. Then the locality took the name of ‘the villages of Abūya’13. This description suggests the presence of a vast province, whose capital may have been Vardāna. Its importance can be inferred by the likely abundance of crops, capable of sustaining many surrounding villages, as well as by the strategic trading location. Vardāna was located beyond the Kampir Duwal14, the earthen wall which surrounded the Bukhāra oasis, just on the border with the steppe where the nomadic populations circulated. This settlement welcomed a well–furnished weekly market15 and local traders periodically brought their products to Bukhāra and to other Sogdian centres16. A commercial vocation for Vardāna can be proposed on the basis that Sogdian markets specialized in trading with the steppe were particularly important17 and the geographical position of Vardāna suggests this to have been the case. Moreover, Naršakhī reports about a well–made zandaniji fabric, which was locally produced. This piece of information could be of some importance because the barter economy paralleling the formal trade economy employed fabrics 18 as currency in place of coins.

Fig. 2: Vardāna, plan of the citadel (O. Cerasuolo & S. Mirzaachmedov).

As far as the history of Vardāna is concerned, the period of the Arab conquest by Qutaiba b. Muslim (705–715) is the most well–documented by the historical sources. Besides Naršakhī, many details of the military events involving Vardān Khudāh have been summarised by Țabarī24. According to the last one, in the year 89 (709) Qutaiba received from al– Ḥajjāj the precise order to direct a military offensive against Vardān Khudāh, addressed as the king of Bukhāra, who was evidently considered very dangerous25. Qutaiba entered in Bukhāra oasis and stopped at Lower Kharqānah, a district confining with the one governed by Vardān Khudāh 26. Here the battle lasted two days and ended with the victory of the Arab army, without the expected invasion of the Vardān Khuda district taking place. That occurrence was rectified by al–Ḥajjāj, who ordered Qutaiba to go back to Sogdiana and defeat definitely Vardān Khudāh 27. The following year, Qutaiba was again in Bukhāra and fought against a coalition of Türks and Sogdians, whose help had been requested by Vardān Khudāh himself28. Naršakhī,

Naršakhī suggests the existence of a privileged relationship with the nomadic neighbours of the north. Despite the legendary Sasanian lineage of the dynasty, he says that Vardān Khudāh was of Türk origin19. This is not surprising, since Bukhara has been included into the Western Türk Kaghanate20 from the second half of the 6th century AD. One of the main effects of the Türk domination was a system of political alliances and the fusion between Sogdians and Türks, particularly evident in the ruling elites21. As far as the provenience of Vardān Khudāh is concerned, S. Stark22 suggested that it has to be found in the Türk clans living in the steppe immediately northward the Bukhāra oasis. In this respect, Naršakhī reports that during the Arab offensive this ruler took refuge in Turkistan several times, information which

9

Justi 1895, 351, § Wardān 2– 3. Justi 1895, 351, § Wardān 7. On the history of Vardān, see also Elishe 1982. 11 Bivar 1969, 67, cat. CG 14; Gignoux, Gyselen 1982, 67, (cat. 20.17). 12 Frye 1954, 32. 13 Frye 1954, 32. 14 On this argument, see Adylov 1995; Stark and Mirzaachmedov 2012. 15 Frye 1954, 16. 16 Smirnova 1970, 149. 17 Smirnova 1970, 139, 148. 18 Smirnova 1970, 155–156. 19 Frye 1954, 10. 20 On the history of Turks in Central Asia, see Stark 2008. 21 De la Vaissière 2005, 200–202. 22 Stark 2008, 228–229. 10

23

An example of neighborhood between sedentary and nomadic population is represented by Taraz, a Sogdian colony in Semireč’e region (Kazakhstan). According to Ibn Hauqal (1964, 489), this city was an important trading post frequented by Türks since the early medieval period. The territory adjacent to the city was protected by some fortresses and the Türks lived just outside this fortified area. 24 Țabarī 1990. On the conquest of Bukhara oasis, see in particular the chapters on the events happened in the years 87, 88, 89, 90. 25 Țabarī 1990, 147. 26 Țabarī 1990, 147. 27 Țabarī 1990, 147. 28 Țabarī 1990, 150.

62

Ancient Vardanā even though he does not attribute this initiative to Vardān Khudāh, offers a precise account of the involved parties and of the military strategies adopted in that occasion 29. Advancing towards Bukhāra, Qutaiba was surrounded by many troops in a rural area west of the city. Among them, there was Tarkhūm, the ruler of Sogdiana, a certain Khunak Khudāh, Vardān Khudāh and Kūr Maghānūm, a Türk king who is said to be the nephew of the Chinese emperor. On that occasion, the struggle ended with a strategic pact between Qutaiba and Tarkhūm, avoiding for some time the occupation of Bukhāra. Only a few months later Bukhāra was finally seized. Except for the previously mentioned struggles, we do not know much more about the many battles cited by Naršakhī 30 that were fought between Qutaiba and Vardān Khudāh. What seems to be certain is that Qutaiba was able to establish Bukhār Khudāh Toghshāda on the throne of Bukhara 31 only after the death of Vardān Khudāh.

Naršakhī hints that Vardān Khudāh usurped the throne of Bukhār Khudāh at the time of Qutaiba’s conquest32. Moreover, after comparing numismatic and literary evidence, A. Naymark33 concluded that Vardān Khudāh has to be identified with Khnk, or Khnk Khudāh, whose name appears on the legends of Bukhār Khudāh drachms dated to the end of the 7th – beginning of the 8th centuries AD. The same person is mentioned also by Naršakhī as the restorer of the palace of Varakhša34. At Vardāna, archaeological evidence dating to the years of the Arab conquest have been unhearted during recent excavations carried out in the citadel35 (Figs. 2–3). According to the preliminary stratigraphic results, the architectural structures belong to a palace whose stable occupation lasted up to the late 8th century AD. Its foundation has to be found around the 6th century AD or a little earlier, in line with the widespread building activity which characterizes the Bukhāra oasis and in general the whole Sogdiana from the 5th century AD onward36.

The role of Vardān Khudāh which emerges from these passages is not that of a petty ruler in the variegate panorama of principalities in the Bukhāra oasis. In fact,

Fig. 3 Vardāna, plan of the excavated structures (Photo: S. Mirzaachmedov). 32

Frye 1954, 10, 32. Naymark 2010. 34 Frye 1954, 17. 35 The Swiss–Uzbek archaeological expedition at Vardanzeh, sponsored by the ‘Society for the exploration of EurAsia’ is directed by the present author and Dž. K. Mirzaachmedov. Archaeological reports are available at http://www.exploration– eurasia.com/EurAsia/inhalt_english/frameset_projekt_0.html 36 Grenet 1996, 372. 33

29

Frye 1954, 45. Frye 1954, 10, 32, 45. 31 Frye 1954, 32. 30

63

SILVIA POZZI As testified in other regional centres 37, this palace was also built on top of ancient structures probably dated back to the 4th–5th centuries AD. Traces of this ancient phase were found under an interlayer of fluvial pebbles and sand 3 meters thick. We do not know yet if the whole area of the citadel pertaining to this old phase was levelled using this system, which assumes the vicinity of a river38. However, the well– preserved state of the ancient structure and the lack of evident phases of abandon suggests that no significant interruption of occupation between these two important constructive periods occurred. Huge quantities of fluvial pebbles result quite unusual in civil building but some analogies can be found in military architecture. Recent investigations at Kizil–tepa39 (eastern border of Bukhāra oasis), focused on a preserved portion of Kampir Duwal, the mudbrick wall that surrounds the whole oasis. Presence of pebble and sand was found in the core of the structure which has been dated to the 4th– beginning of the 5th century AD40 At the present state of research, only the northern part of the early medieval palace has been recovered yet (Fig. 3). An external wall 2.7 meters thick represents the northernmost mural structure (Fig. 4). The excavated portion, 10 meters long, has been preserved at a height of 1 meter and its consistent thickness is compatible with some defensive purposes. The wall was made of big mudbricks measuring 43×26×12 centimetres and the southern face still shows traces of mud plaster. The northern face is reinforced by layers of rammed hearth (pakhsa) alternated with sporadic mudbricks, spread up to the edge of the citadel. At 4 meters south from this external wall stands another parallel wall, 35 meters long and 1.30 meter thick. Only a brief portion of the space between the two walls has been excavated yet, not enough to understand if the two structures were separated by a wide and empty corridor.

Fig. 4: Vardāna, external wall from north–west (Photo: S. Mirzaachmedov).

To the south from the innermost wall there are several rooms, probably used as service spaces. Starting from the east, three aligned rooms use the southern face of the wall as northern perimeter (Fig. 5). Rooms n° 4 and 9, which originally formed an unique room, have an entrance on the southern side. Room n° 8 is characterized on its western side by a passage which leads into a corridor oriented east–west. Another corridor, oriented north– south, intersects the previous one forming a ‘T’ shaped system of corridors which were partially obstructed at the beginning of the 8th century AD. West of the corridor oriented north–south there are other three aligned rooms (Fig. 6). Room n° 6, characterized by a partially preserved sufa on three sides, opens directly into the east–west corridor. Rooms n° 10 and 11 originally formed a single squared room which was subsequently divided into two narrower spaces by a partition wall. To the south from these rooms, whose southern walls result partially covered by a medieval furnace (SU 186), there is an open wide space (courtyard?).

At the present state of research, only the northern part of the early medieval palace has been recovered yet (Fig. 3). An external wall 2.7 meters thick represents the northernmost mural structure (Fig. 4). The excavated portion, 10 meters long, has been preserved at a height of 1 meter and its consistent thickness is compatible with some defensive purposes. The wall was made of big mudbricks measuring 43×26×12 centimetres and the southern face still shows traces of mud plaster. The northern face is reinforced by layers of rammed hearth (pakhsa) alternated with sporadic mudbricks, spread up to the edge of the citadel. At 4 meters south from this external wall stands another parallel wall, 35 meters long and 1.30 meter thick. Only a brief portion of the space between the two walls has been excavated yet, not enough to understand if the two structures were separated by a wide and empty corridor.

Fig. 5: Vardāna, view of the excavated structures from the East (photo S. Mirzaachmedov).

In light of the available information, some reconstructions that slightly changed the initial layout of the palace probably occurred at the end of the 7th – beginning of the 8th centuries AD. This dating, obtained using ceramics, seems to be supported also by bronze ornaments of

37

Silvi Antonini et alii 2010, 59–73. According to Ibn Hauqal (1965, 466–467), the canal named Shafari–Kam (Šapurkam in Naršakhī) arrived up to Vardāna. 39 Stark and Mirzaachmedov 2012, 196–197, 201. 40 Stark and Mirzaachmedov 2012, 201. 38

64

Ancient Vardanā various kind (rings, bracelets, earrings), ivory objects 41 and two Sasanian–style sealstones42. Of great importance is also an iron arrowhead found under wall n° 188, a bad preserved structure which partially covers the partition wall between rooms n° 10 and 11. The arrowhead can be compared with those from Penjikent 43 and is dated to the beginning of the 8th century as well.

Although these traces are still too scarce to formulate a definitive interpretation, we can attempt to reconstruct what happened in those years following known historical events. Considering that Vardān Khudāh took refuge in Turkistan many times, it is probable that the castle of Vardāna was no longer able to offer security from a military perspective. As a result, the obstruction of the corridors could have been an attempt to implement the defensive system through the isolation of the core of the palace. Moreover, following the hypothesis of Naymark (see above), we can suppose that after the restoration of Varakhša palace, Vardān Khudāh could have been moved there. Presumably, he maintained both palaces, even though his court may have moved to the most prestigious one. At Vardāna, the re–arrangement of the excavated spaces is not due to an improvement program for the royal residence, which could be expected for the palace of one of the main political rulers of those years. On the contrary, the abandonment of some rooms suggests a possible reduction of the people living in the castle. We hope that future investigations will provide more evidence to reconstruct the history of Vardāna especially during the years of the Arab conquest, a period in which this settlement and its ruler have probably played a crucial role in the history of the whole Bukhāra oasis.

Fig. 6: Vardāna, view of the excavated structures from the north (photo S. Mirzaachmedov).

The architectural remains pertaining to this phase suggest a reduction of the spaces, as evidenced by the abandonment of some rooms on the east–west corridor.

41

Mirzaachmedov and Pozzi 2013. Pozzi (forthcoming). 43 Raspopova 1999, figs. 14–15. 42

65

SILVIA POZZI Bibliography ADYLOV, Š. T. 1995, “O datirovke steny ‘Kampirak’ i nekatorix voprosax politiceskoi istotii zapadnogo Sogda”, in Markazii Osioda Urbanizaziya Sharayonining Paido Bulishi Va Rivoshlanish Boskichlari, Samarkand, 45–47.

LA VAISSIÈRE, E. 2005, Sogdian Traders: a History, Leiden. LO MUZIO, C. 2009, “Note archeologiche sull’oasi di Bukhara”, in C. SILVI ANTONINI & DŽ. K. MIRZAACHMEDOV (eds.), Gli Scavi di Uch Kulakh (Oasi di Bukhara), Rapporto Preliminare, 1997–2007, Rivista degli Studi Orientali [Supp. 1] 80, 17–40..

ADYLOV, Š. T., MIRZAACHMEDOV, DŽ. K. 2008, “Novye svedenija po istorii obšivanija Varakhšinskogo massiva”, Istorija Material’noj Kul’tury Uzbekistan 36, 61–78. – 2001, “Iz istorii drevnego goroda Vardana i vladenija Obavija”, Istorija Material’noj Kul’tury Uzbekistan 32, 150–157.

MIRZAACHMEDOV, DŽ.; POZZI, S. 2013, Report on the fourth archaeological excavation at Vardanze (ancient Vardāna), by the West Sogdian Archaeological expedition, available at http://www.exploration– eurasia.com/EurAsia/inhalt_english/frameset_projekt_0.ht ml

BARTHOLOMAE, C. 1904, Altiranisches Wörterbuch, Strassburg. BIVAR, A. D. H. 1969, Catalogue of the Western Asiatic seals in the British Museum. Stamp Seals. II, The Sassanian Dynasty, London.

MIRZAACHMEDOV, DŽ.; STARK, S. 2012, “Raboty uzbecko–amerikanskoj ekspedicii po izučeniju oazisnoj steny Buchary Kampir–Devor v 2011 g”, Archeologičeskie issledovanija v Uzbekistane, 2010–2011 g. 8, 191–202.

COMPARETI, M. 2002, “Rapporto delle campagne di scavo 2002 in Uzbekistan, C. Note sul toponimo Vardana– Vardanzi”, Rivista degli Studi Orientali LXXVI, 39–47.

NAYMARK, A. 2010, “Drachms of Bukhār Khudā Khunak”, Journal of Inner Asian art and Archaeology 5, 7–32.

ELISHE 1982, History of Vardan and the Armenian war, translation and commentary by Robert W. Thomson, Cambridge (Mass.).

NIL’SEN, V. A. 1966, Stanovlenie feodal’noj architektury Srednej Azii (V–VIII vv.), Tashkent.

FRYE, R. N. 1954, The History of Bukhara, transl. from a Persian Abridgment of the Arabic original by Narshakhī, Cambridge (Mass.).

POZZI, S. (forthcoming), “Sasanian–style sealstones from Bukhara oasis (Uzbekistan)”, Rivista di Studi Orientali.

GIGNOUX, P.; GYSELEN, R. 1982, Sceaux Sasanides de diverses collections privées, Leuven.

RASPOPOVA, V. I. 1999, Metalličeskie izdelia iz Pendjikenta, S. Peterbug.

GRENET, F. 1996, “Crise et sortie de crise en Bactriane– Sogdiane au IV–V siècles: de l’héritage antique a l’adoption de modèles sassanides”, in La Persia e l’Asia Centrale da Alessandro al X secolo, Roma, 367–390.

SEMENOV, G. L. 1989, “Sogdijskij gorod v rannem srednevekov’e: formirovanie plana”, in Itogi rabot archeologičeskich echspedizij, Leningrad, 128–138.

HORN, P. 1893, Grundriss der Neupersischen Etymologie, Strassburg.

SILVI ANTONINI, C.; FILIPPONI, F.; IVANOV, G.; ADYLOV, Š. T.; MIRZAACHMEDOV, DŽ. K.; LO MUZIO, C. 2009, “Gli scavi di Uch Kulakh (Uzbekistan). Campagne 2008– 2009”, Rivista di Studi Orientali 82, 60–95.

IBN HAUQAL 1965, Configuration de la Terre (Kitab Surat al–Ard). Introduction et traduction, avec index par J. H. Kramers et G. Wiet, vol. II, Paris.

SMIRNOVA, O. I. 1970, Očerki iz istorii Sogda, Moskva.

JUSTI, F. 1895, Iranisches Namenbuch, Marburg.

STARK, S. 2008, Die Alttürkenzeit in Mittel– und Zentralasien: archäologische und historische Studien, Wiesband.

KENT, R. G. 1950, Old Persian, Grammar, Text, Lexicon, New Haven, Conn.

TABARĪ 1990, The Zenith of Marwānid House, Albany (N.Y.).

66

The Italian–Uzbek archaeological mission at Uch Kulakh tepe ILARIA VINCENZI Universitat Autònoma de Barcelona ABSTRACT The work carried out by the joint Italian–Uzbek archaeological mission at Uch Kulakh tepe, located in the western side of the Oasis of Bukhāra, brought to light a very interesting Early Medieval settlement. The archaeological monument with its own peculiarities reflects the urban development occurred during that Age in Western Central Asia. The archaeological data seems to suggest a very strong contact and relationship with the nomadic groups that were crossing the steppe of Central Asia. The archaeological investigation is currently under the direction of a joint Hispanic–Uzbek mission. KEYWORDS Archaeological investigation – Oasis of Bukhāra – Sogdiana – Early Middle Ages – Urban Planning As a member of the archaeological research project La Sogdiana dall'Antichità al periodo Alto–Medioevale, I would like to draw your attention on the archaeological work carried out by the joint Italian–Uzbek archaeological mission at the Early Medieval monument of Uch Kulakh tepe (Oasis of Bukhāra, Uzbekistan) (Fig. 1), during the years 1997–2009, under the scientific direction of Chiara Silvi Antonini (“Sapienza” Università di Roma).  Before moving to the description of the area and of the archaeological monument it is necessary to emphasize the importance that archaeology has, when talking about Central Asia and in particular about Sogdiana or the Oasis of Bukhāra, for the reconstruction of the historical events just due to the absence of a local historiographical literature, at the least up to the period immediately before Islam; for this area we don't have direct written sources. The first explicit evidence about the history of Sogdiana and the people living there before the advent of Islam has come to us through archaeological research, since indirect written sources offer little information about it. Among the historical texts, we should remember the mention of Sogdiana found in the Avesta (Vidēvdāt 1.4; Yašt 10.14)1, 

The Italian archaeological mission has been working in the Oasis of Bukhāra from 1997 to 2009. The results of the excavation activities carried out until 2007 were published in Silvi Antonini & Mirzaachmedov 2009. See also: Silvi Antonini, Noci & Lo Muzio 1996 ; Silvi Antonini & Noci 1999, 221–244; Silvi Antonini 2001, 248–258; Silvi Antonini & Filipponi 2003, 17–26; Filipponi 2006, 47–49; Silvi Antonini 2010, 157–168; Silvi Antonini, Filipponi, Ivanov et alii 2010 , 59–95. 1 In the Avesta there is a reference to Sogdiana (Av. Suγδa, OPers. Sug[u]da) in the list of the lands, with a mention to that

the cuneiform inscriptions of the Achaemenid period: the inscriptions of Darius I (late 6th–5th cent. BC) in Bīsutūn (DB I 16), Hamadan (DH 5), Naqš–e Rostam (DNa 23), Persepolis (DPe 16, DPh 6) and Susa (DSf 38, DSm 9) 2, or the hieroglyphic inscription on the pedestal of the statue of Darius I found in Susa and those of the stems found in Egypt (5th–4th cent. BC) and, afterwards, the inscription of the Sasanian king Shapur I (3 rd cent. AD) in the so–called Ka‘be–ye Zardošt (ŠKZ MPers. 3, Parth. 2, Gr. 4)3. Among the classical authors who dealt with the area we can mention Arrian4, Curtius Rufus5, Plutarch6, Strabo7 and Ptolemy8; however, these are sources, historical documents and fragmentary geographic descriptions relating to the expedition of Alexander the Great. Sogdiana was known by Greek authors as Σογδιανή and located in the area of Central Asia beyond the Oxus. For the period prior to the Islamic invasion a lot of information was gathered by Arabic historians as Naršakhī9, al–Ṭabarī10, and by those of the Arabic geographers whose descriptions are collected and published in the Bibliotheca Geographorum Arabicorum. Without neglecting the Chinese sources such as the territory located between Merv and Chorasmia. Tremblay 2004, 133. 2 Briant 1997, 27–41; See also: Frye 1943, 14–16; Vallat, 1997, 48–55. 3 Lukonin 1969, 42,126. 4 Arr. An. III 8.28, 30; IV 1, 5; 15–18; V 12; VII 6. 5 Curt. VII 10.1; 11.1; VIII 1. 35. 6 Plut. Mor. IV 5; IX 3. 7 Strab. XI 4; XI 11.2. 8 Ptol. Geog. VI 12.3; XI.1; XII 1.5; XIII 1.2; XIV 1. 9 Frye 1954. 10 Yar–Shater 1996, 1987–1997.

ILARIA VINCENZI culture brought to light by archaeological research 14. That would be the beginning of the amalgamation process with those sedentary cultures that had settled along the banks of the Zerafšan and its tributaries.

Annals of the Chinese dynasties which include the events from the 2nd cent. BC onwards and the travel reports by Chinese pilgrims, or the Old Turkic sources, perhaps less extensive but certainly significant; it is worth to remember the inscriptions in Sogdian found at Kultobe 11, located in south Kazakhstan, the administrative documents in Sogdian from Nisa (Turkemenistan) and Mount Mug12, east of Samarkand in present–day Tadžikistan, and the ‘Sogdian ancient Letters’13 found near the site of Dunhuang.

Fig. 2: Map of the Oasis of Bukhāra (A. Petrocchi).

Nevertheless, Sogdiana was subject to the Achaemenid Empire, and consequently it also entered the orbit of Alexander the Great’s successors, but the Oasis of Bukhāra, as both archaeological evidence and historical sources of the region show, always maintained a marginal and autonomous position in comparison with all those historical events, of great international scope, that involved the same Sogdiana (a distinction between Sogdiana of Samarkand and Sogdiana of Bukhāra should be considered). This fact would be a consequence of the amalgamation of two great traditions that have equally contributed to the formation of a cultural landscape which reached the peak during the Early Middle Ages.

Fig. 1: Map of Sogdiana (A. Petrocchi).

The good knowledge of written sources would not be enough without the contribution of the data provided by archaeological research. Thus, archaeology becomes an essential tool, useful to reconstruct the chronological sequence of the correspondences of the events, owing to the lack of written sources. It must be, then, stressed, with regard to Sogdiana, that archaeology and history are two closely connected subjects: it is through the discovery of archaeological remains that it is possible to find correspondences in the chronological sequence of events that have characterized history of mankind.

The period between mid 4th and 5th cent. AD was characterized by important upheavals for Central Asia. New waves of nomadic tribes, which we know with the name of Kidarites, Hephthalites and Türks15, from the Eurasian steppes, disrupted the economic status of the sedentary peoples in different regions of Transoxiana. The situation changed for Sogdiana in the mid 5 th cent. when classical and Chinese sources16 refer that the region

Geographical and historical remarks on the area The archaeological site of Uch Kulakh is located in the western part of the Oasis of Bukhāra, in that region known before the Arabic conquest as Sogdiana, which nowadays is located in central Uzbekistan and western Tadžikistan, between the course of the Oxus (Amu–daryā) to the west and the Iaxartes (Syr–daryā) to the east.

14

Askarov 1965, 53–60; 1981, 99–110; see also Ruzanov 1998 , 141. 15 The Hephthalites put end to the rule of their predecessors, the Kidarites. Under the Hephthalites trade and commerce continued to flourish. In the archaeological sites investigated in the area, the level related to the 5th–6th century is that of the Hephthalites. They were annihilated by the coming up of a new group of nomads: the Türks, who retained control of the region from the mid 6th cent. down to 673 AD, when the Arabs entered the Oasis of Bukhāra. The governor of Bukhāra at the time was the widow of Bidun or Bandun Bukhār Khudāh, as regent for his son Tuġšāda. In 710 AD Qutaiba b. Muslim, after conquering the city of Paykend and defeating his enemies, was installed as ruler of Bukhāra. Frye 1954, 43–48. See also Bartol’d 1968,5– 10; Cannata 1981, 17–19; Frye 1999, 169–181. 16 The Roman historian Ammianus Marcellinus refers to the Chionites first fighting against Saphur II in 356 AD, and two years later concluding an alliance: Amm. XVI.9.4; XIX.1.7. The Chinese text of the Weishu refers that in 457 AD a group, maybe descendant of Xiongnu, occupied the throne of Samarkand. Probably this group could be identified with the

In the Oasis of Bukhāra the environmental geography had a very decisive role in the historical evolution: the presence of the river Zerafšan, the Πολυτίμητος of the Greek sources (Arr. IV.5.6; 6.5; 6.7; Aristob. FGrH 139 F28a; Ptol. VI.14.2; Curt. VII.10.2), and the geomorphology of the soil, rich of loess deposits, provided this strip of land all the essential features to develop large–scale agriculture: from the 3rd millennium BC groups of nomadic and transient herdsmen, probably descendants of the Andronovo culture, began to penetrate this area, as evidenced by various elements of material

11

Sims–Williams & Grenet 2006, 95–11; Sims–Williams, Grenet & Podushkin 2009 , 1025–1033. 12 Livsič 1962. 13 Reichelt 1931.

68

The Italian–Uzbek archaeological mission at Uch Kulakh tepe fell under the rule of a group identified as Chionites or Kidarites. The presence of Kidarites in Sogdiana is indicated by the discovery of seven coins dating back to the 5th cent. and minted in Samarkand: on the observe there is a representation of an archer while on the reverse it is possible read a legend κγδr, that is Kidara17. For the Sogdiana the period under the Kidarites was a time of great economic recovery, as evidenced by the development of urban civilization, agriculture, and the revival of centres located along the main commercials roads such as Bukhāra, Varakhša, Paykend, Samarkand and Penjikent18.

that along the route that linked Rome to China ‒the Oasis of Bukhāra was just in the middle, just the heart of this caravan route track‒ trade involving all types of goods was certainly important, as well as the transmission of ideas, religious beliefs, and artistic expressions. Uch Kulakh tepe The site of Uch Kulakh (Fig. 2) is located in the north– western area of the Bukhāra Oasis, in the Varakhša territory about 40 km south–west from the city of Bukhāra and only a few kilometres, more or less 8 km, from the site of Varakhša, which was from the 7 th cent. the residence of the pre–islamic dynasty Bukhār Khudāh (5th‒8th cent. AD) and the last important urban centre on the road which across the Kizil Kum desert led into Chorasmia.

It was during this period that the phenomenon identified as Central Asian Feudalism manifested that is the Early Middle Ages (5th–8th cent.) the moment in which we can chronologically ascribe the rise and growth of Uch Kulakh tepe. In this period the urban physiognomy of the landscape changes and right beside the cities castles were built. This new architectural reality is characterized, as the Arabic historians describe it, by a social model which provided a king and an aristocracy composed of landowners acting as vassals; they hold in their hands a big portion of the economic power and, once settled in the culture areas, they built the castles that served, on the one hand, as a refuge against enemy attacks and, on the other, as normal residence. As Naršakhī19 refers to us, the lords of the Oasis of Bukhāra for this period were the Bukhār Khudāh (5th‒8th cent.), the last pre–islamic potentates of Western Sogdiana.

