Ancient Macedonia 9783110718683, 9783110718645

Citation preview

Miltiades B. Hatzopoulos Ancient Macedonia

Trends in Classics – Key Perspectives on Classical Research

General Editors Franco Montanari and Antonios Rengakos Series Editors P. J. Finglass, S. J. V. Malloch, Christos Tsagalis Associate Editors Anna Marmodoro and Elena Isayev

Volume 1

Miltiades B. Hatzopoulos

Ancient Macedonia

ISBN 978-3-11-071864-5 e-ISBN (PDF) 978-3-11-071868-3 e-ISBN (EPUB) 978-3-11-071876-8 ISSN 2626-1030 Library of Congress Control Number: 2020946505 Bibliographic information published by the Deutsche Nationalbibliothek The Deutsche Nationalbibliothek lists this publication in the Deutsche Nationalbibliografie; detailed bibliographic data are available on the Internet at http://dnb.dnb.de. © 2020 Walter de Gruyter GmbH, Berlin/Boston Cover image: Domenico Spinosa, Fondo Marino, courtesy of Nicola Spinosa Printing and binding: CPI books GmbH, Leck www.degruyter.com

Preface Two years ago the editors of the new De Gruyter series Key Perspectives on Classical Research kindly invited me to contribute to it by writing a volume which would provide a critical reappraisal of research on the kingdom of Macedon conducted in recent decades and even further back. After a number of epistolary exchanges with Patrick Finglass, Simon Malloch and Christos Tsagalis, the tutelary ‘triad’ of the series, it was decided that my brief would consist in “an in-depth critical presentation of a selected number of topics especially prominent” in the field of Macedonian studies. I proposed exempli gratia (and rather pell mell) “the language or dialect, the political institutions, the assassination of Philip II, the causes of the Third Macedonian War etc.”, all subjects which had interested me recently. On second thoughts, it seemed to me that the selected subjects should be topical enough, so as to interest a public beyond the narrow circle of experts, and at the same time methodically presented, so as to ease the wandering of the reader through the labyrinth of Macedonian scholarship. Consequently, the volume was structured on three main axes (parts II, III and IV) devoted respectively to the land, the people and to a couple of prominent Macedonian personalities who have changed the course of world history. These are preceded by an introduction explaining the significance of the ancient Macedonian kingdom and are followed by an ‘envoi’ intended both as a warning against the unreliability of even one of our best ancient sources and as an assessment of the way travelled through in the volume. From the previous paragraph it follows that the present book does not include an array of subjects. It is not an annotated bibliography, such as an entry from Oxford Bibliographies Online. 1 Nor is it a history of Macedonia. Anyone who wishes to discover the events that raised a small principality on the northern marches of Greece to world power has a number of choices, from the short but convenient A History of Macedonia by R. Malcolm Errington (1990) to the threevolume monumental History of Macedonia by N. G. L. Hammond in collaboration with G. T. Griffith and F. W. Walbank (1972–1988). It is even less a “Companion” volume, such as the excellent Brill’s Companion to Ancient Macedon. Studies in the Archaeology and History of Macedon, 650B.C–300 AD edited by Robin Lane Fox (2011), or the Blackwell Companion to Ancient Macedonia edited by Joseph

 1 For a bibliographical guide see Andrianou/Hatzopoulos 2011. For an overview, see Hatzopoulos 1981; Hatzopoulos 1990; Hatzopoulos 2011c. Cf. Molina Marin 2015. https://doi.org/10.1515/9783110718683-202

VI  Preface Roisman and Ian Worthington (2010), since it does not aim at covering all aspects of Macedonian history, archaeology and culture in antiquity. Readers particularly interested in Macedonian archaeology can also consult with great profit the sumptuous publication of the Louvre Museum Au royaume d’Alexandre le Grand. La Macédoine antique (Paris 2011). Nevertheless, the present volume does have a historical frame, since its timespan extends from the seventh century to the end of Macedonian independence in 168, and it concerns equally the archaic, the classical and the Hellenistic periods, each of which is represented in proportion to the available sources. Some chapters in parts II and III inevitably follow a chronological order and may seem to favour the earlier stages of the formation of the kingdom and of its institutions, because the refoundation of the state took place in the reign of Philip II. However, the source material used in these parts for defining, for instance, the ‘national’ territory, the royal land and the allied cities of Macedonia or for describing the tongue and the the political and social institutions of the Macedonians comes at least as much from the Hellenistic period, while the first chapter of part IV concerns this same period exclusively. Conversely, the Roman period has not been included in this volume, for after 168 the erstwhile sovereign kingdom of the Macedonians was no more. It had been reduced first to a protectorate and then to a mere province of the globalised Roman Empire. By losing its autonomy, it also lost its originality. Its main institutions – King, Companions, Assembly – were suppressed. There were still shepherds, shopkeepers, even a new class of landowners, but the old aristocracy had been exterminated on the battlefield and by mass deportation. As a subject of research, Roman Macedonia is no less interesting than any other Roman province, but it appeals to a different, Roman-oriented, public. There remains the pleasant duty of thanking all those who contributed to the preparation of the present book. First to Patrick Finglass, Simon Malloch and particularly Christos Tsagalis, who offered me the opportunity to write and to publish it in the new De Gruyter series. Then, to my Academy colleague Antonios Rengakos, who encouraged me to accept this offer; to my friend and colleague at KERA Paschalis Paschidis, with whom I have engaged for years in an unending dialogue on all matters Macedonian; to my epigraphist colleagues Klaus Hallof and Demetrios Bosnakis, who kindly communicated to me the new and unpublished Macedonian decrees from Kos; to my German colleague Sabine Müller, who generously sent me her recent book on Perdikkas II; to my Italian colleague Monica D’Agostini, who did the same with the her equally recent monograph on the early years of Philip V; to our librarian and dear colleague at KERA Sophia Saroglidou, who never failed to provide me with information on the publications

Preface  VII

I was after; to my boyhood friend Thanos Veremis, who improved this preface, to my archaeologist colleagues Chrysoula Paliadeli, Angeliki Kottaridou, I. Graekos and Elizabeth Tsigarida, to whose generosity I owe the illustrations of this edition; and last but not least to Véronique Hautefeuille, without whose understanding and unconditional support the present book would not have been written.

Contents Preface  V Note on Transliterations  XI List of Figures  XIII 

Introduction: Why does ancient Macedonia matter?  1

 The Land: Where was Macedonia?  4 . The archaic period  11 .. The meaning of the term ‘Macedonian expansion’  18 .. Material finds and ‘national’ identities  19 ... ‘Sindos’  19 ... Hagia Paraskevi  21 ... Aigeai (Vergina)  21 ... Archontikon  22 ... ‘Thermi’  23 ... Nea Philadelpheia (Klitai?)  24 ... Aiane  24 . The classical and Hellenistic periods  33 . National territory, royal land and allied cities  38 . Some particular cases  41 .. Paionia  41 .. Perrhaibian Tripolis  41 .. Magnesia  42 .. Tymphaia, Parauaia, Atintania  43 .. Derriopos  47  . . . . .. .. . . ..

Who were the Macedonians?  49 The evidence of the ethnics  49 Weder Blut noch Boden: How does one become Macedonian?  50 The origin of the Macedonians  60 The speech of the ancient Macedonians  63 The ‘Prehistory’  64 The ‘pentekontaetia’ revolution  73 Macedonian cults and beliefs  79 Customs and institutions  88 Introduction  88

X  Contents .. .. .. .

Ethnos or polis?  95 Heerkönigtum, feudalism, absolutism or what?  103