C. Silvi Antonini, after a survey of the area of the Oasis, choose the site of Uch Kulakh basically for two reasons: on the one hand, the average size of the tepe (85m width, 75m length and 8m height); and, on the other, the proximity to Varakhša. So everything seemed to point that it could be one of the 700 castles, the residence of a dehqān, present in the Oasis at the time of Qutaiba's conquest, as reported by Naršakhī in the pages of his History of Bukhāra20. The settlement consists of a hill of artificial origin (tepe), originated from the stratification of the structures in terra cruda (mud building technique) and their subsequent abandonment. The profile of the tepe showed three protuberances aligned on the western area, from which the name given to the site comes, since in modern Uzbek language it means “the three hats”. The south–western side of the tepe has a dip, where it could have been situated an access road to the castle, while the different sides of the tepe go down rather uniformly towards the plain below. The tepe covers an area of about 3400m²; the excavations carried out between the 1997–2009 archaeological seasons affected just 50% of the surface of the tepe; although they began with the exploration of the central part of the tepe, the work progressively extended to the western, southern and eastern zones. The main structure identified is a fortified castle equipped with two bastions, placed one on the south–western corner and the other on the south–east, with loopholes; two residential areas outside of the castle, one located on the western side and the other on the eastern side, both had very important representation rooms with paintings on the walls; and finally a fortified southern complex, was also discovered outside the castle.

Fig. 3: Plan of Uch Kulakh Phase VII (4th–5th): Silvi Antonini 2010, 60.

It is also worth remembering that this geographical area was crossed by the great Silk Road, always considered a network of cultural interactions. And it is understandable

The occupation can be dated from the 4 th‒5th to the early 13th cent. AD. In this long period of time, seven phases of building activity have been documented; their chronology has been established on the basis of the archaeological data coming from the stratigraphic sequence; on one

Kidariti. Zeimal’ 1996, 120. See also de La Vaissière 2004, 100–101 = 2005, 95–97. 17 Zeimal’1978, 208. 18 Frye 1999, 173. 19 Frye 1954, 60–65.

20

69

Silvi Antonini, Noci & Lo Muzio 1996, 422.

ILARIA VINCENZI hand, pottery21, terracotta figurines22 and consistent fragments of wall paintings23, on the other the material and the building techniques24.

walls with loopholes, probably to defend the access road to the castle itself, seems to give support to the idea.

In the period between the 4th and 8th cent. the settlement had two living areas to the east and west of the castle, and in the southern part defensive structures have been identified (other walls with loopholes), built later than the system of the castle itself. But after the early 8 th cent., however, the site lost its original function and was only occasionally and partially reoccupied. The Castle25 The complex is located in the central part of the tepe and is characterized, in its northern and central areas, by a series of rooms connected by entrances and passages, while the southern part was kept for defensive purposes, as evidenced by the presence of rectangular angular towers, equipped with loopholes arranged in a vertical position in front and oblique at the corners 26. It formed a chessboard on the surface of the structures, to which rooms related to a defensive system were attached.

Fig. 4: M36 – Western wall (F. Filipponi).

Between the 5th and the 8th cent. the structure of the castle was not substantially modified, despite the construction of two new walls with loopholes (M8 and M13); building activities focused on the fortified southern complex, that would be progressively expanded to the east, with the addition of new walls, with or without loopholes.

On the western defensive wall M36 (Fig. 4), the loopholes were arranged both in vertical and oblique position even if these were not angular27. The technique used for building the defensive wall (in rectangular and trapezoidal blocks of pakhsa of various size 60x50m‒75x30m; 90x50x30m, and big adobe bricks) was a system of double–combined walls, while in the southern side, the two corner bastions were defined by scarp walls with loopholes (Fig. 5). It was built on a podium so that, at least in the earlier phases (4th‒5th cent.), an elevated position was provided by comparison with the living quarters. As early as the 5th century, in the southern part of the tepe, but outside the castle and closely related to it, a fortified complex was built: the presence of several

Fig. 5: South–western bastions (F. Filipponi).

The Western Area28 This section located outside the walls of the castle, was continuously occupied and is characterized by the presence of rooms of a roughly square shape, which are superimposed to one another. It tends to a vertical development showing a topographic organization of the area which characterizes the seven phases of the site. The rooms were connected to each other by means of entrances and corridors.

21

Silvi Antonini 2009, 137–164, 171–180. Lo Muzio, C. 2009a, 165–169. 23 The archaeological finds are preserved in the Archaeological Museum of Bukhāra. The wall paintings are in situ. Lo Muzio 2010, 88–94; Silvi Antonini, 2010, 157–168. 24 Filipponi 2009, 109–136. 25 Silvi Antonini Filipponi & Pozzi 2009, 47–74; Silvi Antonini & alii 2010, 88–94. 26 Some peculiarities of Uch Kulakh deserve further clarification. For instance, in a previous article C. Lo Muzio, when describing the site, reports the loopholes of the defensive system of Uch Kulakh just as “arrow–slits”, instead of loopholes arranged in vertical position in the front and oblique at the corners. He also describes the castle as flanked by an unprotected village, but this is inaccurate, because the complex of Uch Kulakh is characterized by a castle and two residential areas outside of it, one located on the western side and the other on the eastern side. Furthermore, on the western area a protection wall has been identified, and for the eastern area, the prosecution of the work on the field will clarify if there is the presence of a protection wall. Lo Muzio 2009b, 49. 27 This arrangement had a long tradition in the military structures of Chorasmia until the 4th century. See: Tolstov 1948,118; Francfort 1979, 23; Khozhaniyazov 2006, 73–74. 22

During the 4th and 5th cent., the western area is bordered to the west, on the edge of the tepe itself, with a scarp wall with a walkway (A87) along which some open spaces used as lookout points, similar to those found in some of the Chorasmian fortifications, were arranged. The whole settlement was surrounded by protection walls. The arrangement of the rooms of this area, located between the podium of the castle (west side) and the wall of the settlement, their simple decoration and function ‒some kitchens and other transition rooms between the walkway and the castle‒ characterize the area as a defensive support site housing a small community. A change in the layout of the rooms (and therefore in the 28

70

Filipponi 2009, 109–136; Silvi Antonini & alii 2010, 88–94.

The Italian–Uzbek archaeological mission at Uch Kulakh tepe In period V (6th‒7th cent.) a series of remakes shows the presence of new rooms built on the previous structures, with a different intended use and a reduction in the domestic area (Fig. 7).

plan of the area), as well as in their function and appearance (walls decorated with mouldings and niches) would already occur in the next building phase VI (5 th cent.) (Fig. 6).

Fig. 8: Uch Kulach, Phase IV (7th–8th): Silvi Antonini & Mirzaachmedov 2009, 249.

Fig. 6: Uch Kulakh Phase VI (5th–6th): Silvi Antonini & Mirzaachmedov 2009, 245.

Fig. 10: Uch Kulakh Phase III (8th): Silvi Antonini & Mirzaachmedov 2009, 251.

Fig. 7: Uch Kulakh Phase V (6th–7th): Silvi Antonini & Mirzaachmedov 2009, 247.

71

ILARIA VINCENZI

Fig. 11: Uch Kulakh Phase II (8th–9th): Silvi Antonini & Mirzaachmedov 2009, 251.

In the phase IV (7th‒8th cent.) a decline in the layout of the rooms is detected (Fig. 8), as it can be seen in room A14: this room presented a central hearth in the first phase, while in the second phase it had three sufa leaning on the wall, one of them with a mural painting representing a leopard (Fig. 9)29.

A breakdown marks a caesura between the early medieval period and subsequent periods until the 13th cent.; during this interval of time, however, the site was not completely abandoned (Figs. 11–12). The Eastern Area30 This sector was less investigated than others; until 2007 levels belonging to the last phases of the site had only been observed. A series of pavement layers cut by holes, covered with baked bricks, and khum (big jars) were identified, which could suggest that this was a working area; there, the presence of an Islamic drainage system was also identified, consisting of a series of non– bottomed khum inserted inside one another and connected

Phase III (8th cent.) shows the presence of a single room, A9, with four sufa against the walls and the entrance located on the north–western part (Fig. 10). The presence of two square holes in the floor led the researches to assume that these were the remains of two wooden columns supporting the roof of the rooms.

Fig. 12: Uch Kulakh Phase I (9th–13th): Silvi Antonini & Mirzaachmedov 2009, 251.

30 29

Silvi Antonini & Filipponi 2009, 109–136; Silvi Antonini et al. 2010, 88–94.

Silvi Antonini 2007, 157–168.

72

The Italian–Uzbek archaeological mission at Uch Kulakh tepe by cylindrical pottery tubes (Figs. 13–14).

Fig. 16: Uch Kulakh, A84, Western wall, lacerto of the harnessed horse (Silvi Antonini).

Figs. 13–14: Uch Kulakh, Eastern area, Islamic drainage system (left: F. Filipponi; right: I. Vincenzi).

Operative during phases IV and V, and in connection with the rampart of the castle, a room (A13), located in front of the south–eastern bastion was identified. This space has four sufa against the walls and a rectangular structure in its centre. In a hole nearby, but outside it, a Kushano– Sasanian coin was found (Fig. 15).

Fig. 17: Uch Kulakh, A83 mural Western wall, “animal fight” (detail) (C. Silvi Antonini). Fig 15: Uch Kulakh, Kushano–Sassanian coin (F. Filipponi).

At this time, however, the residential area was located in the east, where two rooms (A83 and A84) with wall paintings were found. On the western wall of A84, harnessed horse can be recognised (Fig. 16), while in room A83, whose wall paintings are better preserved, an animal fight seems to be represented (Fig. 17). This room has four sufa against the walls and a central hearth. A large fire broke out in the central and northern parts of the castle marking the end of the castle and other defensive structures that, by that time, were adapted for residential use only. The decline of the settlement which had, already started in the late 8th cent., would be accelerated by this event. The subsequent activities mainly consisted of the construction of structures designed for agricultural and manufacturing purposes such as silos, wells and facilities for drainage of water; these are concentrated in the east and south–east sectors of the site, similar to those also found in Varakhša, Paykend and Vardanzeh, and dated to the 10th or 11th century. After this phase the site, was abandoned and the area is characterized by numerous drain holes (Fig. 18).

Fig. 18: Uch Kulakh, Eastern Area, Phase I (9th–13th) pavement layer cut by holes (I. Vincenzi).

To conclude, I would like to stress the work done in the last season of excavation that brought to light structures dated to the middle of the 4th and 5th cent., in the western sector and, in part, in the northern and south–central portion of the castle. The date of this phase is deduced primarily from the study of materials, combined with the results of radiocarbon analysis. In this period, the settlement would consist of two residential areas, one arranged at the east and the other at the west of the castle,

73

ILARIA VINCENZI As a result of the topographical exploration and survey 33 in the territory of Varakhša, 240 sites have been identified and classified as castles, fortified towns, unfortified towns with village and open sites (settlements with no structural remains often denoting a short–term occupation). As for C. Silvi Antonini, the presence of these unfortified settlements and the appearance of the fortified castle of Uch Kulakh, compared to all the castles in Sogdiana, where a castle is found as the residence of the local lord and a small village next to it, shows two important anomalies: one is that outside the walls of the castle two residential areas were found, one eastwards and one westwards, with wall paintings; and the second is the scarcity of domestic pottery. This seems to point that people lived their everyday life and their representation life outside the castle, and there they would find a refuge only when enemy attacks took place.

and a defensive complex connected to the castle. The western sector is characterized by the presence of a series of rooms, each of them with its own peculiarities and functions, while in the southern inner castle two rooms (A91, A92) were brought to light showing traces of burning with a large quantity of pottery (the C14 date 120–345 AD). In the south–eastern and eastern sectors the occupation relative to the 4th cent. is only documented by two walls (M219, M220) made of right angles in paksha blocks, which, for the position and size cannot be considered walls defining rooms at the present stage of research. In the north–eastern sector, two rooms (A83, A84) were found; their walls have wall paintings31 (preserved only in part), with sufa against the walls and, in A83 a hearth in the centre of the room. On the fragment of wall painting in A83, for example, a fighting scene between two animals, presumably camels, can be seen, which is very different from the representations of fighting animals of the Early Medieval Period. The source of inspiration could found in Sasanian art as shown by the fragments of the mural found in A84 (harnessed horses), without neglecting that the iconography of fighting between camels (A83) is typical of the Eurasian steppes. In my opinion, the latter hypothesis would be the likeliest option, as the topographic investigations of the area seem to show. The relationship between the two ethnic entities, already detected in the coinage, was very strong and the creations of the iconographic patterns characterising the Uch Kulakh paintings could be a fusion of these two different cultural realities. Therefore, the Uch Kulakh paintings would represent a formative stage of what will be known later as the Sogdian school of painting of the Early Middle Ages (whose activity is dated from the 6 th to the 8th cent.).

This factor, along with the individuation of the open site, helps us to understand the relationship between nomads and sedentary populations and clearly indicates that, although there was some kind of conflict, the two cultures certainly entertained a special relationship based on economic and cultural exchanges, as archaeological excavation shows. The site of Uch Kulakh is still under excavation, so only the prosecution of scientific work in the field will give us more information not only on the general plan of the site which must be still completed, but also on the socio– political structure and the economic history of the settlement, as well as on its relations with other settlements in the area. The particular and unique urban pattern of Uch Kulakh allow us to put forward that the site had a very important role within the territory, thanks also to its proximity to the city–capital of Varakhša, and it could be considered a sort of “regional chief place” perhaps, the residence of a very wealthy dehqān of the area.

It can be, then, said that Uch Kulakh paintings are presently the oldest paintings of the Oasis of Bukhāra and should be dated to the 4th or 5th cent32. The work carried out by the Italian–Uzbek archaeological mission, apart from the excavation program, also considered the achievement of a new archaeological map.

31

These fragments of wall paintings were studied by Lo Muzio 2010, 88–94. 32 Lo Muzio 2010, 88–94.

33

74

Cerasuolo 2009, 193–214.

The Italian–Uzbek archaeological mission at Uch Kulakh tepe Bibliography ASKAROV, A. A. 1965, “Novye naxodki Andronovskoj kul'tury v nizov’jach Zaravšana”, IMKU 6, 53–60.

ANTONINI, Dž.K. MIRZAACHMEDOV (eds.), Gli scavi di Uch Kulakh (Oasi di Bukhāra), Rapporto Preliminare 1997–2007, RSO LXXX Suppl. 1, 165–169.

– 1981, Peredatirovke kul’tury Zamanbaba. Kul'tura iskusstvo drevnego Xorezma, Moskova.

– 2009b, “An Archaeological Outline of the Bukhāra Oasi”, Journal of Inner Asian Art and Archaeology 4/2009, 43-68.

BARTOL'D ,V. V. 1968, Turkestan down to the Mongol Invasion, London.

– 2010, “Le pitture”, in C. SILVI ANTONINI, F. FILIPPONI, G. IVANOV et al. (eds.), Gli scavi di Uch–Kulakh. Campagne 2008–2009, RSO LXXXII, 88–94

BRIANT, P. 1997, “De l'Indus aux Balkans: les Rois et leur empire”, Dossiers d'Archeologie 227, 27–41. CERASUOLO, O. 2009, “Indagine Topografica nel territorio di Varakhsha”, in C. SILVI ANTONINI, Dž.K. MIRZAACHMEDOV (ed.), Gli scavi di Uch Kulakh (Oasi di Bukhāra), Rapporto Preliminare 1997–2007, RSO LXXX Suppl. 1, 193–214.

LUKONIN, V. G. 1969, Kul’tura sasanidskogo Irana, Moskva.

CANNATA, P. 1981, Profilo storico del Primo Impero Turco (metà VI–metà VII secolo), Roma,

RUZANOV, V. A. 1998, “K voprosu o xronologii metalličeskogo kompleksa kultury Zamanbaba”, IMKU 20, 41–42.

NIL’SEN, V. A 1966, Stanovlie feodal’noj architectury Srednej Azii (V–VIII vv.), Tashkent.

FRANCFORT, H–P. 1979, Les fortifications en Asie Centrale de l’àge de Bronze a l’époque kouchane, Paris.

REICHELT, H. 1931, Die soghdischen Handschriftenreste des Britischen Museums, II, Heidelberg.

FILIPPONI, F. 2006, “Poselenie Uch Kulakh”, in C. SILVI ANTONINI, Dž.K. MIRZAACHMEDOV (eds.), Ancient and Medieval Culture of Bukhāra. Materials of the conference Based on the Joint Uzbek–Italian Researches in Archaeology and Islamic Studies, Bukhāra 26–27 September 2003, Samarkand – Rome, 47–49.

SEMENOV, G. L. 1989, “Sogdijskij gorod v rannem srednevekov’e: formirovanie plana”, in G. SMIRNOVA (ed.), Itogi rabot archeologičeskix ekspedicij Gosudarstvennogo Ermitaža, Leningrad, 128–140.

SEMENOV, G. L. 1996, Sogdiskaja Fortifikacija V–VIII – 2009, “Materiali e Tecniche Costruttive”, in C. SILVI vekov, Sankt Peterburg. ANTONINI, Dž.K. Mirzaachmedov (eds.), Gli scavi di Uch Kulakh (Oasi di Bukhāra), Rapporto Preliminare 1997– SILVI ANTONINI, C. 2001, “Gli scavi della Missione 2007, RSO LXXX Suppl. 1, 109–136. Archeologica Italiana in Uzbekistan”, in C. SILVI ANTONINI, B. M. ALFIERI, & A. SANTORO (ed.), Oriente e Occidente, FRYE, R. N. 1943, “Sughd and the Sogdians: A Convegno in ricordo di M. Bussagli, Roma 31 maggio 1 Comparison of Archaeological Discoveries with Arabic giugno 1999, Roma, 248–258. Sources”, Journal of the American Oriental Society 63.1, 14–16. – 2010, “Da Uch Kulakh a Varakhsha”, in P. FEDI, C. SILVI ANTONINI, P. MORTARI VERGARA CAFFARELLI, A. A. – 1954, The History of Bukhāra, Translated from a ALABISO, D. SADUN, F. NOCI & T. AURIZI, (eds.), Alla Persian Abridgement of the Arabic Original by Narshakī, Memoria di... Convegno in ricordo di Maria Teresa Cambridge (Mass.). Lucidi, Roma 28–30 maggio 2007, Rome, 157–168. – 1996, The Heritage of Central Asia, Princeton (NJ).

– et al. 2010, “Gli scavi di Uch–Kulakh. Campagne 2008–2009”, in RSO LXXXII, 59–95.

GOEJE, J. (ed.) 1967, Bibliotheca Geographorum Arabicorum, I, Leiden. DE

SILVI ANTONINI, C.; FILIPPONI, F. 2002, “Rapporto delle campagne di scavo 2002 in Uzbekistan. Il sito di Uch Kulakh”, RSO LXXVI, 17–26.

LA VAISSIERE, E. 2004, Histoire des Marchads Sogdiens, Paris. – 2005, Sogdian Traders: a History, Leiden.

– 2003, “Rapporto delle campagne di scavo 2002 in Uzbekistan, Il sito di Uch Kulakh”, RSO LXXVI, 17–26.

LIVSIČ, V. A. 1962, Sogdijskie dokumenty s gory Mug, Moskva. LO MUZIO, C. 2009a, “Le Terracotte”, in C. SILVI

– 2009, “Il settore orientale”, in C. SILVI ANTONINI, Dž. K. MIRZAACHMEDOV (ed.), Gli scavi di Uch Kulakh (Oasi di Bukhāra), Rapporto Preliminare 1997–2007, RSO

75

ILARIA VINCENZI LXXX Suppl. 1, 109–136;

Compte–rendus de l’Académie des Inscriptions et Belles– Lettres, Paris, 1025–1033.

SILVI ANTONINI, C.; MIRZAACHMEDOV, Dž.K. (eds.) 2009, Gli scavi di Uch Kulakh (Oasi di Bukhāra). Rapporto Preliminare 1997–2007, RSO LXXX, Suppl. 1.

TOLSTOV, S. P. 1948, Po sledam drevnechorezmijskoj civilizacij. Moskva.

– 2009, “La ceramica ed altri reperti”, in C. SILVI ANTONINI, Dž.K. MIRZAACHMEDOV 2009 (eds.), Gli scavi di Uch Kulakh (Oasi di Bukhāra), Rapporto Preliminare 1997–2007, RSO LXXX Suppl. 1, 137–164, 171–180

TREMBLAY, X. 2004, “La toponymie de la Sogdiane et le traitement de *χθ et *fθ en iranien”, Studia Iranica 33/1, 113–149. VALLAT, F. 1997, “Les inscriptiones achéménides d’Iran”, Dossiers d’Archeologie 227, 48–55.

SILVI ANTONINI, C.; NOCI, F. 1999, “Rapporto preliminare della campagna di scavo 1997 a Uch Kulakh tepe (Uzbekistan)”, RSO CXXII, 221–244.

YAR–SHATER, E. (ed) 1996, The History of al–Ṭabarī, New York.

SILVI ANTONINI, C.; NOCI, F.; LO MUZIO, C. 1996, “Una ricognizione archeologica in Uzbekistan”, RSO LXIX, 3– 4, 417–428.

ZEIMAL’, E. V. 1978, “Političeskaja istorja drevnej Transoksiany po numismatičeskim dannym”, in Kul’tura Vostoka, Drevnost’ i rannee srednevekov’e, Leningrad, 192–214.

SIMS–WILLIAMS, N.; GRENET, F. 2006, The Sogdian Inscriptions of Kultobe, Shygys (Almaty).

ZEIMAL, E. V. 1996, “The Kidarite Kingdom in Central Asia”, in B. A. LITVINSKY (ed.) History of Civilizations of Central Asia III. The Crossroads of Civilizations: AD 250 to 750, Paris, 123–136.

SIMS–WILLIAMS, N.; GRENET, F.; PODUSHKIN, A. N. 2009, “Les plus anciens monuments de la langue sogdienne: les inscriptions de Kultobe au Kazakhstan”,

76

World is not enough Alexander the Great in Sogdiana: A study in historiography* BORJA ANTELA BERNÁRDEZ Universitat Autònoma de Barcelona ABSTRACT The instauration of a continuous peaceful control over the territories of Bactria and Sogdiana was one of the main problems Alexander the Great must face during his campaign conquering Asia. The harsh reaction of the Macedonian conquerors against the Sogdian and Bactrian rebels has attracted a unequal attention by scholarship. The different answers to explain these facts revealed a kind of topical explanatory tendency frequently based in modern questions more that in a careful analysis of the evidences. KEYWORDS Bactrian & Sogdian Rebellion – Alexander the Great –guerrilla – Mountains – Historiography The following pages are the early part of a wide research about the study of the politics applied by Alexander the Great at the frontiers of his empire: Trace–Ilyria, Sogdiana and India.In this sense, we cannot consider properly frontiers the territories and boundaries of the deserts of Libya and Arabia, as far as these regions received lesser attention by Alexander during the campaign against Persia, in order to defeat the main enemy, but although the unexpected death of Alexander at Babylon blocked Alexander’s plans, we know that both areas, Libya and Arabia, were in the agenda of Alexander’s future projects, so probably if Alexander had not died so soon, we can hardly considered these two areas as authentic frontiers. Also, the geographical position and their desert nature surely simplified in both cases, the Libyan and the Arabian, the decision for Alexander to create a limit, as far as the limit was a natural one. If we review the three cases of conflictive frontiers at Alexander’s Empire, as they were Thrace–Ilyria, Sogdiana and India (at the Indus), we can notice at the light of our ancient sources that the accounts stress several military activities in order to suppress any capacity of revolt or inappropriate disputation against Macedonian rule. Beyond the potential finality as a peacemaking initiative, that can be considered as an unavoidable part, even required, in a process of conquest, it seems also worthy that the sources show the absence of distinction by the Macedonians between combatant and 

Research Group HICOAN – 2014 SGR1111 (Generalitat de Catalunya).

no–combatant population among the peoples of these areas of frontier. This indistinct treatment seems to be more a result of a practical will, as it is the effective and durable elimination of any kind of possible further resistance, even menace, to Macedonian domination in the frontiers of the Alexandrian Empire, that caused by an authentic need or an answer to provocations or hostilities by the local populations. On the other hand, at the light of the information contained in the extant ancient sources it is interesting to review the interpretations and understandings that the scholars on the Ancient World, usually in connection with the contemporary circumstances, have proposed in order to explain the measures and political and military reactions of Alexander in order to control these border lands. In fact, these interpretations reveal, maybe with excessively frequency, the use of specific justifications linked not so much to the data of the sources but to traditional or topic historiographical answers. As far as this is the first step in a wide research in relation with the Alexander’s policies of frontier, during the following pages we centre our attention to the historical situation resulted by the Macedonian conquest on the territories of Bactria–Sogdiana, especially the interpretations given by the scholars to the causes and forms of the Sogdian rebellion to the Macedonians during the period 329–327 BC, which has been usually considered Alexander’s hardest test. In this sense, we can summarize a good part of what seems to be the modern historiography’s common opinion regarding the facts of the Sogdian rebellion and its

BORJA ANTELA–BERNÁRDEZ scholarship2, who initially describes the Sogdian landscape as severely mountainous and mentions its character as frontier between the fertile and rich Bactriana and the wild Scythian country beyond de Jaxartes, although he also points out that the land was rich for agriculture3 at the time that considers their inhabitants as peaceful and industrious4.

suffocation by Alexander in a brief text by Ian Worthington, one of the main modern scholars about Alexander, that can provide us as a piece of evidence for our test. The text, titled ‘Alexander the Great, Nation Building, and the Creation and Maintenance of Empire’, was published as part of a collective book about Military History of the Ancient World, concretely to the creation of strategies in Greece and Rome. In this way, about the Sogdian rebellion and Macedonian reaction, Worthington writes:

Fully documented, especially for his own time, Droysen’s work does not explain the Sogdian resistance as a kind of guerrilla, maybe because this word was, at Droysen’s own times, of very recent creation, as far as guerrilla appeared within the warfare activities of the Spanish unorganized insurgency against the Napoleonic invasion and rule over the country, at the beginning of the XIX century.

“Bessus was quickly joined by Satibarzanes, satrap of Areia, and Bacrtian chieftains such as Oxyartes (the father of Roxane) and Spitamenes, who commanded substantial numbers of men, and especially first class cavalry. To counter this threat, Alexander invaded Bactria and Sogdiana. The speed with which he moved caused these leaders to fall back beyond the Oxus, and not long after Alexander crossed this river, Oxyartes and Spitamenes betrayed Bessus to Alexander, who ordered his execution. Again, the removal of one leader meant nothing, for Spitamenes came to the fore, and the Macedonians were now faced with fierce guerrilla warfare in this different and hostile part of Central Asia. By 327, though, the resistance was over, Spitamenes was dead, and Alexander added cavalry contingents from the two areas to his army”1.

We should come until the work of Ulrich Wilcken, published originally in German in 1931 and quickly translated (the first English and French editions dated of 1932), to find an explanation about the conquest of Sogdiana by Alexander that results very similar to that recently exposed by Worthington’s brief commentary quoted before. First, this fact shows the deep impact that Wilcken’s view of Alexander has on Anglo–Saxon scholarship. Thus, Wilcken writes: “The fighting in Eastern Iran which now faced Alexander was the most severe that he ever had to engage in. It lasted almost three years, from the autumn of 330 till well into 327, and not until then could he regard the conquest of these countries as secured. This is astonishing, when one considers that his unparalleled career of victory from the Hellespont to Media needed only four years to win the throne of Persia. The difference is due to the different quality and mode of fighting of his antagonists. The Aryan peoples of Eastern Iran, especially the Bactrians and Sogdians, had a very highly developed, proud national feeling, and a tenacious enthusiasm for freedom and independence, such as Alexander had found nowhere in the Persian Empire, with the exception of Tyre and a few mountain tribes”5.

In this short excerpt we can find in a generic form the usual statements about the history of the Sogdian rebellion, from its initial context as a result of the pursuit of the usurper Bessus, the murderer of Darius III, until the hard task of suffocation of the Bactrio–Sogdian resistance, the causes of which seems to be, in the common scholar opinion, in the difficulties of the Macedonians to face the fight in a complex, arid landscape, and against a kind of war contrary to what Alexander and his comrades had to confront during the fight against the Achaemenids and the conquest of the Persian Empire. Then, the Sogdian insurgency, under the command of the very skilful Spitamenes, could have actually put Alexander in check thanks to the use of guerilla tactics. This generic mark to understanding the Sogdian rebellion and its consequences that Worthington’s text shows seems to have its support in the explicative couple territory (Worthington writes “different and hostile”, but we do not know to what this territory means to be different) and guerrilla. These two concepts, as we will see, come up as a kind of topic that deserves to be remarked, and as a consequence, to be questioned.

We cannot forget that Wilcken was not, in no way, what nowadays we can consider a kind of Alexander’s specialist6 (as for example was Droysen) as far as Wilcken’s training and research were usually focused more about papirology or Hellenistic and Roman Egypt. Nevertheless, the construction of the topic explanation mentioned in the text of Worthington seems to be explicitly present at Wilcken’s statements on the matter:

To review the basis and causes of this explanatory construction, so generic, the truth is that it seems difficult to find its origin. First, no traces of it can be found in the work of Droysen, the pioneer of the Alexander’s

1

2

About Droysen’s work, see Antela–Bernárdez 2000. At least, in the middle valley of the Jaxartes: Droysen 2012, 262. 4 Droysen 2012, 268. 5 Wilcken 1967, 154 (the first edition in German was published in 1931). Bold and italics in the text are mine. 6 See, for example, Radet’s opinions on Wilcken’s work in Radet 1934. 3

Worthington 2010, 121. Bold and italics in the text are mine.

78

A l e x a n d e r t h e G r e a t i n S o gd i a n a . A s t ud y i n h i s t o r i o gr a p h y “The enemy was difficult to get hold of; they never opposed him in large numbers, but as soon as he advanced, they appeared in his rear, if possible in several places simultaneously; they made a fresh rising; they entrenched themselves in inaccessible rocky recesses, or, if in danger, vanished into the Turkoman steppes, only to break out again suddenly and make a surprise attack upon him. How dangerous and exhausting such a guerrilla war may be, is shown by the fighting of Napoleon in Spain, in which he was unable to get the upper hand”7.

Alexander more continuous trouble than all the embattled hosts of Darius”9. Again, as if Green’s text was a kind of renewed version of Wilcken’s interpretative assessments, Green shows in the same tone as Worthington, the causes, motives and format of the Sogdian insurrection using the concept of nationalism and the binomial territory & guerrilla (“mountain guerrilla fighting”). This repetitive explanation and the unexpected absence of an authentic whole description of the facts and terms that resulted in this interpretative points at a reiterative tendency that must not have had the critical perspective that deserves at the light of the complexity of the historical events described in our sources in relation with the conflict between Macedonian conquerors and local population in Bactria and Sogdiana during the pass of Alexander’s army.

By the way of these two quotations of Wilcken’s work we can see the similarities between them and that of Worthington, although there is a distance of almost eighty years between both authors. Thanks to these two works, Worthinton’s and Wilcken’s, we can show also that the binomial territory (mountainous/dry) & guerrilla has a third element to add in this equation of topical explanations, as it is the national feeling. In fact, the use of the answer of the existence of a national feeling in Bacrtia–Sogdiana to explain the resistance to Alexander seems actually surprising, as far as there is no trace of any other national feeling similar to this one in the other parts of the conquered Achaemenid Empire during Alexander’s crossing of Asia. Likewise, we cannot leave without evaluation the use by Wilcken of the concept of nation or nationalism in his own historical context, especially considering that the idea of nation at Alexander’s times was probably very far from that of the first decades of the XX century in Europe in a time previously to the II World War.

No doubt, the most extreme account regarding our question seems to be that of Milns’s Alexande. Scholars usually focused in military questions relating to Alexander’s army and with chapters and papers about this maters, Milns wrote in 1968, in a date very near to the book of Green, his own Alexander the Great, a some short of generic book if we look at the content which in order to talk about the Sogdian rebellion shows the traditional tendency we have been putting forward. In fact, Milns describes the question in the following terms: “The peoples of both satrapies [Bactria & Sogdiana] were Iranians, organized in a feudal system. The great nobles or barons prized highly their independence and only grudgingly had acknowledged the overlordship of the Persian King. This latter had been an Iranian; but Alexander was an absolute foreigner and as such he was met with an opposition similar to that which opposed the French in Spain under Napoleon.

That Wilcken and Worthington understandings in this matter were the consequences of something more that a coincidence, and in fact there were a clue of the existence of a topical kind of explanation about this matter can be clearly shown if we compare the quoted texts with the statements of the work of Peter Green, whose Alexander the Great, published by first time in 1970, is considered a referential work in Anglo–Saxon scholarship about Alexander, although it is not much more that a review of Alexander’s main episodes in a very old–fashion manner with a very good writing8. Effectively, in his chapter related to Sogdiana, Green writes:

As in the Spanish peninsula, the fragmented state of Bactrian and Sogdiana society determined the character of the resistance to Alexander; in Bactria it took the form of mountain guerrilla warfare; in Sogdiana, the “hit and run” warfare of the desert and steppes. In the words of Major–General Fuller: “No great battles awaited Alexander; he was to be faced by a people’s war, a war of mounted guerrillas who, when he advanced, would suddenly appear in his rear, who entrenched themselves on inaccessible crags, and when pursued vanished into the Turkoman steppes” 10.

“The Macedonians were now advancing into regions where towns, as a Western would conceive them, were largely non–existent, and of which their geographical knowledge was hazy in the extreme. Here they endured three years of mountain guerrilla fighting (330–27) from Afghanistan to Bukhara,from Lake Seistan to the Hindu Kush, against the fierce, most indomitable opposition Alexander had yet been called upon to face. Bessus, and his successor Spitamenes, were fighting a nationalist war, with strong religious overtones; between them they gave

Once more, the binomial territory (“mountain”) & guerrilla appears fully linked, and again nationalism is used as the third element of the interpretation, although this time it seems to be implicitly referred by mentioning the Spanish resistance to Napoleon. It is actually very surprising to find this reference to the Spanish guerrilla during the French invasion of the early XIX century in a

7

Wilcken 1967, 154–155. Bold and italics in the text are mine, once again. 8 Bloedow 1991.

9

Green 1970, 187. Bold and italics in the text are mine. Milns 1968, 167. Bold and italics in the text are mine.

10

79

BORJA ANTELA–BERNÁRDEZ description of Alexander’s fight against the Bactrio– Sogdian insurrection, and in my opinion I consider difficult to see any kind of similarity between the Napoleonic Spain and the landscapes of Bactria and Sogdiana at Alexander’s times. Nevertheless, the intentional reference must be surely linked with the term guerrilla, which was originally related with the Spanish resistance against the French troops of Napoleon. However, Milns assures the existence of a kind of resemblance between the fragmented Iberian Peninsula [sic] and the distant satrapies of Bactria–Sogdiana, and in this similitude I guess he is again redundant in the usefulness of the concept of guerrilla to explain both historical resistance, and then, the equally between the insurgency faced by Alexander and that faced by Napoleon, and maybe even, by any other modern imperial power trying to conquest a hard, mountainous landscape and fighting against a non–organized army with a very intense nationalistic feeling. After all, Spain was lived his own historical process of nationalistic resistance during the decade of the 60/70, in this case against the Francoist dictatorship, by the terrorist organization ETA. No doubt, this explanation is plenty of over–interpretation and contemporarily, simplifying the entire complexity of the historical situations, both that of Spain, in Napoleonic or Francoist times, and that of Bactria–Sogdiana facing Alexander’s conquest. Topics are, then, the base of this explanatory solution.

be more a Greek feeling that a Sogdian. However, despite the parallels, the truth is that the explanation appears dissociated that the tendencies we are showing during the last pages, remarking in fact that, against the topical description of the events usually exposed since Wilcken based in the binomial territory & guerrilla, and sometimes linked with nationalism, we can also describe the existence of a kind of an alternative current of interpretation, which although share some points in common with that of Wilcken and those who seems to follow him, also separate from the simplicity of the interpretative binomial territory & guerrilla and the prominence of the argued nationalism supposed in the Bactrio–Sogdian insurgents. Turning back now to the main question, one of the more recent versions of this topical interpretative tendency appears recorded in the welcomed Alexander the Great by Paul Cartledge, a book that, by the other hand, seems to intend to suggest just even from its subtitle (The hunt for a new past) a review of the traditional perspectives, although finally, at least in the case of the Bactrio– Sogdian insurgency, seems to get stuck in the hunt for this new perspective, and remains in the same scope we have seen until now. Thus, Cartledge writes: “Perhaps early in October 330 Alexander proceeded south from Artacoana. Over the next three years he passed through Drangiana, Arachosia and Paropamisadae, then over the Hindu Kush Mountains to Bactria and Sogdiana and back again to his new foundations of the Alexandria–by–the Caucasus (near modern Begram and Charikar). These vital years saw him adapting brilliantly to meet and overcome unprecedented political and military conditions: vigorous local nationalism, rugged and often appallingly difficult mountainous terrain, plus all the uncertainties of guerrilla warfare”13.

Also during the 70, although later that Green’s, it was published a book that becomes a referential work in the English–speaking academic world: Robin Lane Fox’s Alexander the Great had received both praises and bad reviews as well11. Lane Fox’s view of Alexander, nevertheless, must also be considered as meritorious by its interesting proposal of understanding Alexander with some new scopes and, also, by its well–documented and detailed analysis of the factors implied in the events related with the history of Alexander’s conquest. Surprisingly, in this sense, that Lane Fox sounds more near to Droysen that to Wilcken, when about the Sogdian question he writes:

The similarities with the model we have exposed at the light of the examples quoted from Wilcken, Milns, Green and Worthington is surprising. The binomial territory (“mountainous terrain” & guerrilla appears again linked to nationalism. Although it seems clear that we are in front of an evident case of topical interpretation that resulted from questions more related to contemporary problems and situations that to ancient events, the truth is that it is very hard to define actually the origin and source of this tendency. As we have seen, Droysen cannot be refereed as the father of this kind of interpretation. Also, I personally think that Wilcken cannot be considered the first to use this explanation, as far as he was not a real specialist on Alexander. So, considering also the enormous impact this tendency had in the English– speaking scholarship, and its absence in the German academia, there must exist an author between Droysen and Wilcken who can be regard as the pioneer in the use

“Sogdia [was] no more than a sandy wasteland of stone and scrubby tamarisk, a site which only promised sickness, skirmishes with tribesmen, or more Alexandrias miles from the olive trees they knew at home. Only Bessus had ever lured them into it”12. Surely, the skirmishes can be easily considered a kind of reference to the guerrilla warfare, but at least in this occasion the expression seems does not appear in an explicit way. Likewise, the territory acquires a great prominence in Lane Fox’s view, but nationalism seems to 11

Badian 1976. Lane Fox 2004, 300 (the first edition was published in 1973). Bold and italics in the text are mine. It seems like Lane Fox was following closely Droysen’s perspective, as far as Droysen 2012, 264 also mentions bellicose tribesmen, in reference to the Scythians. 12

13

80

Cartledge 2004, 124. Bold and italics in the text are mine.

A l e x a n d e r t h e G r e a t i n S o gd i a n a . A s t ud y i n h i s t o r i o gr a p h y of the explicative framework binomial territory & guerrilla (+ nationalism).

precipitation. (…) On the basis of this modern sketch, it is quite easy to appreciate an ancient assessment of this rather unusual area: Bactrianae terrae multiplex et varia natura est [Curt. 7, 4, 26]. Although scant and widely scattered, the literary evidence for the geography of Bactria makes it possible to compare present and past conditions. For the most part this will demonstrate that the land has changed very little in the recent millennia” 15.

In this sense, our research had noticed the presence of these components in the work of David G. Hogarth, who maybe was the pioneer we have been seeking, although I am not actually sure at all of this. In effect, Hogart writes: “Almost a whole year, however, had to be spent in desultory campaigns before the back of the revolt was broken finally. With vast deserts to west, and the huge gorges leading eastward to the Roof of the World, Sogdiana and Bactria offered too many refuges to rebel chieftains to be brought under at the first attempt. (…) It was not till the summer of 327 that the Emperor was ready to repass Hindu Kush ; and so obstinate had been the resistance, and so apt did the hill country appear for fresh revolt…”14.

Thus, the existence of a mountainous landscape, with certain areas of desert, means unavoidably geographical and environmental elements that conditioned the adaptation of the local population to the country. This adaptation of the local people resulted in turn into a topical explanation, which stress the fact that the kind of habitat of the mountainous population uses to be of a tribal kind (U. Wilcken, R. Lane Fox), organized in chiefdoms. This kind of habitat occupied by tribes ruled by chieftains seems to validate a more topical parameter, as it is that which assures that the kind of warfare, brilliantly explained by Wilcken in the text of his work quoted before:

Here there is not a direct reference to guerrilla, but the territory appears as a specifically condition that resulted in resistance or insurgency of a kind of guerrilla, if we carefully consider the bold letters. Indeed, Hogarth’s text helps to understand the relationship between guerrilla and territory, at the light of the fact that a mountainous and dry, desert country aims positively the inhabitants to a kind of warfare, lets say dirty, unorganized and, after all, of insurgency based in ‘hit & run’ attacks, taking profit the rebels of the shelters of the favourable entourage of the country, rich in refuges and landforms that an invading army hardly could control, at least in the opinion of the scholars reviewed here.

“They appeared in his rear, if possible in several places simultaneously; they made a fresh rising; they entrenched themselves in inaccessible rocky recesses, or, if in danger, vanished into the Turkoman steppes, only to break out again suddenly and make a surprise attack upon him”16. But let’s turn first to the terrain, which seems to be the first step of foundation of the topical perspective we are analyzing. Although Holt bases his interpretations in the reference of Curtius, who considers Bactria terrae multiplex et varia natura est, and assimilates this kind of geographical conditions also to Sogdiana 17, it is surprising to see how a source like Curtius can be interpreted in so many and different ways. So, for example, did it H. G. Rawlison, the father of modern Assyriology, in his pioneering work about Bactria:

As far as we have highlighted the topical interpretative tendency detected along our research on the conquest of Bacrtia–Sogdiana by Alexander, our purpose now is to review its components. First, we analyze the question of the terrain. This is not, however, a easy issue, because so much have been written about the nature of the terrain of Bactria and Sogdiana during Antiquity. As we have seen, to begin the terms employed to describe the Sogdian landscapes by the scholars we are reviewing in this pages talk about different (“and hostile part of Asia”: I. Worthington), mountain & inaccessible rocky recesses (U. Wilcken), mountain (P. Green), sandy wasteland (R. Lane Fox), mountainous (P. Cartledge), mountain (R. D. Milns), o hill country (D. G. Hogarth). In a book that has become a main reference for the studies of Bactria and Sogdiana, F. Holtz has described the geography of Bactria in a way that seems to confirm effectively the assessments of the historians and scholars we have just been quoting before:

“Bactria was noted for its fertility. It is called by Strabo ‘the pride of Ariana’, and in later days it paid the large sum of 360 talents tribute to the Persian revenues. It was well watered. Besides the mighty Oxus, the Arius (the modern Hari–rud), and several less important streams, irrigate the country. It produced all the Greek products except the olive; and silphium, which was useful as an article of commerce, as well as for fattening an excellent breed of sheep, grew in great quantities on the slopes of the

“These major mountains, rivers, and deserts of Central Asia combine to create conditions ranging from stifling desert wastes to lush semitropical valleys, from mountain–bound glaciers to grazing lands of open steppes. For the most part, the region is arid, with twenty–one centimetres or less of annual 14

15

Holt 1995, 13. Wilcken 1967, 155. As usual, bold and italics in the text are mine. 17 Holt 1995, 21: “Geographically, as one might expect, the area of Sogdiana was not too much different from that of Bactria as a whole”. Nevertheless, it must be noted that Holt’s work shows a master analysis about the sources and the geographical description of Bactria and Sogdiana by the ancient authors: Holt 1995, 13–25. 16

Hogarth 1897, 230. Bold and italics in the text are mine.

81

BORJA ANTELA–BERNÁRDEZ Hindu–Kush. Lucerne, the ‘Medica herba’, as it was called from the place of its origin, grew freely in Bactria, and produced admirable fodder for the famous Bactrian horses, helping, perhaps, partially to account for the reputation which the Bactrian cavalry acquired. The well–known description of Bactrian fertility by Quintus Curtius has been praised by subsequent travellers. ‘The soil of Bactria’, he tells us, ‘varies considerably in its nature. In some spots extensive orchards and vineyards produce abundant fruit of a most delicious quality. The soil there is rich and well–watered. The warmer parts produce crops of corn; the rest is better for pastureland. The fertile portion is densely populated, and rears an incredible number of horses’ [Curt. VII, 10, 1]. It is interesting to compare what is told us by ancient writers with the remarks of a recent visitor to these regions. It will be seen that the agricultural features of the country have altered little; incidentally, the similarity between the two descriptions testifies to the accuracy of the classical geographers”18.

kind of lands allowed, although he also warns about the existence of some parts of those territories that were uninhabited deserts, but it is true that this judgement does not meant, in Curtius’ view, any kind of judgement about conditioning the local population to some kind of mountainous people or to tribal organization. Also, although Curtius mentions deserts, this does not mean that they were this were the usual landscapes of the region. We then must ask ourselves about the reason of the relationship between terrain and guerrilla, the other factor composing the binomial of the topical interpretative tendency. In this sense, it can be very useful the brilliant work of F. Cadiou about the meaning of the concept guerrilla in Antiquity. In this sense, Cadiou defends that: “La tendance de l’historiographie a faire de la guérilla lusitanienne le reflet d’un particularisme ibérique s’accompagne généralement de l’affirmation de la persistance de conceptions stratégiques et tactiques identiques entre les peuples vivants dans la péninsule Ibérique aux IIe et Ier siècles et les populations espagnoles soulevées contre les armées de Napoléon entre 1808 et 1813”21. Likewise, he also writes:

Droysen, as well, tells in his work about the existence of many different information around the richness of the Bactrian lands (and if we validate the assimilation of Holt, then of the Sogdian lands also) in his Alexanders des Grosse19. Undoubtedly, the peculiar conditions of this country have allowed a long and venturous resistance, if those who must do it could really take profit of it. The fertile and densely populated lands of the valley of Marakand, protected from the west by vast deserts and from south, east and north by mountain ranges, were not just easy to defend against any attack, but they were also favourably placed to harry from there to the occupants of the regions of Aria, Parthia and Hircania20.

“La guérilla y est toujours sommairement présentée comme la solution nécessairement privilégiée par les adversaires de Rome face à la puissance militaire écrasante des légions. La faible nombre de cas considérés comme bien attestés dans les sources suscite paradoxalement peu d’étonnement. (…) L’artifice du procède est sensible lorsque l’accent est mis sur l’origine rurale des combattants gaulois, par opposition au recrutement urbain des légionnaires romains. Outre qu’une telle opposition est largement inexacte, puisque, comme l’a bien montrés P. A. Brunt, l’armée romaine est demeurée essentiellement rural tout au long de la République, on s’aperçoit que l’an use est ici influencée par les théories de la guérilla contemporaine qui impliquent l’implantation de celle–ci à partir d’une base sociale rurale”22.

Indeed, the data seems to, at least, put in doubt the assessment of Sogdiana as a poor, mountainous or desert region, and so, it serve as a starting point to put in question the condition of the terrain that we have just mentioned as key reference in the construction of the topical interpretative tendency about the Bactrio–Sogdian uprising against Alexander. As far as it seems, during the XIX century, and at least for the first decades of the XX century, scholars had a very different view from that which became standard about Bactria and Sogdiana in the times of Alexander. Actually, this vision appears first in the texts of our ancient sources. Strabo already mentions the excellent productivity of those lands, resulting for a high fertility of the arable land (Str. 11, 11, 1), and assures the similitude of the Sogdian terrain with that of Bactria (11, 11, 3). Likewise, Curtius, in the text quoted both by Holt and Rawlison, stresses the great amount of natural resources, producing rich harvests of wheat, the optimum pasture lands for the husbandry, the intense demography that this

Although his study is related to the case of Quintus Sertorius’s use of guerrilla tactics, we can certainly consider Cadiou’s statements as very useful and applicable to our analysis. In this way, we are tempted to change here the subject of his sentences and where we say ‘Rome’ we propose to consider ‘Alexander and his Macedonians’, and the result can be also valid. Likewise, the same is acceptable for the case of the continuity of the kind of fight of guerrilla by some populations along the time until the present days23, a statement which assures the continuity of this kind of warfare tactics since Antiquity. If Cadiou explains the case of the habitants of

18

21

19

22

Rawlison 1912, 2–3. Droysen 2012, 260, 263. 20 Droysen 2012 260.

Cadiou 2004, 300. Cadiou 2004, 305. 23 Cadiou 2004, 306.

82

A l e x a n d e r t h e G r e a t i n S o gd i a n a . A s t ud y i n h i s t o r i o gr a p h y the Iberian Peninsula during the Roman Republic, who use the guerrilla tactics since then until, at least, Napoleon, and even nowadays, we then can this propose that, following the same pattern, the ancient Bactrians and Sogdians can surely be understood within the same explicative framework, as far as these populations have had fight against great armies in modern times, as the resistance against the English colonizers during the second half of the XIX century or, more recently, the resistance against the Russian and even USA’s imperialism, then some scholars have thought that their ability to resist against those great armies is a key element of these cultures (as far as it resulted from their adaptation to a specific kind of territory, as we have already seen), and for instance, it can be traced back in time, until Antiquity, that is to say, until Alexander’s times. The difficulties Alexander had suffered in the region can then be considered as parallels of those of the other western armies during their own attempt, in different times, to rule these lands and peoples.

religion seems to turn us back to our present days more that to ancient times, as far as during the XIX, XX and XXI centuries the territory of Afghanistan and the resistance of its populations first against English and later to Russian and even recently to the Western allies (named in a very generic way, certainly arbitrary and maybe deliberately confusing) as a consequence of the post–11S international consequences. If we think that when the scholars linked to the simplistic explicative tendency we have shown, and if they think in the religion of the ancient Bactrians and Sogdians they would just consider Zoroastrism, but the question is actually very controversial; but it seems very difficult to assure, at the light of our actual evidences and the current situation of the research, a statement like this, so categorical like the one which we have exposed here in order to link it with the causes of the resistance against Alexander. On the contrary, if when we talk about ancient religion we are thinking in the contemporary religion of these territories, that is to say, Islam, then the question acquire a new light and perspective, and the guerrilla warfare, as happened with ‘nationalism’, remarking what we have stressed along these papers: that the traditional explicative tendency about the Bactrio–Sogdian revolt against Alexander has its causes in constructions unconnected with the sources and the evidences, trying to develop accounts with the aim of explaining the present circumstances at the light of references in the past. This is not, of course, a reprehensible practice at all, but the historian must make himself conscious of what he is explaining and of the reasons for doing so, in order to explain without mysteries his explicative intentions to the reader.

The very recent actions of the Western coalition of allies in Afghanistan in the beginning of the XXI century would actually renewal the validate of this explicative tendency which, after all, it is caused by very simply and non– critical theoretical topics and has very little in common with a real historical perspective of the facts and the interpretation of the sources. A good clue, for example, is the reference in some authors which we have already quoted here, to the question of the religion, and in fact, to nationalism24. First, we must warn that the term nation, and its derivative, nationalism, have had not the same meaning in the European cultural context of the XIX century when they appeared and during the XX and XXI centuries. Also, the idea of a guerrilla war caused by

24

National feeling (U. Wilcken), nationalist war (P. Green), nationalism (P. Cartledge).

83

BORJA ANTELA–BERNÁRDEZ Bibliography ANTELA–BERNÁRDEZ, B. 2000, “J. G. Droysen e o primeiro Alexandre científico”, Gallaecia 19, 219–249.

HOGARTH, D. G. 1897, Philip and Alexander of Macedon: Two Essays in Biography, London.

BADIAN, E. 1976, “[Review] Alexander the Great, by Robin Lane Fox”, JHS 96, 229–230.

HOLT, F. 1995, Alexander the Great in Bactria, Leiden. LANE FOX, R. 2004, Alexander the Great, London.

BLOEDOW, F. 1991, “Alexander the Great and Bactria”, PdP 256, 44–80.

MILNS, R. D. 1968, Alexander the Great, London.

CARTLEDGE, P. 2004, Alexander the Great. The Hunt for a New Past, London. CADIOU, F. 2004, “Sertorius et la guérilla”, in C. AULIARD & L. BODIOU (eds.), Au jardín des Hespérides. Histoire, société et épigraphie des mondes anciens. Mélanges offerts à Alain Tranoy, Rennes, 297–314..

RADET, G. 1934, “Alexandre le Grand, d’après Ulrich Wilcken”, Revue Historique 173, 80–90. RAWLISON, H. G. 1912, Bactria. The history of a forgotten Empire, London. WILCKEN, U. 1967, Alexander the Great, Norton, New York.

DROYSEN, J. G. 2012, History of Alexander the Great, Philadelphia. GREEN, P. 1970, Alexander the Great, New York/Washington.

WORTHINGTON, I. 2010, “Alexander the Great, Nation– building, and the Creation and Maintenance of Empire”, in V. D. HANSON (ed.), Makers of Ancient Strategy: From the Persian wars to the fall of Rome, Princeton, 118–137.

84

Rex armeniis datus?* Nero, Parthia and the ‘Armenian Question’ JOAN OLLER GUZMÁN Universitat Autònoma de Barcelona ABSTRACT This paper addresses two different yet related aspects of the Armenian question during Nero’s rule. Firstly, it examines imperial foreign policy in the east during the Roman–Parthian War of AD 54–63. Secondly, it focuses on the figure of the emperor himself in order to determine his role in the conflict and contribute to the traditional debate concerning the last Julio–Claudian emperor. The core of the debate questions Nero’s capacity to handle the reigns of the empire. Some authors have denied him any ability whatsoever, either by placing all recognition on his councillors and generals or by directly failing to recognize the existence of any such merit. This position has been softened by other authors, who have pointed out his political and management skills. The ultimate aim is to contribute further to the debate with a detailed analysis of literary sources dealing with Nero’s Armenian campaign, which spanned practically the entirety of his reign. KEYWORDS Rome – Armenia – Nero – Corbulo – Parthians himself to Galba, or apply to the Parthians for protection…”

1. Nero and the Parthians, loyalty beyond death? 

Initial interest in the topic sparked from a set of references in the sources on the image held of Nero by the Parthians. Classical authors note that after the death of the emperor, there remained a positive, even reverential attitude towards Nero. There are three types of such references. – Suetonius (Nero 57.2) mentions how Vologases, king of the Parthians and Nero’s rival in the Armenian conflict, explicitly asked for Nero’s memory to be honoured, while attempting to resume diplomatic relations with Rome after his death.

– Finally, and most significantly, there are the ‘false Neros’1. Several classical authors mention that after the death of the emperor various figures in the east impersonated him. They claimed to have survived the generalized revolt against Nero’s government, escaping to the eastern parts of the empire. These figures were not only accepted, but even honoured. Suetonius (Nero 57.2) refers to one of them, who probably existed during the reign of Domitian2. The Parthians sheltered him and expressed apprehension when turning him over to Rome.

“Nay more, Vologaesus, king of the Parthians, when he sent envoys to the senate to renew his alliance, earnestly begged this too, that honour be paid to the memory of Nero”. – Suetonius (Nero 47.2) also points out how the emperor, seeing his end near, considered several strategies of survival, including asylum in Parthia. “… but part of them, showing no great inclination to comply, others absolutely refusing; and one of them crying aloud: Say, is it then so sad a thing to die?, he was in great perplexity whether he should submit

“In fact, twenty years later, when I was a young man, a person of obscure origin appeared, who gave out that he was Nero, and the name was still in such favour with the Parthians that they supported him vigorously and it was with much difficulty that they were prevailed upon to give him up”. An even more interesting account is found in Zonaras (11.18), concerning a certain Terentius Maximus, who shared a striking resemblance with Nero, besides being an accomplished sitarist. Taking advantage of these 1



This Research has been made within the Research Group HICOAN – 2014 SGR1111 (Generalitat de Catalunya).

About this question the reference work is still the one of A.E. Pappano 1937. 2 During AD 88 – 89; Pappano 1937, 381.

JOAN OLLER GUZMÁN circumstances, in AD 69 he pretended to be the emperor, allegedly hiding in the east until he recovered his power3. He recruited troops in the eastern provinces and crossed the Euphrates, where he was sheltered and fervently supported by a certain Artabanus, who aspired to the Parthian throne. A last instance is found in Tacitus (Hist.2.8), who mentions another ‘false Nero’, also in AD 69, although in this occasion he does not seem to have had any contact with the Parthians.

Graeco–Roman literary sources, which are scarce and mainly focused on Roman interests, therefore greatly biased. In fact, the image gauged from these is not only fragmented, but also condescending, often portraying the Parthian Empire as a buffer state in decadence, lying between the two great Iranian empires – the Achaemenid Persian Empire and the Sassanian Persian Empire 6. Nothing was further from the truth. A careful revision of classical texts, combined with the study of these scarce primary sources and the advances offered by archaeology, have allowed for a reassessment of the Parthian Empire. Recent works, like those by Ch. Lerouge or U. Ellerbrock & S. Winckelmann offer a more accurate panorama on the real significance and power of this empire (Lerouge 2007; Ellerbrock, Winckelmann 2012).

All of these examples demonstrate that Nero retained a certain degree of prestige and reputation among the inhabitants and rulers of the Parthian empire. Different elements converge to make this possible. For instance, messianic traditions in the east spoke of a ‘reborn king’, who would save the world and regain Asian hegemony over Rome4. Although these traditions are relevant for understanding the posthumous influence of Nero on the eastern part of the Roman Empire, it is not the focus of this paper, which centres on his foreign policy and the ways these political measures contributed to building the emperor’s image in the Parthian world.

The Parthian Empire rose from the defeat and expulsion of the last Seleucids in the northeastern quarter of present day Iran towards the mid 3rd C BC. It drank in equal measure from Hellenistic influences and ancient Persian and Iranian traditions7. From this point onwards, the kingdom was ruled by a hereditary and absolutist monarchy, in which monarchs decided on their successors solely based on their personal will. The influence of the Iranian nobility over the throne explains the instability that characterized periods of succession in the Parthian Empire, reflected in the literary sources and so often taken advantage of by Rome and other enemies 8. Despite this, it appears quite clear that, as of the 2nd C BC, particularly after the reign of Mithridates II (124/3 – 87 BC), Parthia became the main political entity in the Near and Middle East. It developed a significant military apparatus and expanded between the Euphrates and India 9. The status quo remained unaltered during the following three centuries, contributing in a decisive manner to the development of this Iranian area10, eventually to become Rome’s main political and military alter ego in the east. In sum, the Arsacid dynasty was neither the successor of a traditional Persian kingdom or a barbarian form of Hellenistic monarchy; it was an original and successful eastern power (Dabrowa 2011, 121).

This favourable image is a direct consequence of the political and military decisions taken by the emperor during the Roman–Parthian War over the control of Armenia, which lasted for most of Nero’s reign. Therefore this paper will focus on the literary references concerning Nero’s actions during the conflict. 2. The Parthian Empire: Rome’s alter ego in the east The Parthian or Arsacid Empire was Rome’s principal political and military rival in the eastern limes from the first Roman interventions in the area in the 1st C BC up to the 3rd C AD. The characteristics and development of the Parthian Empire are complex, so the outline of some basic elements are necessary for understanding its relationship with Rome during the Republic and the Early Empire. Firstly, it is important to bear in mind that primary sources on the Parthians are very limited: some inscriptions in Greek, Aramaic and Parthian found in Bisitun (Iran), Susa and Seleucia on the Tigris; three papyri and varied graffiti; certain cuneiform texts recovered in Babylon; and the ostraka discovered in Nisa. Coin evidence is the principal primary source that records the different monarchs ruling over the empire 5. Nevertheless most of the information has originated from

6

For instance, W. Grajetzki emphasize on how the Parthian art has been traditionally seen as a degeneration of Greek art, because of the Hellenistic influences and not considering its own rules, principles and influences: Grajetzki 2011, 1. 7 In fact, an important part of the arsacide monarchs tried to link directly with this traditional Persian past, assuming the title of "king of kings", original of the achaemenids kings. In the same way, the military expansion was associated in the arsacid period with the expansive program of the achaemenids, reinforcing the idea of continuity between both empires: Rahim Shayegan 2011, 39 – 42. 8 On the Parthian political structure, vid.: Ellerbrock & Winckelmann 2012; Dabrowa 2011; Fowler 2010; Butcher 2003, 33 – 35; Lerouge 2007, 246 et seq. 9 On the Parthian army and its characteristics, vid.: Wilcox 1986. 10 In this respect, some of the archaeological studies taken in the territory from the Parthian empire have demonstrated this decisive contribution in aspects as the reinforcement of the urban life, the growth of the rural sites or an important investment in basic infrastructures, such as routes or hydraulic channels: Wenke 1981.

3

On this case: Pappano 1937, 390 – 391. For example, we have diverse Sibylline Books of Persian origin that speak about this “reborn king”. In the same way, it is necessary to notice the great influence of the Jewish population inside the Roman Empire, a Jewish population who also was defending the arrival of the new Messiah in Judea, concretely in Jerusalem, tradition that was mixed also with the rest of Messianic oriental visions and that would associate equally to Nero's figure. For more details on these visions, its origin and influence in the mentality of the moment and the relation with Nero's figure: Matthews 1937; Charlesworth 1950. 5 On these sources and its problems: Bivar 1983, 21 – 22; Lerouge 2007, 15 – 18. 4

86

N e r o , P a r t h i a a n d t he ‘ A r m e n i a n Q u e s t i o n ’ against the action, which went unanswered. Finally, he counterattacked, sending the Roman troops once more over the Euphrates.

Once Parthia became the main political power in the east as of the 1st C BC, contact with Rome, the aspiring political power in the Near East, was inevitable. The sources cite the first direct encounter between both empires as taking place in 92 BC, when Sulla received an embassy from Mithridates II, which was treated with condescendence and arrogance (Plut.Sull.5.4–5). In fact, different encounters between Rome and Parthia throughout the 1st C BC revealed a considerable lack of tact on the side of the Roman generals and leaders, who dealt with Parthia as with any other eastern kingdom, underestimating its real political, military and demographic importance11. Nevertheless Parthia was not like other eastern kingdoms. After decades of intensified friction between the two, Rome finally grasped the extent of Parthian power. On the other hand, from a Parthian point of view, it was necessary to gain full diplomatic recognition from Rome as a world power, so as to act on an equal basis in the geopolitical international scene and reinforce its position before enemies and allies (Dabrowa 2011, 118).

Rome’s underestimation of Parthia reached a point of inflection during the campaign of Marcus Crassus of 55– 53 BC, which ended with his total defeat at Carrhae15. From this point onwards, Rome became fully aware of the Parthian danger, especially after the invasions of Syria in 52–50 BC and the 40 BC attack of Orodes I in alliance with Quintus Labienus, a former commander in the armies of Cassius and Brutus. These events formed the basis for the “Parthian question”, which aimed at avenging the insults suffered at the hands of the Parthians –especially Carrhae– and conquer the kingdom, in order to establish Rome as the only power in the region (Lerouge 2007, 90 – 100). This was the context that framed the failed expeditions of Caesar (44 BC) and Anthony (36 BC), as well as the first political steps taken by Octavian in the east after his victory in the civil war. It was precisely Augustus, who decided to give Rome’s relations with Parthia a definite turn, setting the stage for the development of the Roman–Parthian conflict of the 1st C AD.

Contact throughout the 1st C BC made Rome realize the real danger posed by Parthia for Roman interests in the east12. As we have seen the very first contact had taken place with Sulla and, in fact, the Mithridatic Wars offered the ideal context for intense diplomatic relations. Thus both Lucullus and Pompey sent embassies to the Parthian king, Phraates III, asking for an alliance –or at least neutrality– in the conflict against Mithridates of Pontus 13. Especially interesting are Pompey’s actions, because they implied the first confrontation between both empires. The stroke took place when –after signing a pact of alliance with the imperator against Pontus– Phraates advanced over Armenia. Pompey, angered, also entered Armenia and expelled the Parthians, advancing as far as the Tigris River (Lerouge 2007, 58 – 59). As the Euphrates River was the established boundary for both empires and their areas of influence, we can imagine that reaching the Tigris would have caused a great impression among the Parthians14. Phraates sent then an embassy to protest

3. Political relations between Rome and Parthia in the 1st C AD: Armenia as an eastern hotspot Augustus saw the lack of political realism enclosed in the idea of militarily subjugating Parthia, so he changed the Roman strategy in the area, focusing on two alternate measures (Grisham 1951, 227; Bivar 1983, 65; Lerouge 2007, 104; Eldwell 2008, 9): – Withdrawal of direct conflict in the area, since it had turned out to be fruitless and dangerous for Roman interests. – Creation of a group of vassal states, allied with Rome, serving as buffer states between Rome and Parthia, and ready to supply Rome with human, military, and logistical resources in case of an armed conflict in Armenia. Among the most significant of these states in the 1st C AD were Commagene, Cappadocia, Sophene, Armenia Minor or, more to the south, Palmyra 16.

11

In relation with it, Sila's attitude would be paradigmatic. Thus, Plutarch indicates how after having received Orobazes, Parthian ambassador, he humiliated him publicly putting him at the same diplomatic level as Ariobarzanes, king of Capadocia, a small allied state of Rome. An attitude that provoked critical comments for his arrogance and that cost the life of the ambassador, executed by his monarch on his return to Parthia for having allowed such humiliation. For an analysis of this episode: Lerouge 2007, 45 – 48. 12 For a description of the political evolution of the contacts between Rome and Parthia during the 1st C BC vid.: Grisham 1951, 223 – 227; Eldwell 2008, 7 – 11; Bivar 1983, 45 – 66; Lerouge 2007, 43 – 93; Sherwin–White 1984, 159 et seq. 13 Firsty during the winter of 69 – 68 BC and secondly during the 66 BC. 14 A frontier that was maintained until the beginning of the Flavian era and that was rather symbolic because until this moment it did not have regular fortifications. About the Euphrates as limit between Rome and Parthia, vid.: Eldwell 2008, 7 – 8.

The success of the change in strategy was reflected in the pact agreed by Augustus and the Parthian king Phraates IV in 20 BC. Augustus recovered the military standards lost in the campaigns of Crassus and Anthony, as well as 15

We cannot forget that, soon before, Gabinius, the Syrian governor, had already planned an attack against the Parthians on the base of a help demand of Mithridates III, pretender of the Parthian throne (D.C.39.56.1–3). 16 A few states where Rome was exercising a sovereign power to impose their monarchs, which had to answer at all time to the interests of Rome; a fact that obviously not always pleased the inhabitants of the above mentioned territories and that has its major reflection in the coinage of coins where these kings were receiving the epithet of “philoromaios”. For more details on the evolution of some of these vassal states, vid.: Hekster 2010; Facella 2010; Sommer 2010.

87

JOAN OLLER GUZMÁN those prisoners that were still alive. It was a great diplomatic success, which Augustus divulged as a victory and sign of the submission of Parthia to Roman will. The conditions imposed in the pact and the subsequent propaganda conveniently coincided with a moment of internal weakness and dynastic crisis affecting the Parthian throne (Lerouge 2007, 104 – 106)17. This situation was increased when Phraates IV sent on his own initiative his nine sons and grandsons as hostages – one was later claimed by the Parthians during another dynastic crisis between AD 6–9. At a first glance it would seem that with Augustus Rome did impose certain superiority over Parthia, at least at a symbolic level, expressed in the delivery of hostages and Rome’s participation in the designation of the Parthian monarch. In reality, the situation was not what it seemed. For example, Phraates probably sent his descendents to Rome as a tactic to eliminate competitors for his favorite candidate, Phraataces (Bivar 1983, 68); whereas the Roman candidate, Vonones I, did not last on the Parthian throne, because he was not accepted by his people, and was subsequently defeated by the anti–Roman Artabanus III of Iranian origin (Grisham 1951, 227; Bivar 1983, 68 – 69; Lerouge 2007, 113)18.

Armenia had already revealed itself as strategically important, and both Lucullus and Pompey were driven to intervene in the area against the powerful Armenian king Tigranes II. Nevertheless it was only after the 1st C AD that Armenia became the main battle field. In this regard, Rome was always fully aware of the need to hold elements of influence and power over Armenia, as a counter–weight to the traditional Arsacid influence in the territory (Cizek 1982, 323). In fact, from this point onwards, Rome directed its foreign policy in Armenia according to three possible main lines of action: – Military conquest and submission of Armenia. – Real sovereignty through the imposition of a roman candidate to the throne, that is to say a “rex datus”. – Nominal sovereignty allowing Parthia to choose its own candidate to the throne, although with Rome’s sanction. As of Augustus, and after signs of crisis in the Parthian dynasty, Rome opted for the second line of action, imposing candidates on the Armenian throne (Hammond 1934, p. 82)20. This was certainly not a situation that agreed with the Parthian monarchs, generating constant instability and conflict in the area. The first great moment of crisis took place under Tiberius and Artabanus III, more specifically in AD 35, when the latter expelled Mithridates, the Roman candidate to the Armenian throne, imposing his own son. The ensuing conflict ended in the negotiation between Vitellius and Artabanus and the restoration of peace conditions (Bivar 1983, 73 – 76; Lerouge 2007, 130). At the death of Artabanus in AD 38, Rome imposed Mithridates once more on the Armenian throne, a clear indication that the conflict was far from resolved. For Rome the Parthian question remained a priority. In spite of the fact that armed conflict was avoided, the aim was to constantly promote internal strife in the Parthian kingdom, as seen in the actions of Tiberius and Claudius, who in different occasions supported the causes of royal hostages in Rome attempting to take over the Parthian throne (Lerouge 2007, 133). This was the historical context that framed the Roman–Parthian War that would take place after Vologases I rose to the Parthian throne in AD 51.

Nevertheless there is one aspect of this entire situation that is really interesting for us and it is when Rome imposed a monarch on a foreign throne to serve its interests. This kind of indirect interference in the internal affairs of a foreign territory is known by the term ‘rex datus’. The measure was popular among Roman emperors and became one of the main instruments of political propaganda, displayed on architecture, sculpture, and especially on coins19. In our case, it is related to the foreign policy adopted by Rome in the east after Augustus. Thus, after the pacts signed by the princeps, armed conflict between Romans and Parthians was abandoned for a time. However frictions between the two powers were shifted to another territory that would concentrate most of the military activity of the following century, becoming a real hotspot in the east: Armenia. It was a strategically positioned kingdom that acted as a buffer state, separating Rome and Parthia between the areas of the Black Sea and the Caspian Sea, which controlled direct passage both to Asia Minor and Syria (Roman territories) and to Media (Parthian territory). In the Republican Period, during the Mithridatic Wars,

4. Nero vs Vologases: the Roman–Parthian War for Armenia

17

Maybe the most famous example of this propaganda would be the system of symbols associated to the August of Prima Porta statue, where appears the image of the defeated Parthian, giving the standards to the Roman emperor. On this symbols and its interpretation problems, vid.: Zanker 1990, 189; Simpson 2005. 18 In fact, contemporary authors like Strabo, Pompey Trogue or Velleius keep referring in its writings to Rome and Parthia as the biggest empires of the world, therefore showing that it did not exist an image of Roman superiority on Parthia: Lerouge 2007, 119. 19 We have, for instance, some well known cases of monarchs using the legend “rex datus” as propaganda for their foreing policy actions, such as Antoninus Pius, Traian or Marcus Aurelius and Lucius Verus: Pani 1972, 264 – 265 ; Suspène 2012.

With Vologases begins a period of consolidation and development of Parthian power after years of internal strife. This monarch attempted to recover the past glory of the Parthian kingdom, based on three objectives: – Iranization of the state, reflected, for example, in the dissemination of Zoroastrianism and the gradual 20

So the year 1 AD took place a reunion between Gaius Caesar, adopted son of August, and Phraates V, where it was agreed that Ariobarzanes, philoroman, will be the king of Armenia: Lerouge 2007, 129; Eldwell 2008, 11.

88

N e r o , P a r t h i a a n d t he ‘ A r m e n i a n Q u e s t i o n ’ abandonment of Hellenistic elements present in many cities, such as Susa or Babylon (Farrokh 2007, 146).

Before going into any detail on Nero’s decisions concerning the Armenian question, it is necessary to understand the sequence of events25. As previously mentioned, when Nero came into power, he was confronted by the forwardness of Vologases I, who had imposed his own candidate, Tiridates, on the Armenian throne, clearly breaching the agreement reached with Augustus. Nero was therefore obliged to also launch an offensive campaign in the area, which initiated the Roman–Parthian War of AD 54–6326. The emperor decided to reinforce the position of his vassal states, placing several pro–Roman monarchs on the thrones of kingdoms, such as Commagene, Armenia Minor, Sophene or Judaea. He also sent an embassy to Vologases I, demanding an end to hostilities and the delivery of hostages, while putting his armies in the hands of a reputed general, Corbulo, who had already served Claudius. The sources detail how the general trained his legions harshly, although during the first part of the war (AD 55–56) he avoided armed conflict, preferring to negotiate with Tiridates in Armenia, a tactic which would prove to be the predominant strategy throughout the war. Although the Parthians withdrew for a time, peace negotiations were not fruitful, and in AD 57 Tiridates took over the throne once more, giving way to the first violent confrontation. Corbulo shrewdly took advantage when Vologases was forced to face a rebellion in Hyrcania, in the eastern part of his empire, leaving his brother Tiridates alone in Armenia. Corbulo then led a triumphal march over Armenia between AD 58–60, taking over the two main cities, Artaxata and Tigranocerta. Afterwards Nero decided to continue with the political tradition initiated by Augustus and placed Tigranes V on the Armenian throne, a former hostage of Rome and member of the royal family of Cappadocia. So Nero opted for the second line of action: real sovereignty over Armenia. Tigranes V was supported by 1000 legionaries, 300 auxiliaries, and the forces of neighbouring vassal states27. This state of affairs did not favour the Parthians and Tiridates immediately undertook a failed attempt in AD 60 to recover the throne. However the conflict did not reach its climax until AD 61, when Vologases returned from suppressing the Hyrcanian revolt and dedicated himself to recovering the Armenian throne

– Internal stabilization to put an end to the chronic dynastic problems that had affected Parthia in the last decades. To achieve this, Vologases placed his brothers Pacorus and Tiridates on the Median and Armenian thrones respectively (Wolski 1993, 165). – Aggressive foreign policy in Armenia, where Vologases imposed his brother on the throne, breaking with the existing agreements with Rome (Wolski 1993, 164). Parthia’s aggressive foreign policy led to a confrontation with Rome. Once in power, Vologases attempted to impose his brother Tiridates on the Armenian throne and to that effect invaded the region. He expelled Radamist, who in turn had deposed Mithridates and managed to be accepted by Rome. After a series of military actions, Tiridates finally defeated Radamist and imposed himself as king, a situation which greeted Nero at the beginning of his reign in AD 54 (Bivar 1983, 79 – 81). Nero, emperor between AD 54–68, is one of the most debated Roman figures. The polemic originates in the image offered of him by the classical literary sources, framing him into a kind of ‘Julio–Claudian black legend’21. Indeed, if ever there was an emperor despised and mistreated, as much by the classical sources as by modern scholarship, it is without a doubt Nero, who has been characterized as sadistic, degenerated, matricide, etc22. Furthermore he has been described as a bad governor, who took many of his political decisions according to personal interest rather than for the good of the empire (Van Overmeire 2012, 753 – 755). The biased vision given of the emperor by the classical sources is extremely problematic. Nevertheless recent approaches to the study of his biography have reviewed this image and reclaimed his role as an emperor, attempting to understand the context of his actions and highlighting his achievements and innovations23. Nero tried to reform the empire, introducing important social and political changes in the conception of the state, which included accentuating its conversion into an autocratic monarchy of eastern characteristics (Fernández 1991, 202 – 206). Hence classical sources should be reviewed in detail, reread, and even re–interpreted24.

the rest, with Nero governing by himself, of mismanagement and political drift, is not so clear, appearing some of the most remarkable actions of the emperor (as the monetary reform or the resolution of the Armenian conflict) in this supposed bad second period of government. Vid.: Oller 2009. 25 For the description of the events we have used the already mentioned works that resume the narration of the literary sources: Hammond 1934, 92 et seq..; Grisham 1951, 228; Cizek 1982, 323 et seq.; Bivar 1983, 79 et seq.; Wolski 1993: 165 et seq.; Lerouge 2007, 130 et seq.; Eldwell 2008, 17. 26 It has been some discussion about the exact chronology of the war and its different episodes, basically because we depend exclusively on the literary sources, which not always coincide. As it is not the main aim of this paper, we will use the classical chronology. On this question, vid.: Henderson 1901; Carter 2004. 27 On the question about the support of these vassal states, vid. A.A. Barret 1979.

21

As we will see later, the main literary sources that gave us information about Nero’s government are Tacitus, Cassius Dio and Suetonius. 22 Having as a magnificent example the famous Nero played by Peter Ustinov in Quo Vadis?, paradigm of the self–worship, the vices and the madness. 23 For instance, as works that follow this vision: Griffin 1987; Elsner & Masters 1994; Champlin 2003. 24 We can use as example our previous work on the influence of the political ideology of Seneca in the first years of Nero’s government, from which we could determine that the traditional differentiation of this government, born in the literary sources, between a few first five years of good government under the influence of Burrus and Seneca (the quinquennium Neronis) and

89

JOAN OLLER GUZMÁN for his brother. He justified his attack by allegedly aiding Adiabene, a vassal state of Parthia that had just been subject to incursions by Tigranes. Vologases unleashed a great attack on two fronts: Armenia, expelling Tigranes and replacing Tiridates on the throne; and the Euphrates, against Corbulo, who was now governor in Syria. Corbulo asked Nero for the help of another commander to lead the conflict in Armenia, while he defended the limes along the Euphrates. The emperor sent Caesennius Paetus, who in AD 62 disregarded Corbulo’s previous caution and experience and openly attacked Tiridates, failing completely, and finally accepting a shameful surrender28. After their victory, the Parthians sent an embassy to Rome in AD 63, asking Nero to recognize Tiridates as king of Armenia. Nero did not only deny their petition, but awarded Corbulo absolute power in the east to solve the Armenian question once and for all. Neither Corbulo nor Vologases were interested in a violent confrontation, so after negotiating, it was decided that Tiridates would symbolically renounce his crown in front of a statue of Nero, and later travel to Rome to be crowned by the emperor himself, a ceremony that took place with great splendour in AD 66. The change from the second line of action (real sovereignty) to the third line of action (nominal sovereignty) would ensure 50 years of peace between Rome and Parthia along the frontier of the Euphrates River.

Nero’s actual role during the Roman–Parthian War in Armenia may be gauged by reviewing the events described in the literary sources that have been used to support such different interpretations. 1) Nero’s first orders In AD 54, the year the war began, Tacitus (Ann.13.6–8) mentions that Nero sent new legions east, called on the support of his vassal states, built bridges over the Euphrates, and very particularly, named Corbulo as commander in chief of the troops. In general his orders were perfectly coherent with the logical development of an overseas conflict. The efficiency of his measures were never put to test, since Vologases was forced to withdraw temporarily due to a revolt led by his son Vardanes. However Tacitus remarks that the senators were very approving of the appointment of such a prestigious general as Corbulo, which cleared any initial doubt surrounding the emperor’s capacity to manage delicate foreign affairs. “Apart from the routine of sycophancy, they felt genuine pleasure at his appointment of Domitius Corbulo to save Armenia: a mesure which seemed to have opened a career to the virtues” (Ann.13.8). Everything indicates that Nero responded swiftly and adequately to the situation in Armenia. After Vologases’ withdrawal, he ordered the troops to be divided between the command of Ummidius Quadratus, governor of Syria, and Corbulo. Both generals would later compete over leadership of the campaign. Tacitus describes the tensions between the generals when the time came for Vologases to relieve hostages, in order to know with whom of them these hostages had to stay. After some debate, it was finally decided they should stand under Corbulo’s protection, which clearly upset the provincial governor (Ann.13.8–9). This episode illustrates Corbulo’s problematic nature, which would reappear in later years with negative consequences. Nevertheless it was Nero’s second successful political intervention. He reconciled both generals by bestowing upon them identical honours, showing sagacity and political vision in times of military campaign.

5. Victory… was it Nero’s?… Corbulo’s? … or the Parthians’? It is time to address Nero’s role in the Armenian conflict. The principal sources of information are the classical texts29: Suetonius in the Life of Twelve Caesars; Cassius Dio in Book LXII of Roman History; and especially Tacitus in his Annals (Books XIII–XVI), who provides the most detailed description of events30: From these sources, historians have followed two main approaches to the Armenian conflict, which in fact coincide with the two existing visions on Nero’s figure as an emperor. For example, Henderson (1903) and Schur (1963), in their classic works on the emperor, present him as a capable and intelligent ruler. This image has been rejected by other authors, who place the success of his campaign on the decisions of imperial councillors, such as Seneca and Burrus, or on the diplomatic and military abilities of the commander in charge, Corbulo (Momigliano 1931). Alternatively, historians have also interpreted Nero’s actions as negligent and damaging for the development of the war (Hammond 1934). Some have even pointed out that the victory was actually Parthian, while Rome suffered an important setback in foreign affairs (Lerouge 2007).

2) Initial combats and the first contact with Tiridates The next remarkable event took place during the first violent clash between Corbulo and Tiridates. It was a victorious campaign for the Romans, taking the cities of Artaxata and Tigranocerta (Ann.13.36–37). At the very beginning both generals avoided direct conflict, preferring minor skirmishes and practical guerilla warfare. Tiridates did not achieve his objectives, so he sent an embassy to Corbulo, claiming that Rome had deprived him of the throne that belonged to him by right. The Roman commander’s answer is quite suggestive. He recommended that Tiridates direct his petition to the Ceasar, instead of spilling more blood (Tac.Ann.13.37).

28

The defeat was so shameful that even Tacitus compares it with the Caudine Forks (Ann.15.15). 29 Though, as we will see, also the numismatics can give as some interesting clues about this question. 30 In our approach of these authors and their texts, we have used the following editions: for Tacitus, the translation by Jackson 1969; for Suetonius, the translation by Rolfe 1979; for Cassius Dio, the translation by Cary 1968/70.

“Corbulo, who had sure information that Vologeses was detained by the revolt of Hyrcania, rejoined by

90

N e r o , P a r t h i a a n d t he ‘ A r m e n i a n Q u e s t i o n ’ advising Tiridates to approach the emperor with a petition: – “A stable throne and a bloodless reign might fall to his lot, if he would renounce a dim and distant hope in order to pursue one which was within his grasp and preferable”

decision seems intelligent and adapted to the geopolitical situation in the area, since Rome’s victory was achieved during a period of Parthian weakness that would not last indefinitely. In fact, prior Roman experience in the area had proven it difficult to exercise direct control over a territory largely under Parthian influence 32. Hence Nero’s position was much more realistic from a political point of view. On the other hand, nowhere in the sources can it be discerned that Nero and Corbulo were seriously at odds. If that were the case, the general would not have been distinguished with the governorship of Syria, an extremely high ranking position, which de facto allowed him to retain control over the Euphrates border. Furthermore, three years later, Nero granted him with absolute powers in the east, after a new rise in hostilities in the area. In sum, Nero’s decisions at this point in the conflict were perfectly coherent and logical in relation to military events.

It is important, because it reflects clear imperial guidelines from the very first moments: armed confrontation should be avoided at all costs by diplomatically accepting Tiridates as king of Parthia but always with Rome’s sanction. It seems therefore that Nero finally chose the third line of action (nominal sovereignty) to deal with the Armenian question. This approach was enforced by Corbulo’s actions throughout the conflict, with one exception, as will be seen ahead. 3) Tigranes, king of Armenia The only moment of exception in Rome’s attitude took place right after these events. Corbulo, taking advantage of Vologases’ problems in Hyrcania and the lack of decision displayed by Tiridates – who refused to present his petition before the emperor in Rome–, decided to take action. In consequence, Armenia’s two principal cities fell to the Romans. After such a decisive victory, it became necessary to decide on how to organize the new political situation in Armenia. Nero and Corbulo had different opinions on what measures should be taken, a disagreement used by Nero’s detractors to justify his presumed negligent attitude31.

4) The campaign of Caesennius Paetus Unfortunately, Tigranes was not very efficient at keeping peace in the area33. Incursions into Adiabene offered the perfect excuse for Tiridates, now fully aided by Vologases, to launch a new campaign aimed at regaining the Armenian throne (Tac.Ann.15.1–2). At this point Corbulo asked for help and Nero sent Caesennius Paetus to direct part of the troops. This episode has also been widely used to discredit Nero’s government, since the campaign led by Paetus resulted in complete failure, endangering all prior achievements in the Armenian question. For example, Hammond points out his bad decision in choosing Paetus and suggests that this is the moment when Nero actually started to personally direct foreign affairs in Armenia, whereas before he had been under the counsel of Seneca and Burrus (Hammond 1934, 94).

Tacitus offers a key text to understand this moment in the conflict. The Roman historian explains how Corbulo, after his victory, intended to take possession over Armenia (possessionem Armeniae usurpabat), but Tigranes returned to his throne with Nero’s sanction. Corbulo was forced to retreat and was subsequently assigned as the new Roman governor in Syria (Ann.14.26).

Although his decision did coincide, more or less, with Seneca’s and Burrus’ withdrawal from Nero’s circle of councillors, interpretations such as Hammond’s actually stem from a standard assumption: when Nero’s decisions were correct, he was influenced by his councillors, but when faced with failure, he had necessarily acted on his own. Such an insight is impossible to prove; it originates from the image portrayed of Nero by the classical sources. While the influence of Nero’s councillors on his political thought is generally accepted and particularly felt in the

“Having harried with fire and sword all whom he had ascertained to be against us, he began to take possession of Armenia, when Tigranes arrived, whom Nero had selected to assume the sovereignty”. It appears that Corbulo, encouraged by his military victory, decided to take initiative in the area and incorporate Armenia into the Roman Empire. A change was almost taken towards the first line of action, conquest, which would have converted Armenia into a Roman province. Nevertheless Corbulo was forced to abide by Nero’s decision. Although the emperor’s disposition towards the region had also changed after the general’s victory, he opted for a less intrusive measure, the second line of action or real sovereignty, which consisted on placing on the Armenian throne a monarch chosen by Rome.

32

And that will continue in the future, as will show the inability in maintaining the control over Armenia after the victories of Traian, Lucius Verus or Septimus Severus. 33 We do not share the theory of J. Wolski, who considers that Tigranes attacked Adiabene pushed by Rome, aiming to provoke Vologases (Wolski 1993, 167). Basically because we do not see what benefit will obtain Rome of this provocation, which will generate a situation that could break the balance existing in the area, now clearly favourable to Rome. We think that rather it was an individual action on the part of the new Armenian monarch, moved by the need to guarantee his new position. A movement which would show that, though to a general level Nero's decision to impose a pro–roman monarch seemed to be ideal, probably the choice of the Capadotian prince was not the best.

Differences in opinion between emperor and commander did not necessarily stem from personal dislike. Nero’s 31

For instance: Hammond 1934, 92.

91

JOAN OLLER GUZMÁN the imperial house36, emperors rarely received direct and reliable information of what really happened in their military campaigns (Millar 2004a, 194).

so–called quinquennium Neronis, going as far as considering Nero a puppet of his advisors is extreme. The next section will concentrate on this issue with more detail.

There are several aspects to highlight from Caesennius Paetus’ disastrous campaign. Firstly, the general proved himself incapable of running the campaign efficiently. Nevertheless his appointment was not random or impulsive; Nero was merely satisfying the request of his main man in the east, Corbulo. The emperor reacted swiftly, faced by a complicated situation and the demands of his general, showing coherence in his decisions and trust in Corbulo. The mistake lay in the specific choice of military command. Likewise other aspects must be taken into consideration that did not involve the emperor directly, such as the friction between the two generals, caused by their different characters and personal ambitions.

Going back on events, let us recall that Paetus suffered a great defeat when leading the Roman forces in Armenia, which ended in Tiridates recovering the throne 34. The Roman commander was clearly inept, lacking caution and military ability. Nevertheless it is worth noting other aspects that led to his disaster. For instance, Tacitus stresses the personal problems existing between Caesennius Paetus and Corbulo and openly speaks of their mutual dislike for each other. It is an interesting fact, because Corbulo had also fallen out with his previous peer, Ummidius Quadratus. Paetus reproached Corbulo for not having sealed a real and definite victory over the Parthians (Ann.15.6)35. On the other hand, Corbulo was once again unwilling to share his glory with any other commander; whereas events started raising suspicions in Rome, since after a first failed Parthian attack, Tiridates and Vologases were allowed by him to retreat (Ann.15.6). This situation affected the remainder of the campaign and, according to Tacitus, when things started getting complicated for Paetus, Corbulo preferred to delay his intervention in order to increase the glory of his future victory.

5) Absolute powers for Corbulo After Paetus was defeated, Corbulo managed to save the situation by stopping Vologases at the Euphrates and keeping the frontier, even if Armenia had been lost. In fact, the provincial governor and the Parthian king agreed to put an end to hostilities. The Romans would withdraw from their positions over the Euphrates, while the Parthians would abandon their garrisons in Armenia (Ann.15.17). Nevertheless the Armenian question remained unresolved and Vologases sent an embassy to Nero asking for Tiridates to be recognized on the throne. The emperor’s reaction to the request offers some useful insight into his actions as a ruler.

“Only with a struggle, it is said, could he be brought to admit the hostile pressure to Corbulo. Nor was there any haste on the part of Corbulo himself, who hoped that, if the dangers came to a head, the glory of a rescue would also be heightened” (Ann.15.10).

Nero interpreted the letter as a mock asking to confirm a reality that already existed de facto. He rejected the Parthian request and decided to consult on the next step to take among the principal Roman citizens (Tac.Ann.15.25).

Paetus also decided to send false reports to Nero, speaking of great victories and progress in Armenia, when his initial military movements had not been successful at all. Misinformed, the emperor did not issue new orders until the situation got so extreme that news from Armenia inevitably reached Rome (Tac.Ann.15.8):

“The irony of the barbarians in asking for what had been taken was now obvious, and Nero held a council of state to decide the choice between a hazardous war and an ignominious peace”.

“… he therefore led back the army, and, to give the impression that the war was now closed, indited a letter to the Caesar, as grandiloquently phrased as it was void of content”. Therefore, the problems experienced throughout the Roman world to convey information across great distances was definitely another influencing factor. In fact, during the early imperial period, with military command still in the hands of senators not belonging to

Therefore we enter again into the question of the Imperial councillors and their influence to the throne. F. Millar has studied the role played by these imperial councillors (Millar 2004a; Millar 2004b). Based on literary sources, the author claims that questions related to military and foreign affairs were closely linked to the emperor. If there was a select group of secretaries or specialists dealing with these affairs, no evidence has survived to prove it. The only certainty is that the emperor could at any point consult with people of his close circle, such as praetors, consuls, centurions or senators, who would in all likelihood, be the ‘principal citizens’ (primores civitates) mentioned by Tacitus (Millar 2004a: 166 – 168). In

34

In favour of Petus, it is important to remark that, unlike in Corbulo’s campaign, in this case he had to face the totality of the Parthian forces, since Vologases already did not have to be dependent on distractions as the Hyrcanian revolt (Farrokh 2007, 149). 35 And, in fact, he was not lacking certain reason, since Corbulo had conquered Artaxata and Tigranocerta without practically having to use the force, since both cities surrendered and opened their doors in front of the Roman troops.

36

In fact, the Roman–Parthian war of Nero was the last great military campaign in the East directed by this type of commands, not linked to the imperial house (Millar 2004b, 225). From this moment, they will be the emperors in person who will direct the campaigns and will negotiate with the enemies.

92

N e r o , P a r t h i a a n d t he ‘ A r m e n i a n Q u e s t i o n ’ the policy of the 3rd C emperors. Finally, Tacitus’ narration is much more coherent with the overall evolution of the conflict. Therefore, in our opinion, Tacitus is probably a more reliable source.

reality, councillors of the early empire did not wield any legal power, so the emperor could consult with them at his free will; the final decision was always taken by him (Millar 2004a, 170). The fact that Nero consulted with trusted men, first with Seneca and Burrus and later with these principes civitates, does not necessarily indicate weakness or lack of ability, but more probably the usual actions taken by Roman emperors in the making of decisions related to foreign policy and also some sensible prudence by Nero in front of a highly delicate situation.

6) The crowning of Tiridates in Rome Even though armed confrontation ended in AD 63 with the agreement between Tiridates and Corbulo, the conflict was not finalized until three years later, when the Arsacid king fulfilled his promise and travelled to Rome to receive the crown from the hands of Nero himself. The description of the ceremony may be found in Suetonius (Nero.13) and Cassius Dio (63.1–7), since Tacitus’ Annals are lost as of AD 66. Both authors coincide in the grandiloquence of the event and Dio quotes that it cost the Roman treasury 80,000 denarii38.

Nero finally resolved to go to war and for that purpose granted Corbulo absolute powers, which gave him the capacity to recruit troops and the submission of all vassal kings and tetrarchs, as well as all governors, procurators and praetors from neighbouring provinces. Tacitus states that such absolute power had not been seen in Rome since Pompey (Ann.15.25)37. From this point onwards, Corbulo threatened the Parthians with open war by unleashing a series of small attacks that spread fear among the enemy. Nevertheless, in the end, he reached an agreement with Tiridates. The Parthian king would renounce his crown with the promise that it would be returned later from the very hands of the emperor in Rome. This passage is a reflection of the three main aspects that characterize Roman policy during the conflict: Nero’s complete confidence in Corbulo; diplomacy over armed conflict; and commitment to the third line of action or nominal sovereignty.

The importance of the ceremony must have been very great, both for Nero and the Parthians. This moment has been considered as the culmination of Nero’s political activity (Fernández 1991, 204). The whole celebration must be understood as an act of reaffirmation and political propaganda, very particularly in the eyes of foreign allies and enemies, but also internally, broadcasted through the celebration of games, the embellishment of Rome, etc. According to Dio, Nero presented himself as supreme governor of the known world. His words very openly addressed his enemies, especially the Parthians, extolling on the capacity of the Roman emperor to rule over the destinies of all peoples and their monarchs 39. In this way Nero formalized the concept of ‘rex datus’ (even if Tiridates had not been initially chosen by him) and at the same time celebrated the success of his government, which was reflected on a coin issue depicting a closed temple of Janus, symbolizing peace throughout the empire40. The complexity of the whole affair was full of messages to the Roman people. Somehow Nero sold an image of victory, trying to show that he had achieved what Rome had been yearning for since the Republican period, because not even Augustus had succeeded in subduing the Parthians (Lerouge 2007, 147 – 149). In addition, we have there the best example of Nero’s political intention, which was to progressively veer the Roman government towards an eastern style monarchy.

It is interesting to notice that the historical account offered by Cassius Dio differs a lot from Tacitus. He interprets these events as a clear demonstration of Nero’s lack of ability in comparison to Corbulo’s good sense. According to this other source, Nero did not want a peaceful agreement and planned to continue the armed conflict at any cost, even if it meant leading it himself (D.C.62.22). “Nero had intended to accompany the expedition in person, but he fell while performing a sacrifice, so that he did not venture to set out, but remained at home”. In the end, Corbulo secretly advised Vologases to send Tiridates to Rome to supplicate the crown from Nero, forcing the emperor to act as a condescending ruler. Dio offers a pejorative image of the emperor, comparing him to a much more capable and shrewd general. However the account offered by Tacitus reveals itself as the most plausible one. To start with, it was written much closer in time to events. Furthermore in the 3rd C AD Dio’s political stance was contrary to the senatorial faction that supported territorial expansion in the east promoted by emperors such as Septimius Severus (Lerouge 2007, 52). Nero’s legacy of the 1st C AD would not have been considered a good example among those who criticized

38

It seems that both descriptions are based on the detailed account that Pliny offered in his Naturalis Historia, unfortunately lost except for some little details about the decoration of the city done in honour of the visit of the Armenian king (NH.19.24; XXXII.54). 39 This is especially evident in the reconstruction that the classical author offers of the words that the emperor directs to Tirídates after this one knelt down and begged for his crown (63.5.3): “Well hast thou done to come hither in person, that meeting me face to face thou mightest enjoy my grace. For what neither thy father left thee nor thy brothers gave and preserved for thee, this do I grant thee. King of Armenia I now declare thee, that both thou and they may understand that I have power to take away kingdoms and to bestow them”. 40 From 64 AD until 66–67 AD Nero coins diverse monetary series with the legend: PACE P.R. TERRA MARIQ. PARTA IANUM CLUSIT S.C. On these issues and the question of the closing of Jano’s temple by Nero, vid.: Townend 1980.

37

It would consist in an imperium maius for the entire east (Vervaet 2002, 136). Pompey obtained it the 67 BC during his campaign againts the pirates.

93

JOAN OLLER GUZMÁN However for some authors the whole coronation was nothing more than an act that the Parthians agreed to, so as to ascertain their hold on their territory and achieve their political goals. For example, Ch. Lerouge defends that there was no real submission of the Parthians to Rome. They only wanted to avoid armed conflict with Rome and for that reason agreed to certain gestures that would appease Nero in a situation in which forces had proven to be equal (Lerouge 2007, 140 – 144). It was unquestionably also a great victory for the Parthians, because they managed to retain their pre–eminence over their traditional area of influence, Armenia, and to do so with Rome’s sanction. Moreover both powers were treated as equals in the treaty. Even though they accepted Nero’s power to crown their king, it was the first time that Rome granted a status quo in terms of equality in political and military affairs in the east. Therefore the war did end as a great diplomatic victory for Vologases, although it does not mean that Nero’s role should be darkened because of it. Rome not only gained nominal sovereignty over the Armenian throne, but also humiliated a member of the Arsacid royal family. Cassius Dio explains that the humiliation only took place, because of the great gains for the Parthians. However the act was highly charged with symbolism, which reflected the power and influence of Rome in the east during Nero’s rule.

general and the emperor, or at least, did not influence the development of events. 6. Conclusion: lights and shades in Nero’s eastern campaign The Roman–Parthian War fought in Armenian territory marked a point of inflection in the diplomatic relationship between both powers. Nero’s government continued in the Roman tradition of confrontation with the Parthians, which had caused great instability in the eastern limes since the beginning of the 1 st C BC. However, during Nero’s government, a new period of conflict was inaugurated with episodes of armed confrontation that had not broken out since the agreements reached in times of Augustus. Roman policy was centred around two main strategies: avoid armed conflagration when possible, evidenced in Corbulo’s preference for negotiation; and install nominal sovereignty in Armenia, described in this paper as the ‘third line of action’, which meant that any Parthian sitting on the Armenian throne would need to be sanctioned by the Roman emperor. The consistency in Rome’s foreign policy would only be broken after Corbulo’s victorious campaigns between AD 58 and 60, which granted Rome a very favourable position to attempt imposing real sovereignty with the crowning of the Roman client king Tigranes. The ‘second line of action’ failed, evidencing that the initial policy, defended by the imperial throne, was the most adequate. The ‘second line of action’ had already been attempted prior to Tigranes, during the reign of the Julio–Claudian emperors Tiberius and Claudius, with no successful result. Real sovereignty was never accepted by the Parthian monarchs and therefore no real stability was possible in the region. On the other hand, the ‘first line of action’ –conquest and incorporation of new territories to the empire as a Roman province– was never in reach. Failed attempts transpired during the 1st C BC and in later campaigns by the Antonines and Severans. Rome never managed to establish permanent control over any territory beyond the Euphrates River. In conclusion, Nero’s advocacy for nominal sovereignty may be considered as a definite point of inflection in the Armenian question and a great success in his foreign policy, leading to fifty years of peace in the area, which allowed Rome to reinforce and consolidate the eastern limes. Since Rome’s arrival in the east, a century and a half earlier, the region had not enjoyed such a favourable political atmosphere, which would not repeat itself for another 150 years. The end to the Armenian question was not only aided by a successful imperial outlook on the geopolitical situation of the area, but also by the favourable disposition of the Parthians, who accepted the ‘third line of action’ as a means to maintain control on a historically Arsacid area of influence. Both Rome and Parthia could claim to have been successful. Although Rome lost some influence over Armenia, the empire still played a fundamental role in the sovereignty of the kingdom. Moreover peace brought stability to the eastern limes and converted an enemy into an ally. Meanwhile Parthia recovered its influence over Armenia

7) Corbulo’s Death Although Corbulo’s death transpires after the end of the conflict, it offers valuable insight for the debate on Nero’s figure. Domitius Corbulo committed suicide in AD 67 in Greece, after learning that Nero intended to have him executed. This event has been used to highlight the hypothesis backing Nero’s dislike for the general (Hammond 1934, 83). Several aspects have already been mentioned above to defend their relationship. There is no clear evidence for an enmity between emperor and general during the war. Much to the contrary, the solution to the Armenian question only arrived after Nero granted absolute power to Corbulo in the east. In AD 67 the war was long gone and the political situation had become very complex, with Nero trying to maintain himself in power by applying very harsh measures of repression. One of them was directed against Corbulo, a man of great reputation, esteemed by senators and the people41. F. J. Vervaet stresses Nero’s complete trust in Corbulo from the very beginning of the conflict and points that it was a conjunction of varied situations that led to the break up of their relationship. Furthermore he rejects the general’s participation in the revolts of Rubelius Plautus or Piso. It was probably the suspicion that Corbulo’s son in law participated in the revolt of Vinicianus, which put the emperor on guard. Given the reputation, power and influence of the general, Nero decided to take the most expedite of decisions (Vervaet 2002, 136 – 172). However, throughout the resolution of the Armenian question, no personal problems existed between the 41

As indicates Dio, who says that a lot of people were considering him to be sufficiently qualified to replace the emperor at the head of the Empire (D.C.62.19.3–4).

94

N e r o , P a r t h i a a n d t he ‘ A r m e n i a n Q u e s t i o n ’ and gained recognition from Rome as its political and military alter ego in the Near East.

the Parthians. Nero presented himself as regent, not only of the empire’s destiny, but also of all of the kingdoms of the known world, in tune with his ambition of turning Rome into an eastern autocratic monarchy.

The second aspect analyzed in this paper was Nero’s role as a ruler during the conflict, questioned by the diverse and extreme interpretations, which have surrounded his figure. A detailed revision of the literary sources shows that the image of the Julio–Claudian emperor should be reassessed. This paper has shown that his foreign policy concerning the Armenian question was largely consistent from the beginning of his reign. Furthermore his decisions proved to be successful, especially the appointment of Corbulo as chief military commander in the area. His trust in him was ratified by granting him absolute powers in AD 63. Naturally not all his decisions were the right ones, the most evident mistake being his choice of Caesennius Paetus to aid Corbulo. Nevertheless these errors are not a reflection of an erratic or capricious foreign policy, but only punctual mistakes in an overall coherent direction. It is difficult to determine the degree of influence played by his imperial councillors, such as Seneca or Burrus. However it is highly improbable that Nero delegated his decision making completely, reduced to the role of a passive agent. The ceremony that crowned Tiridates is the clearest example of the importance that the emperor placed on his eastern affairs. The occasion was used as a means of internal propaganda, as well as an outspoken message directed to Rome’s foreign enemies, especially

The propaganda was coherent with the concept of ‘reges dati’, which presented Nero as the giver/taker of kingdoms42. In reality, we know that Nero did not actually ‘give’ the throne to Tiridates, but just confirmed it. In this sense, this episode cannot be considered a case of ‘rex datus’43. Rome would have to wait until the 2nd C AD to exercise such a power in Armenia.44 Finally, let us recall the ‘false Neros’ mentioned at the beginning and the regard found for this imperial figure in Parthia. After having understood the events leading to the conclusion of the Roman–Parthian War, it is easier to explain Nero’s fame. Imperial policy allowed Parthia to reinforce its position in Armenia and put an end to confrontation with its most dangerous enemy. Furthermore Parthia obtained a crucial diplomatic recognition from Rome, which positioned the kingdom as the leading Asian power in the east. For these reasons, Nero’s figure remained in high esteem among the Arsacids. He was also liked by the popular classes, favoured by messianic traditions and his outspoken admiration for the eastern provinces of Hellenistic influence. It comes as no surprise then that his memory remained vivid for a long time in the east.

42

Especially through the already mentioned passage of Dio. Vid. n. 49. 43 As said by Lerouge 2007, 139 – 143, who defends that, in spite of the fact that Tiridates appears as slave of the emperor and to the different elements of submission imposed in the ceremony, actually we would be in front of symbolic acts accepted by the Parthians that were not implying a real submission. 44 For example, under Traian who in the year 117 AD imposed Parthamaspates as king in Parthia. Or under Marcus Aurelius, when in 163 AD Lucius Verus imposed Sohemus as king in Armenia: Bivar 1983, 91 – 93.

95

JOAN OLLER GUZMÁN Bibliography BARRET, A. A. 1979, “Annals 14.26 and the Armenian settlement of A.D. 60”, CQ 29.2, 465 – 469.

GRAJETZKI, W. 2011, Greeks and Parthians. Mesopotamia and beyond. 331B.C. – 224 A.D., London/Bristol.

BIVAR, A. D. H. 1983, “The political history of Iran under the Arsacids”, The Cambridge History of Iran, 3 (I). The Seleucid, the Parthian and the Sasanian periods, Cambridge, 21 – 99.

GRIFFIN, M. T. 1987, Nero. The end of a dynasty, London. GRISHAM, R. 1951, L’Iran. Des origines à l’Islam, Paris.

BUTCHER, K. 2003, Roman Syria and the Near East, London.

HAMMOND, M. 1934, “Corbulo and Nero’s eastern policy”, HSCPh 45, 81 – 104.

CARTER, M. 2004, “The lex portorii and the Armenian campaign of Caesennius Paetus”, Latomus 63/2, 370 – 379.

HEKSTER, O. 2010, “Trophy kings and Roma power: a Roman perspective on client kingdoms”, Kingdoms and principalities in the Roman Near East, Stuttgart, 45 – 56.

CARY, E.(transl.) 1968/1970, Cassius Dio: Roman History, London,

HENDERSON, B.W. 1901, “The chronology of the wars in Armenia, A.D. 51 – 63”, CR 15.4, 204 – 213.

CHAMPLIN, E. 2003, Nero. Harvard. — 1903, The life and principate of the emperor Nero, London.

CHARLESWORTH M. P. 1950, “Nero: some aspects”, JRS 40, 69 – 76. CIZEK, E. 1982, Néron, Paris.

JACKSON, J. (transl.), 1969, Tacitus: Annals, V. XIII – XVI, The Loeb Classical Library, Cambridge.

DABROWA, E. 2011, “The Parthian kingship”, Studia Graeco–Parthica. Political and cultural relations between Greeks and Parthians, Wiesbaden, 111 – 122.

LEROUGE, CH. 2007, L’image des Parthes dans le monde gréco–romain. Du début du Ier siècle av. J.C. jusqu’au la fin du Haut–Empire romain, Stuttgart.

ELDWELL, P. M. 2008, Between Rome and Persia. The Middle Euphrates, Mesopotamia and Palmyra under Roman control, London/New York.

MATTHEWS, E. 1937, “Nero and the East”, HSCPh 48, 75 – 103. MILLAR, F. 2004a, “Emperors, frontiers and foreign relations, 31 B.C. to A.D. 378”, Rome, the Greek world and the East, vol. 2: Government, society and culture in the Roman Empire, London, 160 – 194.

ELLERBROCK, U., WINCKELMANN, S. 2012, “Die Herrschaftsstruktur der Parher”, Die Parther. Die vergessene grossmacht, Darmstadt, 117 – 134. ELSNER, J., MASTERS, J. 1994, Reflections of Nero: culture, history and representation, London.

MILLAR, F. 2004b, “Government and diplomacy in the Roman Empire during the first three centuries”, Rome, the Greek world and the East, vol. 2: Government, society and culture in the Roman Empire, London, 195 – 228.

FACELLA, M. 2010, “Advantages and disadvantages of an allied kingdom: the case of Commagene”, Kingdoms and principalities in the Roman Near East, Stuttgart, 181 – 198.

MOMIGLIANO, A. 1931, “Corbulone e la politica verso i Parti”, Atti del IIº Congresso Nazionale di Studi Romani, Roma, 368 – 375.

FARROKH, K 2007, Shadows in the desert: ancient Persia at war, Oxford.

OLLER, J. 2009, “Quan uellem nescire litteras. El De Clementia de Sèneca i la seva aplicació durant el quinquennium Neronis”, Comprendre 11.2, 41 – 61.

FERNÁNDEZ, P. 1991, “Nerón y neronismo: ideología y mito”, Espacio, tiempo y forma Serie II Historia Antigua 4, 199 – 222.

PANI, M. 1972, Roma e i re d’Oriente da Augusto a Tiberio. Cappadocia, Armenia, Media Atropatene, Bari.

FOWLER, R. 2010, “King, bigger king, king of kings: structuring the power in the Parthian world”, Kingdoms and principalities in the Roman Near East, Stuttgart, 45 – 56.

PAPPANO, A. E. 1937, “The false Neros”, CJ 32, 385 – 392.

96

N e r o , P a r t h i a a n d t he ‘ A r m e n i a n Q u e s t i o n ’ RAHIM SHAYEGAN, M. 2011, Arsacids and Sasanians. Political ideology in post–hellenistic and Late Antique Persia, Cambridge.

TOWNEND, G. B. 1980, “Tacitus, Suetonius and the temple of Janus”, Hermes 108/2, 233 – 242. VAN OVERMEIRE, S. 2012, “According to the Habit of Foreing Kings: Nero, Ruler ideology and the Hellenistic Monarchs”, Latomus 71.3, 753 – 779.

SCHUR, W. 1963, Die Orientpolitik des Kaisers Nero, Neudruck der Ausgabe. SHERWIN–WHITE, A. N. 1984, Roman foreign policy in the east. 168 B.C. to A.D. 1, Norman.

VERVAET, F. J. 2002, “Domitius Corbulo and the Senatorial opposition to the reign of Nero”, AncSoc. 32, 135 – 194.

SIMPSON, C. J. 2005, “Where Is the Parthian? The Prima Porta statue of Augustus Revisited”, Latomus 64.1, 82 – 90.

WENKE, J. 1981, “Elymeans, Parthians and the evolution of empires in southwestern Iran”, Journal of the American Oriental Society 101 nº 3, 303 – 315.

SOMMER, M. 2010, “Modelling Rome’s eastern frontier: the case of Osrhoene”, Kingdoms and principalities in the Roman Near East, 217 – 228.

WILCOX, P. 1986, Rome’s enemies (3). Parthian and Sasanid empires, Oxford.

ROLFE, J. C. (transl.) 1979, Suetonius: Life of the twelve Caesars, Volume II. The Loeb Classical Library. London.

WOLSKI, J. 1993, L’empire des Arsacides, Lovanii. ZANKER, P. 1990, The power of images in the Age of Augustus, Ann Arbor.

SUSPENE, A. 2012, “Images royales en contexte romaine: les rois étrangers sur les monnaies romaines (II siècle av.J.C. – II siècle ap.J.C.)”, Cahier des Études Anciennes 49, 259 – 279.

97

The relations of the Sasanian Empire with the Eurasian Steppe NÚRIA OLAYA MONTERO Universitat Autònoma de Barcelona ABSTRACT This paper presents a multidisciplinary approach to the study of the relations of Sasanian Iran (AD 224–651) with Central Eurasian steppe nomads, which can be classified in four categories: military conflicts, alliances, taxation and trade and, finally, cultural contacts. This paper offers a sample of historical examples about these relations based on the information provided by Classical and Oriental sources using several sciences to analyse the data. KEYWORDS Sasanian Empire – Central Eurasian steppe nomads – sources The main objective of this paper is to present our study on the relations between Sasanian Iran, named after the ruling dynasty in Persia from the first quarter of the 3rd century AD to the second half of the 7th century AD, and the Central Eurasian steppe nomads (Iranian and / or Turkic speaking). Our study focus on the Northeastern borders of the Persian Empire, paying special attention to the steppes of Central Asia, according to the data collected by contemporary and later sources in different languages. Documents written in Middle Persian, the language spoken under the Sasanian dynasty, are rather scarce, but they are of great importance given that they furnish us with the most direct and first–hand information. Nonetheless, we have to rely on information in other languages in order to gain access to more data. Mostly, we work with Greek and Latin, but also Bactrian, Sogdian, Armenian, Syrian, Georgian, Persian, Arabic, Chinese and Ethiopian sources. Although the steppe nomads sporadically knew some writing systems after adopting the alphabets devised for Northeastern Iranian languages (Bactrian and Sogdian), they did not create a literature stricto sensu, even though there are some written records like inscriptions, letters and numismatic legends. I am going to focus on the Northeastern Sasanian border, which included the Persian provinces of Gorgān, Merv, Ḫārizm, Ḫorāsān, Sīstān, Bactriana or Ṭoḫārestān and Sogdiana. Currently, these regions correspond to the Northeast of Iran, Turkmenistan, Uzbekistan, Southern Kazakhstan, Tajikistan, Afghanistan and Northern Pakistan. I also add the old adjacent territories of Tūrān, Hindukuš, Hind, Gandhara, Čīn and the vassal kingdom of the Kušānšahr until its destruction in the 4th century AD. Therefore, the geographic delimitation of this study is expanded up to China and Northern India, but such an approach is justified by the Persian expansion or contact with these regions. The Caucasian boundary is not covered here but I would like to stress its importance due

to the fact that it was a focal point of penetrations for some nomadic populations that came from West Central Asia and the Eurasian steppes, like Huns, Ogurs, Sabirs, etc. A detailed discussion of each one of the Central Asian peoples and regions that maintained some kind of relations with the Sasanians is beyond the aim of this paper, although they are dealt with in the commentaries of the historical examples provided. However, a short presentation of these peoples and their origins is given for the sake of clarity. The first category is that of the Iranian peoples, mostly associated with Northeastern Iranian languages, like Bactrians and Sogdians; the Kušāns, who adopted the ‘Bactrian’ language and writing, but also used others (eg. Bhrāmī); the Chorasmians, settled in Ḫārizm; the inhabitants of Sīstān; the region of Rob, located almost in the centre of Afghanistan and the region of Tūrān, originally used to describe the territories occupied probably by Iranian peoples beyond the Oxus / Amu Daryā river ‒considered the natural border between Iran and the nomads. The second group is that of the Turkic peoples, among whom are the Turks and the same region of Tūrān, which over time came to designate the Turks and the Khazars. Thirdly, the trio integrated by Chionites, Kidarites and Hephthalites, on which there is a great discussion about their language, differently identified as Northeastern Iranian or Turkic. Finally, the Huns, whose linguistic affiliation cannot be asserted with certainty. Before explaining the contacts between Persians and nomadic peoples, a serious problem connected with the sources should be considered: the anachronism. Many authors, in order to update their text or following literary tradition, use wrong names for steppe peoples, calling all kind of nomads from Central Asia Huns and, later on, Turks, regardless of their chronology, which leads to inaccurate statements. Another problem is the ambiguity of the repeated mentions of the nomadic peoples as

NÚRIA OLAYA MONTERO ‘neighbours’ or ‘enemies’, without any further description, which complicates the identification of the nomadic group involved in the contacts. In addition, the name of Tūrān in Ferdūsī’s Šāhnāme presents occasional synonyms with Čīn, identified as Chinese Turkestan and China, but it is also used for the Turks, like in the title ḫāqān–e Čīn, the Great Qaγan of the Turks. The relations between Sasanians and Central Asian peoples can be divided into four main categories: military conflicts, alliances, taxation and trade and cultural contacts. For the sake of brevity, only one example of each category is described in detail. Military conflicts Military conflicts or confrontations are already mentioned from the beginning of the Sasanian dynasty with Ardašīr I until the destruction of the Persian Empire under Yazdgird III. Numerous struggles with nomadic peoples on the borders with Central Asia are documented, except during the reigns of 12 Sasanian kings or queens 1. However, it should be noted that the absence of confrontations with Central Asian nomads under these monarchs in the sources coincides with relatively short reigns or periods of internal conflicts in Persia, which could be the reason for the lack of further information. However, in some cases, the numismatic evidence offers us more data than the sources and helps us to fill the gaps left by the authors. Concerning the other 21 kings, there are accounts of their struggles and occasional wars against nomadic populations. Conflicts were almost always motivated by Sasanian expansionist ambitions or by nomadic destructive incursions into the Persian borders. In order to stop and prevent raids from the North, the Persian kings built several defensive walls with garrisons in the borders, such as the wall of Gorgān and the wall of Tammiše. It also appears mentioned in the sources that many of these nomadic peoples took advantage of the Sasanian king’s death to attack Persia in its weakness. Example The date of Ardašīr I’s occupation of some Central Asian territories is disputed: a first theory says that the events took place in AD 225, a year after Ardašīr destroyed the Parthian Empire and once he obtained the title of ‘king of kings’ (šāhānšāh)2; a second one situates it after the wars with the Romans, between AD 233 and AD 235 3; and a third one places the occupation under the reign of Ardašīr’s son, Šābuhr I4. The closer source to the events is the inscription of Šābuhr I at Ka‘be–ye Zardošt (post 260 AD). This inscription recorded several offerings for the souls of 1

Vahrām III (293), Hormizd II (302–9), Ardašīr II (379–83), Šābuhr III (383–8), Vahrām IV (388–99), Kavād II (628), Ardašīr III (628–30), Šahrvarāz (630), Bōrānduxt (630–1), Pērōz II (631), Āzarmīgduxt (631), Hormizd V (631–2). 2 Beckwith 2009, 85; Bivar 2006, 203 and 209; Widengren 1971, 747. 3 Harmatta 1965, 186–194. 4 Frye 1984, 295; Narain 1994, 169.

100

many dignitaries who had lived under Ardašīr. I highlight three individuals5: *Sadāluf, King of Ab(a)rēnag6, Ardašīr, King of Merv7 and Ardašīr, King of the Sakas8. The votive enumeration in the inscription corresponds to a list of local kings who were semi–independent of the Sasanian central authority9. *Sadāluf (disputed etymology) was the ruler of Ab(a)rēnag, corresponding to Abaršahr or Nīšābūr in Ḫorāsān. There is a possibility that *Sadāluf might be a Kušān vassal who offered his submission to Ardāšir10. The King of Merv governed the city and the province which corresponded to ancient Margiana, nowadays the Southern territories of Turkmenistan. The King of the Sakas ruled the ancient Achaemenid regions of Drangiana and Arachosia in Southern Afghanistan. The name of the Sakas comes from the Eastern branch of the Scythians who overran in the second half of the 2nd century BC those areas near the regions of the Indus and the Northwest of India, adjacent to the kingdom of the Kušāns. Ṭabarī and Ibn al–Aṯīr relate that Ardašīr I travelled around the provinces of Siǧistān, Ǧurǧān, Abaršahr [Nīšābūr in Ibn al–Aṯīr], Merv, Balḫ and Ḫārizm to the furthest boundaries of Ḫorāsān killing many people and returning again to Merv. Shortly afterwards, the king received in the city of Ǧūr embassies from the Kušāns and Ṭūrān offering their submission11. Dīnawarī confirms the presence of Ardašīr in Ḫorāsān and Siǧistān12. The Nihāyat al–Irab asserts the same statement but adds the length of the stay in the city of Merv (one year) and the collection of land taxes13. Ferdūsī only mentions the extension of Ardašīr’s realm, from Rūm to Čīn and from the land of the Turks to Hind 14. Ardašīr I seems to conquer or make vassals of all the peoples on the Northeastern border according to Oriental sources, although it is doubtful that the kingdom of the Kušāns (established as from the 1st century BC), which included the regions of the Hindukuš mountain range, was completely annexed. Perhaps various territorial divisions and small local dynasties maintained a relative independence until the total conquest accomplished by

5

ŠKZ §41 (Huyse 1999, 54): mp. *Sadāluf ī Abrēnag šāh; Ardašīr ī Marw šāh; Ardašīr ī Sagān šāh | gr. Σαταροπτ Ἀβρηναχ βασιλέως; Ἀννρταξάρου; Μαρου βασινλέως; νἈρτναξιρ Σιγαν βασιλένως. 6 MP. ’plynk; Ar. Abaršahr. 7 MP. mlw’, mwlw’ /marv, murv/; OP. Margu–; Av. marγu–, mouru–; Sogd. m(r)[γ]; Gr. Μαρου, Μαργιανή; Arm. Marg, Marv; Ar. Merw; Chin. Mo末. 8 MP. sk’n, skst’n /sakān, sakistān/; Sogd. sgst’n; Gr. Σιγαν, Σεγιστανή, Σακαστανή; Lat. Sacae, Segestani; Arm. Sagastan, Sakastan; Syr. Sgstn; Skt. śaka; Ar. Saǧastān, Siǧistān; , Sīstān; Chin. Sai 塞. 9 Lukonin 2007, 729. 10 Frye 1984, 294–295. 11 Ṭabari I, 819–20 (Bosworth 1999, 15). Ibn al–Aṯīr (Widengren 1971, 767). 12 Dīnawarī (Widengren 1971, 772). 13 Nihāyat al–Irab (Widengren 1971, 770). 14 Šāhnāme 22, 431–2 (Mohl 1976 [V], 369).

The rel at i ons of t he S asani an Em pi re wi t h t he Eurasi an S t ep pe Šābuhr I15. The king of the Kušāns at that time was Vāsudeva I or II, who has been linked with the king of the Da Yuezhi 大月氏 known as Bodiao 波調 (229–30). Bodiao sent an embassy with taxes to the court of the Chinese Emperor Cao Rui 曹叡 (226–39) of the Wei 魏 dynasty in 229/30, reported by the Weishu 魏書16. This embassy is considered a call for help or assistance against the Sasanian occupancy17. Persian rule was established in the Hindukuš, the so–called Kušānšahr, a region with Kušān population but ruled by Sasanian princes until the arrival of the Chionite / Kidarite Huns during the 4th century AD. In the sources, the term Ṭūrān does not designate the people beyond the Oxus River or the Turks as usual, but it refers to the Baluchistan region in Southern Persia, near the Persian Gulf18. Alliances Both military and dynastic alliances are recorded. Military alliances could be a coalition against common enemies, i.e. a partnership in which both parties had a shared interest or profit attacking others, or, in certain cases, a mercenary agreement. The latter type of association could be caused by a previous Persian subjugation and, therefore, the participation of the nomad population was ‘forced’ or, in other cases, encouraged by the hiring of nomadic troops leaded by their kings. Alliances were not only between Persians and nomads, as there were also alliances between Romans / Byzantines and nomads against Persia. The dynastic alliances were achieved through arranged marriages between steppe princesses and Persian kings, whether the marriage was claimed by the Sasanian king himself or it was a nomadic offer. Also noteworthy is one case of alliance in which the Persian monarch was willing to send his son to the court of one of these outlying kings to receive his education. Example In the year 558, the Hephthalite Huns 19, a Central Asian nomadic people inhabiting areas of Bactria and Sogdiana and ruling them since c. AD 450, fell under the onslaught of a personal alliance between the Persian king Husrav I (AD 531–79) and the leader of the Western Turks, the Qaγan known as Silzibul or Sizabul in Greek sources (Σιλζίβουλος, Σιζάβουλος) and as Sinǧibū ḫāqān in Arabic tradition (Old Turkic Sri yabγu), though, his true Turkic name seems to have been Ištämi and the former forms would be variants of a Turkic title. However, there are critical voices against this identification20.

15

Bivar 2006, 203; Chegini–Nikitin 1996, 41–42; Dani 1996, 167; Narain 1994, 169; Widengren 1971, 748. 16 Weishu III 6a (Zürcher 1968, 371). 17 Narain 1994, 170. 18 Bosworth 1999, 15. 19 Bactr. ηβοδαλο; Sogd. *Heβtalak [pl. *Heβtalit]; Gr. Ἐφθαλῖται οr Νεφθαλῖται / Ἀβδέλαι; Syr. eptalī / ạβdel; Arm. Hep‘t‘alk‘; Ar. Haiṭāl [pl. Haiāṭela]; MP. hēftāl / hyōn; Chin. Yida 挹怛 / Hua 滑 / Yanda 嚈噠. 20 Sinor 1994, 302–305.

101

The war and the subsequent Hephthalite fall is narrated in Greek and Eastern sources, although with some variants. Firstly, the fall of the Hephthalites seems to be attributed to the Turks. Menander the Guardsman reproduces a speech by Silzibul in which he stated that he was willing to make the war against the Hephthalite Huns21. In addition, a Turkic embassy confirmed to the Byzantine Emperor Justin that the Turks, led by Sizabul, had conquered the Hephthalite Huns and had imposed taxes on them22. Theophylact Simocatta also attributes to the Turkish Qaγan the victory over the Hephthalites23. The Zhoushu 周書 specifies that AD 558 was the last year in which the Hephthalites sent an embassy to the Zhou 周 dynasty and were smashed by the Tujue 突厥, the Chinese name for the Kök Türk, the Celestial or Blue Turks24. Ištämi was the ruler of the Western Turks, as it is written in the Old Turkic inscriptions on the funerary complexes of Kül Tigin (†AD 732) and Bilgä Qaγan (†AD 735) in Northern Mongolia. These epigraphs show that Bumin (AD 534–52), first Great Qaγan of the Turks who had revolted against their oppressors, the Ruanruan 蠕蠕, and his brother Ištämi expanded their realm up to the Iron Gate (Tämir qapïγ) and conquered all those territories25. The aforementioned ‘Iron Gate’ is a gorge located between the cities of Balḫ and Samarqand, the capitals of Bactria and Sogdiana, in the centre of the Hephthalite territories. Therefore, these inscriptions demonstrate the Turkish sovereignty over those lands and corroborate the information provided by the sources. However, Menander adds a speech by Zich (gr. Ζίχ)26, a Persian ambassador to Byzantium in AD 561, who stated that Husrav I had destroyed the power of the Hephthalites 27. This assertion seems to contradict all previous information, even that by Menander himself. Persian and Arabic sources deal with these facts in a different way, alluding to a Turkish–Persian alliance against the Hephthalites, ratified by the marriage between Husrav I and the daughter of the Qaγan, the mother of Hormizd IV, Husrav’s successor. There are three lines of literary tradition in the sources about the fall of the Hephthalites28: firstly, Ferdūsī and Ṭabarī attribute to the Ḫāqān–e Čīn / Sinǧibū ḫāqān the death of the Hephthalite King (called respectively Ġātfer and W.r.z = *Warāz?, etymologies unknown); secondly, Dīnawarī, Ṭabarī again, Masʽūdī and Taʽālibi bestow the victory on Husrav I, 21

Menander the Guardsman fr. 4,2, l. 7–11 (Blockley 2006, 46). Menander the Guardsman fr. 10,1, l. 70–79 (Blockley 2006, 114). 23 Theophylact Simocatta VII 7, 7 (De Boor–Wirth 1972, 256). 24 Zhoushu fol.15a, section F, columns b–d (Miller 1959, 12, 72). 25 Kül Tigin E 1–3 and Bilgä Qaγan E 3–4. (Tekin 1968, 232, 263–4). 26 A family name? His possible true name might have been Yazdušnasp (Gr. Ἰσδιγούσνας / Ἰεσδεγουνσάφ). 27 Menander the Guardsman fr. 6, 1, l. 207–208 (Blockley 2006, 64). 28 For further study see Widengren 1952, an excellent article about the fall of the Hephthalites and the Turkish–Persian alliance. 22

NÚRIA OLAYA MONTERO although this time the Hephthalite king is called Aḫšunwār29; thirdly, Bal‘amī explains how the Turkish– Persian alliance against the Hephthalites was formed. According to Bal‘amī, the Turks and the Hephthalites had maintained friendly relations until Husrav I asked for help from their Turkish relatives ‒taking advantage of his marriage with the daughter of the Qaγan one year before‒ and both parties agreed to march against the country of the Hephthalites, attacking simultaneously all their borders. Oriental authors state that the war took place in an Eastern city, Balḫ or Buḫara, both in the heart of the Hephthalite territories. All versions agree that the realm of the Hephthalites was divided between the Sasanians and the Turks. The region to the north of the Oxus River was assigned to the Turks while the South bank of the river was granted to the Persians, obtaining both nations a benefit from this conflict. Taxation and trade Taxes could be ‘bidirectional’, i.e. taxes received by the Sasanians due to the previous subjugation of nomadic peoples, which were usually carried to the Persian court by embassies; or, conversely, taxes paid by the Sasanians to other nations because of a Persian defeat. I also distinguish two types of trading: the first one when the Sasanians or the nomadic population were the final buyers of the products and the second one when the Persians or the nomads acted as intermediaries. It should be noted that the lucrative Silk Road crossed the northern regions of Iran and that the Persians asked for taxes so that goods could be carried through their lands. Example One of the best known contacts between Persians and Huns are the two wars of Pērōz I (AD 457–84) with the Hephthalite Huns (AD 474 and 484) and its subsequent consequences: the death of the Sasanian king at the hands of the Hephthalites and the captivity of his son Kavād I (AD 484); the ascent to the throne of Balāš and his later dismissal by an alliance between Kavād I and the Hephthalites (AD 488); Kavād’s flight to the country of his former Hunnic allies due to his deposition by the Persian nobility (AD 496); and his return with an Hephthalite army against king Zāmāsp and his recovery of the power (AD 498). The amount of sources dealing with these issues is very impressive and shows the importance and the memory of these events in later tradition. Many of those sources report a Hunnic tribute levied on the Persians in exchange for peace and for services provided to Kavād I. Procopius of Caesarea tells us that Pērōz (Gr. Περόζης) succumbed together with the whole Persian army and the Sasanians became tributaries of an unnamed Hephthalite king30. He also tells us that some years later, in AD 502, Kavād (Gr. Καβάδης) could not cope with the payment of the tribute and asked the Roman Emperor Anastasius for help to pay the taxes. 29

A title? < Sogd. ’xš’wn /’xšwnδ’r ‘king’ Procopius of Caesarea, De Bello Persico I 4, 35: Dewing 2001, 30. 30

102

Anastasius refused to help him and this caused a war between the Romans and the Persians31. Although Ghirshman32 asserted that some coins from AD 484 to AD 516 had the numismatic legend ΗΑΠΑΤΑΛA and he linked that name with a possible Hephthalite king Heftal III (Yandaiyilituo 厌带夷栗陁), nowadays his assumptions are considered obsolete33. Related to the war and the subsequent Hephthalite rule in Bactria, we have the Bactrian Documents, a group of first–hand letters and lists written in Bactrian and dated following the date of Kaniška, so we have to add 222 years to the date offered in the document34. In the first place, I stress several letters which can certify an effective Hephthalite occupation of the Bactrian region, which bears witness to the victory over the Persians narrated in the sources. In 239 (AD 461) and 252 (AD 474), Meyam, the king of the people of Kadag35, the current Mazār–i Šarif, in the northeast of Afghanistan, was known as the ‘governor of the famous and prosperous king of kings Peroz’36. At the end of the 5th and the beginning of the 6th centuries AD, under the reign of Kavād I, Kilman appeared as the king of the people of Kadag and the ‘governor of the famous and prosperous yab[γu] of Hephthal’37. The title of yabγu38 was associated with nobility, although it is a Turkic loanword. In Two dignitaries appear in two documents of the same period bearing the title of yabγu of the Hephthal: Ormuzd (Bact. ωρομοζδο) and Sart (Bact. σαρτο), who also carried the title of King of Rob39. We have no more data about these dignitaries but it should be noted that the first notable, Ormuzd, had an Iranian name. There are more Bactrian Documents that also attest the payment of taxes or fees to the Hephthalites by private individuals. Two cases from the city of Malr ‒currently Madr in Northern Afghanistan‒, one dating back to 260 (AD 482) and the other one to 295 (AD 517), show several individuals unable to cope with the tax payment established by the Hephthalites; they did not have any asset left in their houses and, therefore, they were forced to sell their only property40. The document dated from 260 specifies that the sellers had to give gold and sheep to the Hephthalites. An undated list enumerates various monthly taxes which a private unnamed individual had to pay to the Hephthalites and only in one occasion to the Persians. His taxes were used to purchase goods or food: five dirhams or drachmas as a subsistence allowance for 31

Procopius of Caesarea, De Bello Persico I 7, 1-3: Dewing 2001, 48, 50. 32 Ghirshman 1948, 17–19. 33 Göbl 1967, II 149. 34 Following the dating system used in Sims-Williams 2007, based on the hypothesis established by de Blois 2006. 35 Bactr. *καδαγο; MP. ktk; Chin. Huolu 活路. 36 Documents ea and ed (Sims–Williams 2007, 108–9 and 114– 5). 37 Document ja (Sims–Williams 2007, 124–5). 38 Bactr. ιαβγo; OT. yabγu; Chin. xihou 翕侯. 39 Documents eh and jb (Sims–Williams 2007, 122–3 and 126– 7). 40 Documents I/Ii and J (Sims–Williams 2012, 44–9).

The rel at i ons of t he S asani an Em pi re wi t h t he Eurasi an S t ep pe the Hephthalites and for the Persians, eight dinars for the ‘Hephthalite levy’ used to buy glazed and silver vessels, but also for the offerings and for sheep fodder. He also paid a monthly tax to an unnamed king of the Hephthalites41. With all these data, it seems that the Hephthalites received a tax not only from the Persians, but also from the subjugated population. Cultural Contacts There were many cultural relations between Sasanian Persia and the nomadic peoples: iconography, religion, loanwords, clothing, exchange of people and ideas, etc. Activities like commerce encouraged a dialogue between different cultures, not only between the Persians and the other Northeastern Iranian inhabitants or the nomadic population, but also with far away countries like China. Example One silver plate in the State Hermitage Museum (Inventory Number S24), dated from the 3rd or 4th century AD and produced in a local workshop in the East of the Sasanian Empire, shows a Sasanian king (Vahrām II or IV? Šābuhr II?) or a Kušāno–Sasanian King (Vahrām Kūšān Šāh?) in a typical hunting scene 42. I want to emphasize the shape of the crown: with ram’s horns on top. This is not a royal Persian model. The ram’s horns

crown is characteristic of the Kušāno–Sasanian Kings and the Kidarite Huns in Bactria and Gandhara, as we can see displayed in their coinage, e.g. the Kidarite coins of Vahrām II Kūšān Šāh in Balḫ or Pērōz Kūšān Šāh in Pešāwar. However, Ammianus Marcellinus narrates that Šābuhr II (Lat. Sapores) exchanged his diadem for a headgear in the shape of a ram’s head during the siege of Amida (AD 359), coincidentally short after he had defeated his Eastern enemies, likely the Kidarites 43. Furthermore, this plate has a Sogdian inscription written in archaic script in the back indicating that this object belonged to a Sogdian merchant. It was found next to a Bactrian bowl in Kerčevo (Perm region, Russia). Conclusions The contacts between Sasanian Iran and the Central Asian steppes can be classified in different categories as I wanted to point out and all of them were of great importance. However, the study of these relations cannot be accomplished without a detailed study of the sources using a combination of various disciplines such as philology, history, prosopography, archaeology, numismatics, epigraphy, etc. This helps us to enhance, extend or correct the data provided by the sources. In this sense, I wanted to show examples representing not only the different types of contacts, but giving also the opportunity to work in a multidisciplinary way.

41

Document al (Sims–Williams 2012, 164–5). Brunner 1978, 92; Herles 2012, 233–5; Lerner 2009, 221–2; Tanabe 1989, 53–5. 42

103

43

Ammianus Marcellinus XIX 1,3.

NÚRIA OLAYA MONTERO Bibliography BECKWITH, C. I. 2009, Empires of the Silk Road. A History of Central Eurasia from the Bronze Age to the Present, Princeton/Oxford.

HERLES, M. 2012, “Zur Darstellung der Wildschweinjagd im Alten Orient”, Stories of long ago. Festschrift für Michael D. Roaf, Münster, 219–40.

BIVAR, A. D. H. 2006 [1983], “The History of Eastern Iran”, in E. YARSHATER (ed.), The Cambridge History of Iran 3(1), Cambridge 181–231.

HUYSE, PH. 1999, Die dreisprachige Inschrift Šābuhrs I. an der Ka‘ba–i Zardušt (ŠKZ). Band 1. Corpus Inscriptionum Iranicarum. Part III: Pahlavi Inscriptions. Vol. I: Royal Inscriptions, with their Parthian and Greek versions. Texts I, London.

BLOCKLEY, R. C. 2006 [1985], The History of Menander the Guardsman: Introductory Essay, Text, Translation and Historiographical Notes, Liverpool. BRUNNER, CH. J. 1987, Sasanian Stamp Seals in the Metropolitan Museum of Art, New York. BOSWORTH, C. E. 1999, The history of al–Ṭabarī. Vol. V: The Sāsānids, the Byzantines, the Lakhmids, and Yemen, SUNY Series in Near Eastern Studies, Albany.

LERNER, J. A. 2009, “Animal Headdresses on the Sealings of the Bactrian Documents”, Iranica 17. 215–26. LUKONIN, V. G. 2007 [1983], “Political, Social and Administrative Institutions: Taxes and Trade”, in E. YARSHATER (ed.), The Cambridge History of Iran 3(2). The Seleucid, Parthian and Sasanian Periods, 681–746, Cambridge.

CHEGINI, N. N.; NIKITIN, A. V. 1996, “Sasanian Iran – Economy, Society, Arts and Crafts”, In B. A. LITVINSKY (ed.), History of Civilizations of Central Asia. The Crossroads of Civilizations: 250–750, Vol. III, Paris, 39– 80.

MOHL, M. J. 1976 [1866], Šāhnāme, vol. V, Paris

DANI, A. H. 1996, “Eastern Kushans, Kidarites in Gandhara and Kashmir, and Later Hephthalites. Part I. Eastern Kushans and Kidarites in Gandhara and Kashmir”, In B. A. LITVINSKY (ed.), History of Civilizations of Central Asia. The Crossroads of Civilizations: 250–750, Vol. III, Paris, 167–180.

SINOR, D. 1994, “The Hun period”, in D. SINOR (ed.), The Cambridge History of Early Inner Asia, Cambridge, 177– 205.

DE BOOR, C.; WIRTH, P. 1972, Theophylacti Simocattae. Historiae. Bibliotheca Scriptorum Graecorum et Romanorum Teubneriana. Stuttgart. DE BLOIS, M. F. 2006, “Du nouveau sur la chronologie bactrienne post-hellénistique: l'ère de 223-224 ap. J.-C.”, Comptes rendus des séances de l'Académie des Inscriptions et Belles-Lettres 150/2, 991-7. DEWING, H. B. 2001 [1914], Procopius. History of the Wars. Books I-II, Cambridge, London.

NARAIN, A. K. 1994, “Indo–Europeans in Inner Asia”, in D. SINOR (ed.), The Cambridge History of Early Inner Asia, Cambridge, 151–76.

SIMS–WILLIAMS, N. 2007, Bactrian Documents from Northern Afghanistan II: Letters and Buddhist Texts. Studies in the Khalili Collection III. Corpus Inscriptionum Iranicarum Part II, London. –2012, Bactrian Documents from Northern Afghanistan I: Legal and Economic Documents (Revised edition). Studies in the Khalili Collection III. Corpus Inscriptionum Iranicarum Part II, London. TANABE, K. 1989, “A Discussion of one Kushano– Sasanian Silver Plate and Relation to Gandharan Art”, Orient 25, 51–80. TEKIN, T. 1968, A Grammar of Orkhon Turkic, Indiana.

FRYE, R. N. 1984, The History of Ancient Iran, München. GHIRSHMAN, R. 1948, Les Chionites–Hephtalites, El Cairo. GÖLB, R. 1967, Dokumente zur Geschichte der iranischen Hunnen in Baktrien und Indien I/II. Wiesbaden. HARMATTA, J. 1965, “Minor Bactrian Inscriptions”, Acta Antiqua Academiae Scientiarum Hungaricae 13/1–2, 149–205.

104

WIDENGREN, G. 1952, “Xosrau Anôshurvân, les Hephthalites et les peuples turcs”, Orientalia Suecana 1, 69–94. WIDENGREN, G. 1971, “The Establishment of the Sasanian Dynasty in the Light of New Evidence”, La Persia nel Medioevo, Roma, 711–82. ZÜRCHER, E. 1968, “The Yüeh–chih and Kaniṣka in the Chinese Sources”, in A. L. BASHAM (ed.), Papers on the Date of Kaniṣka, submitted to the Conference on the Date of Kaniṣka, London, 20–22 April, 1960, Leiden, 358–90.

From the conquest of al–Andalus to the Muslim majority Catalonia between the eighth and the tenth century RAMON MARTÍ Universitat Autònoma de Barcelona & MERCÈ VILADRICH Universitat de Barcelona ABSTRACT The aim of this paper is to summarize the earliest stages of this process, focusing on the actions of the Caliphate of Cordoba and the lands of the Catalan counts during the tenth century, and to assess the evidence that can help to understand how these two societies were founded. Although much progress has been made by recent examinations of Latin and Arabic sources and archaeological analysis, the subject remains controversial, distorted sometimes by the inaccurate claims of medievalists. Indeed, the issue remains a fertile ground for historical research. KEYWORDS Late Antiquity – Dark Ages – Catalonia – Al-Andalus – Islamization – Archeology – Palatium Created by the fusion of the counties under the dominion of house of Barcelona and the conquest of the Muslim regions of Lleida and Tortosa, the birth of Catalonia in the mid–twelfth century represented the establishment of a new political entity in an area characterized by enormous physical and social differences. Obviously, the methods used in the peaceful, gradual integration of the counties and the military conquest of the Islamic lands have nothing in common, and reflect the fact that the two areas were separated by huge gap which pitted the two rival cultures, Christian and Muslim, against each other. The creation of a deeply Arabized and Islamized society in northern al–Andalus had been the result of a long and multifactorial process; profoundly influenced by the Umayyad conquest and the Carolingian response during the eighth century, Catalonia was the scene of a confrontation between two worlds. 1. The formation of al–Andalus, a history under construction Perhaps inevitably, the hypotheses regarding the study of the Islamic period in Catalonia should be understood inside the framework of Spanish historiography. Deeply rooted in a religiously–motivated idea of the conflict, present–day interpretations are still influenced by Claudio

Sánchez Albornoz’s impassioned defence of a view in which the indigenous society was preserved from any Islamic contamination. According to this maximalist approach, the Muslim conquest was followed by large– scale migrations, the abandonment of lands, the search for mountain refuges and the development of new settlement strategies to preserve indigenous cultural wealth. Disregarding the impact of the Umayyad seizure of power, this rather naive medievalism has preferred to focus on repopulation processes. For instance, Ramon d’Abadal wrote influential studies of the resettlement of many of the Catalan central lands annexed by Wilfred the Hairy at the end of the ninth century (Abadal 1958). In fact, Abadal wavered between defending a major role for the resettlement of the lands of Catalunya Vella1, including Septimania, to a more nuanced approach that substantially reduced its importance in the areas under Islamic rule until the eighth century (Abadal 1986). Resettlement was repeatedly used by historians as a 1

Catalunya Vella refers to the area to the north–east of modern– day Catalonia, occupied for a relatively short period by the Muslims. It was predominantly Carolingian and is considered the starting–point of the Christian expansion towards the southern area, known as Catalunya Nova.

RAMON MARTÍ & MERCÈ VILADRICH convenient paradigm for explaining complex political and military events (Sabaté 1996) or the economic growth of the Catalan counties (Salrach 2004). Américo Castro took the opposite view, underlining the ethnic and cultural mixture in the shaping of Andalusi society, a hypothesis based nevertheless on very limited heuristic resources. Thanks to the use of anthropological and ethnographic methods, this approach gradually acquired scientific support, especially after Pierre Guichard proposed his theory of a social and cultural absorption of most of the indigenous peoples by the foreign minority (Guichard 1976). Meanwhile, the French cultural institution Casa de Velázquez organized pioneering archaeological work and Pedro Chalmeta underlined the role of peaceful submission agreements during the conquest of al–Andalus (Chalmeta 1975). From a wider perspective, Richard W. Bulliet noted that it was during the Caliphate of Cordoba that the Muslim majority was attained (Bulliet 1979). Thomas F. Glick’s work broadened the vision of changes introduced in al– Andalus though cultural, scientific and technological transfer from the East (Glick 1992). All these studies opened up a new window on Andalusi history, although the wide range of the views expressed often led to heated epistemological discussions. Considered as a cultural crossroads, the Iberian Peninsula allowed a multidisciplinary comparison of historical, anthropological and philological approaches. Still, at this initial stage the main priority was to bring to medievalism an understanding of Islamic culture, which was virtually unknown at the time. Pere Balañà did a great deal to introduce a Catalan readership to the Islamic heritage (Balañà 1997) and Dolors Bramon published a Catalan translation of Arabic sources on Catalonia until the end of the Caliphate of Cordoba (Bramon 2000). Likewise, the introductory works by the late Miquel Barceló on Islam in the Catalan territories had a noticeable impact. His anthropological approach generated new theories of radical cultural change at an early stage of the conquest, and his views of social and economic development highlighted the link between the process of acculturation and the use of irrigation by peasant communities (Barceló 1992). By questioning the legitimacy of the Arabists and conventional archaeological methods, he prescribed as an alternative what was termed “hydraulic archaeology” (Kirchner & Navarro 1993). However, this approach often provided rather superficial descriptions and presented serious limitations as regards its retrospective application. In recent years, significant progress has been made by bringing together historical documentary and archaeological sources. This collective task cannot be described here in detail, but a few examples will suffice to reflect its importance. Regional archaeological research has flourished in a new multidisciplinary atmosphere, acknowledging the slow but inexorable progress of cultural change in the south–east of the peninsula (Gutiérrez Lloret 2013). Manuel Acién demonstrated how

106

the Andalusi ideology did not manage to generate its own synthesis until Abbasid times (Acién 1998). In Catalonia, as in other northern Iberian areas initially included in the settlement paradigm, the launch of medieval archaeology has faced different challenges in Christian and Islamic areas. First, Andalusi archaeology is a minority discipline which has encountered serious difficulties, especially the research carried out outside urban areas and focusing on the earlier chronological periods. Second, the most recent discussion of Islamic rule in Catalonia dates from fifteen years ago (Islam 1998). At that time, the main results were still limited to the cities of Catalunya Nova – Balaguer, Lleida and Tortosa – for which very few data were available prior to the tenth century (García, Giralt, Loriente & Martínez 1998). Further research was required to bridge the gap between the Umayyad conquest and this date. Our research team’s strategy was to explore documents and place–names together, in order to identify archaeological sites of interest. Unexpectedly, the results showed that the main centres of the Arab conquest coincided with generic Latin place names, such as the term palatium and its derivatives, which we interpreted as properties under Islamic taxation (Martí 1999). We also studied the group of terms derived from pharus, related to the first network of watchtowers built in al–Andalus (Martí 2004). At the same time, we expanded our knowledge of early mediaeval farming in Catalunya Vella (Folch, Gibert & Martí 2010) and tested our hypotheses in the broader scope of four “Villa” seminars studying the area north of the River Ebro (Sénac 2006; 2007; 2010; 2012). Understanding the major political, social and economic transformations in the European West in the eighth and ninth centuries remains a great challenge due to the precarious nature of the sources and the inaccuracies in the interpretation of the archaeological record. However, recent efforts to understand the shaping of al–Andalus society have resulted in an avalanche of data and ideas. A series of collective studies commemorating the thirteen hundredth anniversary of the invasion of 711 have presented a range of analytical perspectives on miscellaneous subjects (Varia 2012; 2013). In the next section we summarize the contribution of studies of Catalonia to this productive debate. 2. The solid foundations of a new domain To be able to explore the social transformations brought about by the conquest in the Catalan lands, first we need to define the late Visigoth Hispanic society and the problems that it faced. Pedro Chalmeta noted that Latin chroniclers attributed the Arab conquest to divine providence, regarding it as punishment for the grave sins committed by the local population (Chalmeta 1994). It is tempting, and appropriate, to review these transgressions of the divine law, unless we decide to ignore contemporary Christian opinions.

F r o m t h e c o n q u e s t o f a l – A n d a l u s t o t he m u s l i m m aj o r i t y Interestingly, the archaeological evidence in the farms of Late Antiquity suggests an extreme degree of social division. Recent excavations have led to some striking conclusions regarding Visigothic social organization. For instance, human corpses have been found thrown among dead animals and waste areas, instead of being buried in cemeteries like the other farm inhabitants (Roig Coll 2011). The palaeo–anthropological studies of individuals abandoned in silos, wells or ditches reveal a shorter life expectancy than for people buried in tombs, and a higher prevalence of nutritional deficiencies and pathological diseases (Ruíz, Villa & Subirà 2007). This marginal group also showed symptoms of demanding physical work, in general absent in the other category. In a single settlement, this oppressed group could comprise dozens of individuals of both sexes and of all ages, suggesting the existence of two social classes which were strictly endogamic and isolated from each other. Scores of archaeological sites from the fifth to the eighth centuries reveal segregated social environments of this kind in the region of Barcelona. They are especially numerous in the Vallès area, where they first attracted the attention of historians, though complementary archaeological evidence has been found in Tarragona, Lleida, Girona and outside Catalonia. Although not all these gruesome discoveries reflect identical circumstances, there is no doubt that certain forms of ancient slavery were widespread in these farmlands. These conclusions recall Marc Bloch’s observations from many years ago regarding the increase in the number of slaves during times of invasion, and also Pierre Bonnassie’s emphasis on the dehumanized condition of the slaves reflected in the written sources until the extinction of the practice in the second half of the first millennium (Bloch 1947; Bonnassie 1985). However, this “slave system” was not based on immutable latifundiae, but on legislation dating from the Late Roman Empire, in which the farmer was tied to the land and the artisan to his labour (Dossier 1989). Indeed, legal sources from the Late Imperial period refer profusely to serfdom, distinguishing between the status of the serfs according to their origins (servers, mancipia, coloni) and showing even freedmen (liberti) and their progeny remained in thrall to the master (Barbero & Vigil 1978). The Liber iudicum and the ecclesiastical canons imposed the obligation of menial labour on public (servi fiscales), private, or ecclesiastical serfs. Of course, the serfs might escape, and the laws included a complete book on fugitives and the liability of those who concealed them (IX De fugitivis et refugientibus). These legislative provisions should not be regarded as mere rhetoric, although there must have been significant geographical differences in terms of their application. Moreover, peasant farmers’ rights to the land would have been an important issue. Probably, the initial bond of dependence involved the peasant being tied to his master’s land in accordance with the old ordinances, although with time there seems to have been a trend

107

towards more attenuated forms of servitude such as tenancies. However, tenant farmer status was no panacea: whether tenants were subject to public authorities or to private landowners was of little consequence, since the ownership of land and of individuals had become the very essence of power at that time. It is clearer now that the Middle Ages did not start with large slave–owning farms, but we can also confirm that certain groups lived in conditions of extreme subordination and were entirely deprived of freedom. This is demonstrated by the persistence of domestic hereditary slavery in the seventh and eighth centuries, when the western lands were thrown into political and social turmoil in the wake of the Islamic expansion. We prefer to attribute the extinction of slavery to a social conquest, rather than to recruitment difficulties, the role of the Church, or technical progress, even though it did not imply the total suppression of servitude (Martí 1999b). Far from ending, the widespread flight of serfs actually increased towards the end of the Visigothic period, when King Egica enacted “a truly bizarre law, mobilizing the entire free population of Hispania as a patrol force for the hunting of slaves” in 702 and where any poorly–dressed stranger immediately aroused suspicion (Bonnassie 1985, 337). So the Islamic expansion found the Iberian Peninsula in a state of extreme social division and serious political conflicts, while Roderic’s coup d’état in Toledo challenged the legitimate election of King Akhila in Narbonensis and Tarraconensis. Traditionally regarded as weak, early Andalusi rule in fact overhauled public property and taxation in Catalonia and Septimania and established a strong new system of territorial defence. However, in Catalonia, as in other Iberian lands, the territorial rearrangement was not uniform, and two core areas emerged as a result of the process of conquest and the resistance offered in the field. Ṭārīq and Mūsā’s hasty initial invasion up to the year 713 seems to have subdued most of Catalunya Nova, but did not secure the Caliphate’s fifth of the spoils (khums). In contrast, the successful extension of the campaign to Narbonne, led by al–Ḥurr and particularly by al–Samḥ (719–721), was followed by the distribution of undivided lands and the delivery of the share corresponding to the state, in accordance with cUmar ibn cAbd al–cAzīz’s instructions (Gibert 2011, 248–253; Martí 2012). In practical terms, this meant the assignation of more than 200 farms in Catalunya Vella and Narbonensis to the Islamic state, the so–called palatium, or palau in later Latin sources, a term related at that time not to a building, but to public property or land taxation (fiscus; Fig. 1). A recent publication provides new data on the subject, placing it inside the broader framework of the first Islamic taxation programme (Pastor & Ballestín 2013). Overall, rural palatia perform as homogeneous logistical centres, distributed throughout the area and especially around cities and towns. Their names can prove revealing. For instance, fiscal activity is reflected in the name “Palau de Soldans / Sultan’s Palatium” (P. Salatane), and

RAMON MARTÍ & MERCÈ VILADRICH in Latin forms such as “Palau del Tresor/Treasure Palatium” (P. Fisco) and “Palau del Rei/King’s Palatium” (P. de Reig). Sometimes, the names tell of the ethnicity of their occupants, as in four unmistakable allusions to Palacio Moro, P. Maurorum or Palamors, although it is unclear whether the reference is to Berbers or whether maurus might also refer to Arabs. Nevertheless, the owners of certain palatia were Arabs, for instance Abū Thawr (P. Abtauri) in Jacfar/Jafre (Baix Empordà), the probable Abū Zayd (P. Auzido) in Ripollet (Vallès Occidental) and a possible Abū cĪsà (P. de Avezia) in Rocafort de Bages (Bages). The name P. Monnoso in Roussillon may refer to the Berber Munūsa, who revolted in Cerdanya in 731. Other names hint at Arabic terms, like P. Meserata or P. Moronta, or at least a possible Eastern origin for unusual local names like P. Dalmalla, P. Dalba, P. Aries, P. Danum, P. Rafano and P. Vitamenia. Some indigenous names also appear among palatia owners, such as P. Rodegarii, P. Frugelli, P. Borrelli and P. Felmiro, suggesting their active role as muladi clients or, more likely, as owners at a later date under Carolingian rule. Other names allude to local characteristics, such as P. Siccus, Monte Palacio and P. Serpentis, or to the settlement’s age, such as P. Antiquum, P. Vetere and P. Fracto. Still others seem to be linked to a previous name, as is probably the case of P. Merlago, although most palatia are referred to by a simple name without any additional details in plural or diminutive forms (Palol, Palou, Palaus, Palous), or by a specific geographical location (P. Tordaria, for example). They all refer to small or slightly larger rural farms, usually considered as villae or loci, and reinforced by a tower, as in Palau–sator (Alt Empordà) and in Turrem de Azar que vocant Palaciolo (Vallès Oriental).

state can be identified in old names deriving from al– qaṣr, while the Arabic term balāṭ seems to correspond to the Latin palatium (Marti 2013). However, there were huge differences in the strategies of distribution: while there were two hundred palatia in the northern and eastern districts of Catalonia, there were only four or five rural qūṣūr in Catalunya Nova. From the archaeological perspective, the data available today regarding these rural palatia may appear ambiguous, for different reasons: either because they come from areas which have been heavily exploited and occupied or because the archaeological research carried out was not systematic and did not provide reliable dates. The sites of greatest archaeological interest tend to be located around settlements from the Roman period or Late Antiquity both on the coastline or inland: for instance, in the old factory at Pla de Palol (Platja d’Aro, Baix Empordà), where no medieval objects were found (Folch 2013), in Terrassa (Vallès Occidental), and also in P. Fracto, in l’Aiguacuït (Soler, Ruíz 2013), an old settlement with huts and silos from Visigothic times containing abandoned human corpses, and where a larkspur was recovered (Roig, Coll 2013). Another emblematic site is Palous de Camarasa, which attesting settlements from ancient, Andalusi and later periods; with a Visigothic necropolis and an early eighth century grave in which period tools like a manure scraper and a lily– shaped buckle were found, commonly attributed only to the Visigoth period (Alòs, Solanes 2013). In Catalonia, in the absence of conclusive dates, it is debatable whether certain debris and structures (for example, several palatia) correspond to the end of the Visigothic era or the beginning of the Islamic period (the latter option was not considered by the first scholars). In Septimania, the finding of Islamic graves on ancient mosaics in Les Palats, in Laure–Minervois (Aude), certifies the presence, albeit brief, of the newcomers in northern locations (Gibert 2013). Among the first products of the Islamic conquest is the long network of watchtowers connecting eastern Catalonia, flanking the Via Augusta from the Pyrenees to the southern counties of Barcelona. It consists of a series of similar circular towers, referred to as “lighthouses” (pharus) in documents, built in opus quadratum cut stone (Figures 2 to 4). Always considered to be Roman constructions, this attribution has recently been challenged by the results of archaeological studies (Martí ed. 2008; Martí, Folch & Gibert 2008).

Fig. 1: Location of Palatia in Catalonia.

The areas of Lleida and Tortosa were the first to be conquered and must have been subject to land division and distribution. A part would have been reserved for the state and the maintenance of the Muslim community. Here, as in areas of al–Andalus occupied for longer periods by the Arabs, the rural patrimony of the Islamic

108

Highly symbolic and strategic, the network is an extraordinary achievement which must have been carried out by a state power, but it remains difficult to date on the strength of the archaeological evidence. Onomastic evidence provided in the Turó de Montcada may be able to resolve the question, for it appears to have been the location of a huge watchtower along the route from Vallès to Montjuïc and Barcelona. The name Montcada (Monte Catano) is believed to correspond to the designer of the entire network, cAbd al–Malik ibn Qaṭan al–Fihrī, an

F r o m t h e c o n q u e s t o f a l – A n d a l u s t o t he m u s l i m m aj o r i t y aristocrat appointed by the Emir of Ifrīqiya, and who is criticized in the sources and seems to have been obsessed with mastering peaks and mountains. Sent to al–Andalus circa 732 and deposed in 734, he returned around 741, only to be executed the following year later in the aftermath of the great Berber revolt which shook the entire West. Although place–names with faro and its derivatives are common all over the Iberian Peninsula, there was no other network carrying out similar functions in al–Andalus, a fact that stresses the prominent role of Catalonia and Septimania in the context of the early Muslim expansion.

obvious innovations of Eastern or North African inspiration.

Fig. 4: Torre de les Gunyoles, possibly a watchtower (Avinyonet del Penedès).

This apparent lack of cultural and chronological definition is not restricted to the Catalan area, and it even raises doubts about other exceptional examples in the Emiral archaeological record which display unmistakable Umayyad influences such as architectural medallions. Among Mozarabic churches, the rural monastery of Santa Maria de Melque, near Toledo, built in a balāṭ (Balatalmelc) towards the middle of the eighth century, is an archaeologically certified example (Caballero, Moreno 2013). Likewise, although considered as pre–Islamic, the great Valencian aristocratic residence of Plà de Nadal (Riba–roja de Túria) was built by a certain Theudimer, namesake of the count who surrendered the lands of Alacant and Murcia to the Arab conquerors (though probably unrelated, given the distance between their domains) (Gutiérrez Lloret 2013). According to another plausible explanation, Theudimer, whose influence and personal privileges would have increased after his submission, may have being rewarded with this unusual Oriental–style building near the fortified town of Valencia la Vella. Like him, some Visigothic aristocrats would benefit substantially from the conquest of al–Andalus, as did King Vitiza’s heirs, whose ownership of three thousand Lusitanian farms was confirmed (ḍiyac). This issue merits further research, for some of the indigenous aristocracy shared territorial dominion with the Arabs and maintained the strong oppression of the peasantry for decades to come.

Fig. 2: Locations of watchtowers and similar structures.

Fig. 3: Torrassa “del Moro” or “del Far” (Llinars del Vallès).

We can continue to argue whether the foundation of the network was Roman or Islamic, although the gap between them is more than nine centuries (!). However, we stress that some of the excavations show no evidence of ancient occupation, and the route of the network is quite unrelated to the Roman conquest of Hispania. The ambiguities arise from the similarities in the materials and the structures which recall preceding traditions and the absence of any

109

The survival of the previous social and economic order is also observed in the form of the fortified settlements that expressed territorial dominance, but these settlements were the first to be affected by political and military contingencies and changes. Fortified towns like Serrat dels Tres Hereus (Casserres, Berguedà), which dates from the eighth century, appears to have carried out specialized economic functions in the surrounding rural areas (Folch, Gibert, Martí, Rodrigo 2007). The excavation of its ruined early mediaeval levels yielded a repertoire of mainly iron tools related to livestock farming, such as wool carders, a

RAMON MARTÍ & MERCÈ VILADRICH skiver and a scraper among other implements, arrowheads and nails, but no agricultural devices (Morell 2006). Without any reliable data on the historical context it is difficult to determine the function of this fortified town, although it was probably a centre of metal production and livestock farming. Elsewhere, the fortified Visigoth town of Puig Rom (Alt Empordà) which dominated the Bay of Roses from the seventh century onwards and was evacuated during the Islamic conquest is a clear forerunner of settlements of this kind. Here, the variety of abandoned tools and weapons reflect the inhabitants’ farming and fishing activities (Palol 2004). At a different scale, the fortifications of l’Esquerda (Roda de Ter, Osona) and Olèrdola (Alt Penedès), described as “cities” in later documents, provide additional examples of these early mediaeval colonies. Because of their particular geographical position, these fortified nuclei complemented or replaced ancient settlements which had been important centres inside their area of influence and often possessed a cathedral or a monastery. The career of the Church in the eighth century is somewhat contradictory, similar in a way to that of the collaborationist Visigothic aristocracy. Moving between the maintenance of tradition and the enlargement of its social base, all Episcopal dioceses in Catalonia were guaranteed continuity, while the alleged flight of Archbishop Prosperus of Tarragona may be an invention of recent historiography (Gonzalo 2011). Also, towards the end of the century, submission to Toledan hierarchy and precepts is still evident, as witnessed by the quarrel over dogma involving Bishop Felix of Urgell, a victim of Carolingian expansion (Martí, Viladrich 2000). Islamic rule may even have contributed to the spread and popularization of Christianity by imposing protected legal status (dhimma) on the natives and obliging them to pay tax (jizya), policies which in the long term may have bound the natives closer together. In the ninth century, the collaborationist Church hierarchy ended in collapse in most of the provinces, while Islamization and Arabization made substantial progress (Aillet 2013). However, the initial hegemony of Arabs and their allies was soon interrupted by regional or local Berber revolts in the Maghrib and al–Andalus, and ended with the overthrow of the Umayyad by the Abbasid revolution. The arrival in al–Andalus of the emir cAbd al–Raḥmān, in 754, opened up a new period in which power gravitated definitively around Cordoba; the strategic interest of the northern limits began to wane, and Narbonne was irreversibly lost in 759. 3. Forging a new frontier society During the last quarter of the eighth century the Carolingian response provides the historical context of a society in which the Arab lineages struggled to maintain their old privileges dating back to the time of the conquest. Several governors of the cities switched

110

allegiance from the Emir to the Franks, a dramatic move in itself and one that was capitalized upon by others ready to replace them even if it meant continuing their submission to Cordoba’s dawla (Ballestín 1999). Abū Thawr, governor of Girona (and not of Huesca, as was traditionally believed) is a sound example of this trend. He possessed the palatium of Jacfar/Jafre (Empordà) as well as a villa on the outskirts of the city in Campdorà, which both appear recorded under his name. In 778, along with other Arab leaders Abū Thawr supported Charlemagne’s expedition against Zaragoza, while circa 790 he sued for peace with Charlemagne when it appears that his domains were transferred to the enemy. The action confirms the detail in the Chronicle of Moissac which states that it was the city’s inhabitants who surrendered the city in 785, though this date may not be accurate (Martí 2013). Like Abū Thawr, many Muslims must have remained in the counties of Girona and the Empordà; this is demonstrated by the presence of Arabic names still in use in the early tenth century. A similar process was underway in Barcelona, where Sulaymān ibn al–Acrābī participated in the Franks’ expedition in 778, when he sought to extend his rule to Zaragoza. In the end, the adventure cost him his life. The same fate befell his son Maṭrūḥ a decade later; as governor of Barcelona, Maṭrūḥ tried to capitalize on opposition to the new emir Hishām on the frontier. Once Girona and the Pyrenean areas had defected, Cordoba’s replacement of the governors of Barcelona was to no avail, and the city surrendered to a major Frankish offensive in 801. As in Girona, some Muslims preferred to remain under Carolingian rule and some even volunteered for the Frankish army, as can be inferred from a fatwa (Sénac 2002, pp. 59–69, note 36). A significant number of Muslims retreated to the rearguard or to positions of greater safety; the evacuation of Tarragona may be a case in point. It is commonly accepted that the Frankish offensive ended at the Llobregat and Cardener rivers, though campaigns were waged against the city of Tortosa and the valley of the Ebro. However the establishment of a new imperial order prioritizing the Frankish component encountered serious problems in Catalunya Vella, where, throughout the ninth century, disputes between leading aristocratic clans proliferated under the Carolingians. An example is the joint revolt of cAyšhūn and Guillemó d’Osona, dispossessed members of the lineages of Sulaymān and Berà, count of Barcelona in 826; as a result of the uprising the central regions of Catalonia remained a no man’s land for half a century. With no intervention from either the Carolingians or from Cordoba, the independence of these districts evidenced the contradictions of the Frankish model, as well as the social cohesion and vitality of this rural area particularly when Arabic and local forces worked together (Gibert 2011). During this period of political and military impasse of the ninth century, there were profound changes in the social

F r o m t h e c o n q u e s t o f a l – A n d a l u s t o t he m u s l i m m aj o r i t y organization and in the rural habitat. In the Carolingian area, the city of Barcelona and the castle of Terrassa became the main fortifications in the defence of the frontier, as witnessed by a privilege granted by Charles the Bald to the county’s gotos sive hispanos in 844. In being exempted from certain taxes, the Catalan Christians were granted many of the same rights and duties as the Franks. Like them, they had to serve in the army and in the active defence of the frontier, performing surveillance or supplying horses for royal emissaries. Rules concerning the possession and ownership of land were part of the specific hispani tradition that had developed since the conquest of Septimania and had encouraged defections to Andalusi frontier districts. In fact, the new aprisio regime upheld the rights of the first occupier or the count, who was thus able to benefit from the services of the newcomers (Dupont 1965). At this point the freedom of choice of the newcomers seems to have ended, under the tutelage of their chosen lord, who provided them with lands to settle. In the Islamic lands, the persistence of certain features of the Catalan palatia that have been identified by archaeological study, confirms a certain permanence in the social order. Sooner or later, however, these farms abandoned ancient traditions and introduced new practices. Within a few generations old social and economic customs had died out. Palatia are by no means an isolated case, since numerous sites in Catalonia during Late Antiquity remained operational during the Islamic period, although they tended to decrease in size as previous social inequalities faded away. A few datable examples discovered over the last decade should suffice to show this. For instance, the site of Aubert (Vall d’en Bas, Garrotxa) consists of a main dwelling divided into several areas which were used between the sixth and the eighth centuries, and was replaced by new structures in the Carolingian period. It was abandoned definitively in the tenth century (Folch 2012). Likewise, in Monistrol de Gaià (Gaià, Bages), which is still being excavated, a large walled area dating from the sixth century was partially reused in Islamic times and became a forge under the Carolingians (Gibert 2011). Lastly, the briefly inhabited tower of Santa Coloma (La Selva) endorses the Islamic origin of the network of watchtowers (Folch, Gibert, Llinàs, Martí 2007), while the necropolis of Pertegàs (Calders, Bages) is a fine example of the survival of the indigenous settlement at least during the eighth century (Folch, Gibert, Carrascal 2011). However, the main indicators of religious change – Muslim graves, for example – are few and far between and are limited to a few sites and ceremonies from Late Antiquity that are not attached to churches. We have a couple of isolated examples of Islamic burial sites in Catalunya Vella – Sarrià de Ter (Gironès; Folch 2012, pp. 437–438) and Altimiris (Sant Esteve de la Sarga, Pallars; Sancho 2010) – which were common throughout al– Andalus (Gutiérrez Lloret 2012), and also found in Catalunya Nova in Riudoms and Vilanova de Prades (Baix Camp, Conca de Barberà, Gonzalo 2011; Menchón

111

2012) and in Benifallet (Baix Ebre, Negre, 2013, p. 489). Significantly, a recent selection of studies on burial practices in Catalonia did not include Islamic sites, and burials in lateral decubitus are scarce, even in the counties to the west of the Llobregat (Molist & Ripoll 2012). With no thorough–going data on Islamic conversion, we can only state that it is during the tenth century that the rural cemeteries near Balaguer, Lleida and Tortosa display obvious effects of a general process of Islamization. However, in the ninth century ribats were built which seem to have played an important part in in the process of Islamization and territorial defence. In coastal areas and areas around cities we find place–names like Vallfogona de Balaguer (Noguera) and Sant Carles de la Ràpita (Montsià), where towers are recorded, and Santa Margarida i els Monjos (Alt Penedès) and La Riba (Alt Camp). The areas of central Catalonia strengthened the watchtowers and castles to preserve regional and local positions. In the Andalusi sector, round towers evolved into stylized shapes, reducing their diameter and increasing their height, combining stone and mortar or building inner false–vault covers, at the same time improving the defences. Before the tenth century, the tower of the castle of Sant Pere de Ribes (Garraf) had a solid basement and a horseshoe–arched door of tripartite lintel (Adell, Riu 1992; Plate 5). Similarly, the magnificent crenellated tower of Vallferosa in Torà (Segarra) dates from the Islamic period (Menchón 2011). Prior to Carolingian rule, the main regional nuclei of central Catalonia – Berga, Cardona, Solsona, Manresa and Olèrdola – would have had Moorish fortifications, although little is known about their ninth–century inhabitants. The Sant Miquel cemetery in the castle of Olèrdola dates from the late eighth century, but no traces of Muslim burials remain (Molist & Bosch 2012). In Lleida we occasionally find unusual rural fortifications, very different from the standard building practices of the counts, with archaic masonry patterns that merit further study (Plate 6). In any case, the political unity of the Franks collapsed towards the end of the ninth century under the revolt led by Wilfred, Count of Urgell, Berga, Osona and Barcelona, and his heirs. Able to rally local leaders, the counts pushed back the frontier to the Conca de Barberà, bringing the lands under their control towards the middle of the tenth century. From this point on, we begin to have documents in the archives reflecting the imposition of their power in each area and setting the definitive boundaries between the different counties. These documents contain only a few Arabic place names in central Catalonia, no more than in Catalunya Vella (Pérez 2008), while in the newly conquered lands the Muslim presence is practically non–existent. This suggests that the Muslim minority living there would have retreated to safer areas, while the advance of the counties of Cerdanya affected mainly the indigenous population, where the new

RAMON MARTÍ & MERCÈ VILADRICH power forcibly submitted the area and insisted on its Christian character.

Fig. 6: Castellnou d’Ossó, emiral building (Ossó de Sió).

Fig. 5: Castell de Ribes’ watchtower (Sant Pere de Ribes)

Wilfred set up new religious and military institutions and built numerous castles and churches. The new fortifications proliferated to the west of the River Llobregat, and by the year 1000 there were more than 130 castles in the counties of Barcelona alone. In the hinterland, the old institutions were less affected by change, although certain monasteries as Ripoll, Sant Joan de les Abadesses and Sant Cugat del Vallès benefited from the patronage of the counts and became powerful institutions able to compete with the authority of the diocesan bishops, building churches and establishing the first rural parishes. These castles and parishes provided the bases for new general taxes, which were no longer linked to the relationships of property and ownership of the land, and which remain almost unexplored today. In documents from the period both castles and parishes claimed the same jurisdictional rights to the biblical decima, revamped into a public tax in Carolingian times. Meanwhile, the situation of the Islamic cities in Catalunya Nova also underwent a radical change as they joined the organization of the Cordoban state. Towards the end of the ninth century the favoured Umayyad client lineage of Banū Qāsī (Lorenzo 2010) built the walls of the city of Lleida and the splendid fortress of Castell Formós de Balaguer, with ashlar masonry (Giralt 1994; García, Giralt, Loriente, Martínez 1998). At the same time, the city of Tortosa demonstrated a high degree of autonomy and naval power, able to intervene throughout the Mediterranean (Ballestín 1999; Guichard 2001).

112

Despite Tortosa’s loyalty to the Caliphate, it was not until the end of the Islamic presence that the military and ideological force of the occupiers began to find expression in the form of public works (Martí, Negre in press; Plate 7). For instance, the shipyard built under cAbd al–Raḥmān III near 944–945 is attested by a foundation stone (Yzquierdo 1998); commonly believed to have been located in the Remolins sector, it was more probably constructed in the south–western point of the city, near the Portal de la Rosa and towards the final section of Barranc del Rastre, as part of a general overhaul of the city walls. Al–Ḥimyārī states that Tortosa’s mosque was built towards 956–957, with five vaults and a spacious courtyard (Bramon 2000, p. 125–127). It may have been located near the cloister of the cathedral or the surrounding area, where certain pieces were reused, like two small porphyry and marble columns similar to the ones in al–Ḥakam II’s miḥrāb in the Cordoba mosque, built shortly afterwards. Also, one of city baths may have been built on the Caliphate’s initiative, if the misnamed miḥrāb of Tarragona, a marble piece with an epigraphic inscription which attributes the work to caliph cAbd al– Raḥmān and his fatā and mawlā Jacfar circa 960–961, turns out to be from Tortosa. The piece is similar to the windows in the baths in the upper terrace in Madīnat al– Zahrā', near Jacfar's house (Aceña 1998). This decisive state intervention transformed the urban structures and the agricultural economy of Catalunya Nova. However, this new order established itself in the tenth century, and before this time the rather simplistic hypothesis that automatically associates Andalusi society with irrigated agriculture should be rejected, in spite of the arrival of plants and agricultural techniques from the East (Negre 2013). In addition to the strong state intervention, each of the rural constituencies in Tortosa was linked to a specific castle. Though fewer in number than the fortifications in the Christian areas, the Caliphate castles of Amposta, Miravet and Siurana de Prades were nonetheless solid strongholds which acted as territorial and political centres in their respective regions.

F r o m t h e c o n q u e s t o f a l – A n d a l u s t o t he m u s l i m m aj o r i t y the Christians. Nonetheless, local dhimmī communities survived as relatively strong groups in border areas such as the foothills of the Pyrenees.

Fig. 7: Castell d'Amposta's Torre de Sant Joan, caliphal building.

The truces signed in 940 between the Caliph of Cordoba and the Count of Barcelona and other leaders in the area seem to mark the turning point in the two divergent processes of rural Islamization and Christianization, until the resumption of hostilities in the last quarter of the century. Political and military leaders on both sides would finance their operations through a religious tax – the zakāt among the Muslims and the ecclesiastical decima among

113

An agreement made in 987 between the villages Aguilaniu and Josseu regarding the exploitation of a salt well, supervised by a priest named Fortuny, who was the judge of the Christians in Lleida (iudicem cunctis Christian Leridensis), involved the payment of the zakāt (post azeka cecidit super illos, Marti 1995, p. 42). An even clearer example is the village of Àger in La Noguera, where a significant Mozarabic community was documented from the ninth century onwards and where the effects of Arabization were noticeable when the valley was conquered by the Christians in the eleventh century (Benet 1983; 1994). In fact the influence of the Islamic conquest was evident throughout the Christian lands of the north of the Peninsula, where a series of cultural, fiscal, technical and scientific innovations from the East were implemented by native societies without any apparent contradiction (Pastor & Ballestín 2013). The resulting societies were quite different from the ones that had preceded them.

RAMON MARTÍ & MERCÈ VILADRICH Bibliography ABADAL, R. 1958, Els primers comtes catalans, Barcelona. – 1986, Catalunya Carolíngia I–1. El domini carolingi a Catalunya, Barcelona. ACEÑA, R. 1998, “Arquet de Tarragona”, in L’islam i Catalunya. Catàleg, Barcelona, 44–46. ACIÉN, M. 1998, “Sobre el papel de la ideología en la caracterización de las formaciones sociales: la formación social islámica”, Hispania: Revista española de historia 58, 915–968. ADELL, J.–A.; RIU, E. 1992, “Castell de Ribes (o de Bell– lloc)”, Catalunya Romànica. El Penedès. L’Anoia 19, 268–270. ALÒS, C.; SOLANES, E. 2013, “El jaciment arqueològic de Palous (Camarasa, la Noguera). Dades sobre el poblament hispano–visigot al mig Segre”, in Lo que vino de Oriente. Horizontes, praxis y dimensión material de los sistemas de dominación fiscal en al–Andalus (siglos VII–IX), Oxford, 211–222. AILLET, C. 2013, “La formación del mozarabismo y la remodelación de la península ibérica (s. VIII–IX)”, in De Mahoma a Carlomagno. Los primeros tiempos (siglos VII–IX), Pamplona, 285–310. BALAÑÀ, P. 1997, L'islam a Catalunya (segles VIII–XII), Barcelona. BALLESTÍN, X. 1999, “L’anomenada frontera superior d’al–Andalus: aṯ–ṯaġr al–a'là i la resistència a la instauració de la dawla dels Banu Marwan”, in M. BARCELÓ (ed.), Musulmans i Catalunya, Barcelona, 61– 76. BARBERO, A.; VIGIL, M. 1978, La formación del feudalismo en la Península Ibérica, Barcelona. BARCELÓ, M. 1992, “Quina arqueologia per al– Andalus?”, in II Coloquio Hispano–Italiano de Arqueología Medieval, Granada, 243–252. BENET, A. 1983, “Una comunitat mossàrab a la Vall d'Àger. Estudi Antroponímic”, Butlletí interior d'Onomástica 13, 3–7. – 1994, “Els mossàrabs a Catalunya. El cas de la vall d’Àger”, Catalunya Romànica. La Noguera 17, 28–31. BLOCH, M. 1947, “Comment et pourquoi finit l'esclavage antique”, Annales. Économies, Sociétés, Civilisations, 2e année–1er, 30–44.

BONNASSIE, P. 1985, “Survie et extintion du régime esclavagiste dans l'Occident du haut moyen âge (ive–xie s.)”, Cahiers de Civilisation Medievale 28, 307–343. BRAMON, D. 2000, De quan érem o no musulmans. Textos del 713 a 1010, Vic–Barcelona. BULLIET, R. W. 1979, Conversion to Islam in the Medieval Period: An essay in Quantitative History, Cambridge. CABALLERO, L.; MORENO, F. J. 2013, “Balatalmelc, Santa María de Melque. Un monasterio del siglo VIII en territorio toledano”, in Lo que vino de Oriente. Horizontes, praxis y dimensión material de los sistemas de dominación fiscal en al–Andalus (siglos VII–IX), Oxford, 182–204. CHALMETA, P. 1975, “Concesiones territoriales en al– Andalus”, Cuadernos de Historia. Anexos de la revista Hispania 6, 1–90. – 1994, Invasión e islamización, Madrid. DOSSIER 1989, “Dossier: Esclaus, serfs, pagesos. De l'esclavisme al feudalisme”, L'Avenç 131, 21–49. DUPONT, A. 1965, “L'aprision et le régime aprisionaire dans le Midi de la France”, Le Moyen Age 71, 179–213, 375–399. FOLCH, C. 2012, Els territoris del Nord–est de Catalunya durant l’Alta Edat Mitjana (segles vi–xi d.C): organització territorial i arqueologia del poblament, [Diss], Bellaterra, Universitat Autònoma de Barcelona: http://www.tdx.cat/ handle/10803/107883 – 2013, “Els palatia dels territoris del nord–est de Catalunya: indicis toponímics de la primera organització fiscal islàmica”, in Lo que vino de Oriente. Horizontes, praxis y dimensión material de los sistemas de dominación fiscal en al–Andalus (siglos vii–ix), Oxford, 149–159. FOLCH, C.; GIBERT, J.; CARRASCAL, S. 2011, “La necròpolis de l’Alta Edat Mitjana de Pertegàs (Calders, Bages) (segles VIII–IX)”, IV Congrés d’Arqueologia Medieval i Moderna a Catalunya, 2010, Tarragona, 113– 122. FOLCH, C.; GIBERT, J.; LLINÀS, J.; MARTÍ, R. 2007, “Excavacions arqueològiques a la torre del Far (Santa Coloma de Farners, La Selva). Una torre alimara a la Marca Superior d’al–Andalus”, in III Congrés d’arqueologia medieval i moderna a Catalunya, Sabadell, 529–534. FOLCH, C.; GIBERT, J.; MARTÍ, R. 2010, “Hábitats y sociedades rurales en Catalunya (siglos vi–xi): toponimia,

114

F r o m t h e c o n q u e s t o f a l – A n d a l u s t o t he m u s l i m m aj o r i t y documentación y arqueología”, in Villa 3: Historia y arqueología de las sociedades del valle del Ebro (ss. vii– xi), Toulouse, 125–155. FOLCH, C.; GIBERT, J.; MARTÍN, J.; RODRIGO, E. 2007, “L’ocupació de l’Alta Edat Mitjana del jaciment del Serrat dels Tres Hereus (Casserres, Berguedà)”, in III Congrés d’Arqueologia Medieval i Moderna a Catalunya, Sabadell, 753–757. GARCÍA, J. E.; GIRALT, J.; LORIENTE, A.; MARTÍNEZ, J. 1998, “La gènesi dels espais urbans andalusins (segles VIII–X). Tortosa, Lleida i Balaguer”, in L’islam i Catalunya, Barcelona, 137–165. GIBERT, J. 2011, L’Alta Edat Mitjana a la Catalunya Central. Estudi històric i arqueològic de la conca mitjana del riu Llobregat, [Diss.], Universitat Autònoma de Barcelona, Bellaterra: http://www.tdx.cat/ handle/10803/79098 – 2013, “De Guissona a Magalona: consideracions a l’entorn del topònim palatium en els extrems de la seva dispersió en terres de la Tarraconesa oriental i la Septimània”, in Lo que vino de Oriente. Horizontes, praxis y dimensión material de los sistemas de dominación fiscal en al–Andalus (siglos vii–ix), Oxford, 160–181. GIRALT, J. 1994, “Castell Formós”, Catalunya Romànica. La Noguera 17, Barcelona, 225–238. GLICK, TH. F. 1992, Tecnología, ciencia y cultura en la España medieval, Madrid. GONZALO, X. 2011, Dinàmiques d’organització social i territorial in territorio Terrachone. Estudi històrico– arqueològic del territori de la ciutat de Tarracona / Ṭarrakūna / Terrachona i els seus marges (s. vii–x dne), Mémoire M2R Études Médiévales, Université Toulouse II Le Mirail, Toulouse: http://www.academia.edu/1471985/ GUICHARD, P. 1976, Al–Andalus : estructura antropológica de una sociedad islámica en Occidente, Barcelona. – 2001, “Les zones litorales de la région orientale d’al– Andalus aux VIIIe–XIe siècles”, Castrum 7. Zones côtieres litorales dans le monde méditerranéen au Moyen Âge: défense, peuplement, mise en valeur, Rome–Madrid, 273–290.

Islam 1998. L'Islam i Catalunya, Museu d’Història de Catalunya. Barcelona. KIRCHNER, H.; NAVARRO, C. 1993, “Objetivos, métodos y prácticas de la Arqueología Hidráulica”, Archeologia Medievale, 20, 121–150. LORENZO, J. 2010, La dawla de los Banū Qasī. Origen, auge y caída de una dinastía muladí en la frontera superior de al–Andalus, Madrid. MARTÍ, R. 1995, “Territoria en transició al Pirineu medieval”, in Actes del 3er curs d'arqueologia d'Andorra, Andorra, 37–83. – 1999a, “Palaus o almúnies fiscals a Catalunya i al– Andalus”, in Les sociétés méridionales à l'âge féodal. Hommage à Pierre Bonnassie, Toulouse, 63–70. – 1999b, “Conquestes i capitulacions pageses”, in Catalunya a l'època carolíngia. Art i cultura abans del romànic, Barcelona, 59–63 [English translation: “Peasant Victories and Defeats”, Id., 448–451]. – 2004, “Fars de terra endins. A propòsit de la Torrassa del Moro”, Laietania, 16, 185–198. – (ed.) 2008, Fars de l’islam. Antigues alimares d’al– Andalus, Barcelona. – 2012,”Los territorios catalanes en la encrucijada del 711”, 711. Arqueología e historia entre dos mundos, Zona Arqueológica, 15.2, 11–23. – 2013, “El palatium rural, una institución fiscal del siglo viii”, in Lo que vino de Oriente. Horizontes, praxis y dimensión material de los sistemas de dominación fiscal en al–Andalus (siglos vii–ix), Oxford, 133–148. MARTÍ, R.; FOLCH, C.; GIBERT, J. 2008, “Fars i torres de guaita a Catalunya: sobre la problemàtica dels orígens”, Arqueologia Medieval. Revista catalana d’arqueologia medieval 3, 30–43. MARTÍ, R.; NEGRE, J. 2014, “Fortificaciones y edilicia de prestigio en el extremo oriental de la Marca Superior: Ṭurṭūša y su entorno”, in 1st International Medieval Meeting Lleida 2011, Lleida [in print]. MARTÍ, R.; VILADRICH, M. M. 2000, “Guissona, origen del bisbat d'Urgell”, El Comtat d'Urgell 4, Lleida, 37–66.

GUTIÉRREZ LLORET, S. 2012, “La necrópolis de Vistalegre (Aspe, Alicante) a la luz de la arqueología del siglo xxi”, in La necrópolis de Vistalegre (Aspe, Alicante) 1985–1986, Alacant, 133–151.

MENCHÓN, J. 2011, “La torre de Vallferosa (Torà, La Segarra). Aproximació a la seva arquitectura i datació radiocarbònica”, in Actes del iv Congrés d’Arqueologia Medieval i Moderna a Catalunya, Tarragona, 769–782.

– 2013, “De Teodomiro a Tudmīr. Los primeros tiempos desde la arqueologia (s. vii –ix)”, in De Mahoma a Carlomagno. Los primeros tiempos (siglos vii–ix), Pamplona, 229–283.

– 2012, “Necròpolis de l’antiguitat tardana i de l’alta edat mitjana a les comarques del Camp de Tarragona, Conca de Barberà i Priorat”, in Arqueologia funerària al nord– est peninsular (segles vi–xii), Barcelona, 125–154.

115

RAMON MARTÍ & MERCÈ VILADRICH MOLIST, N.; BOSCH, J. M. 2012, “El cementiri medieval de Sant Miquel d’Olèrdola (Olèrdola, Alt Penedès)”, in Arqueologia funerària al nord–est peninsular (segles vi– xii), Barcelona, 469–494. MOLIST, N.; RIPOLL, G. (eds.) 2012, Arqueologia funerària al nord–est peninsular (segles vi–xii), Barcelona. MORELL, N. 2006, “Forjant el passat. Recerques arqueològiques entorn la metal·lúrgia del ferro al Serrat dels Tres Hereus”, L’Erol: revista cultural del Berguedà 91, 19–23. NEGRE, J. 2013, De Dertosa a Ṭurṭūša. L’extrem oriental d'al–Ṯagr al–Aˁlà en el context del procés d'islamització d'al–Andalus, [Diss.], Universitat Autònoma de Barcelona, Bellaterra: http://www.tdx.cat/ handle/10803/116319 PALOL, P. 2004, El castrum del Puig de les Muralles de Puig Rom (Roses, Alt Empordà), Girona. PASTOR, E.; BALLESTÍN, X. (eds.) 2013, Lo que vino de Oriente. Horizontes, praxis y dimensión material de los sistemas de dominación fiscal en al–Andalus (siglos vii– ix). BAR. Límina / Límites. Archeologie, storie, isole e frontiere nel Mediterraneo (365–1556), BAR Publishing, Oxford. PÉREZ, S. 2008, “Toponímia d’origen islàmic a la Catalunya Vella”, Societat d’Onomàstica: butlletí interior, 110, 64–87. ROIG, J.; COLL, J. M. 2011, “Esquelets humans en sitges, pous i abocadors als assentaments rurals i vilatges de l’Antiguitat Tardana de Catalunya (segles v–viii): evidències arqueològiques de la presencia d’esclaus i serfs”, in Actes del iv Congrés d’Arqueologia Medieval i Moderna a Catalunya, Tarragona, 75–82. – 2013, “Los palatia altomedievales del Vallès (siglos X– XI): la aportación de la arqueología”, in Lo que vino de Oriente. Horizontes, praxis y dimensión material de los sistemas de dominación fiscal en al–Andalus (siglos vii– ix), Oxford, 223–257. RUIZ, J.; VILLAR, V.; SUBIRÀ, M. E. 2007, “La població visigòtica de Can Gambús (Sabadell)”, in III Congrés d’Arqueologia Medieval i Moderna a Catalunya, Sabadell, 799–806.

116

SABATÉ, F. 1996, L'expansió territorial de Catalunya (segles IX–XII): conquesta o repoblació?, Lleida. SABATÉ, F. (dir.); BRUFAL, J. (coord.) 2011, IV Curs internacional d’arqueologia medieval: Els espais de secà, Lleida. SALRACH. J. Mª 2004, Catalunya a la fi del primer mil·lenni, Lleida. SANCHO, M. 2010, “Els Altimiris”, Villes et campagnes de Tarraconaise et d’al–Andalus (vie–xie siècle) : la transition : Villa 3, Toulouse, 67–90. SÉNAC, PH. 2002, Les Carolingiens et al–Andalus (viiie– ixe siècles), París. – (ed.) 2006, De la Tarraconaise à la Marche Supérieure d'al–Andalus (ive–xie siècle). Les habitats ruraux : Villa 1, Toulouse. – (ed.) 2007, Histoire et archéologie des sociétés de la vallée de l’Ébre (vii–xie siècles) : Villa 2, Toulouse. – (ed.) 2010, Villes et campagnes de Tarraconaise et d’al–Andalus (vie–xie siècle) : la transition : Villa 3, Toulouse. – (ed.) 2012, Histoire et archéologie de l'Occident musulman (viie–xve) : Al–Andalus, Maghreb, Sicile : Villa 4, Toulouse. SOLER, J.; RUIZ, V. 2013, “Els palaus periurbans de Terrassa. Una xarxa fiscal andalusina al voltant de la seu d’Ègara”, in Lo que vino de Oriente. Horizontes, praxis y dimensión material de los sistemas de dominación fiscal en al–Andalus (siglos vii–ix), Oxford, 205–210. VARIA 2012, 711. Arqueología e historia entre dos mundos, Zona Arqueológica, 15, Alcalá de Henares. Varia 2013, De Mahoma a Carlomagno. Los primeros tiempos (siglos vii–ix), Actas de la xxxix Semana de Estudios Medievales de Estella, 17 al 20 de julio de 2012, Pamplona. YZQUIERDO, P. 1998, “Làpida fundacional de les drassanes de Tortosa”, in L’islam i Catalunya. Catàleg, Barcelona, 47–48.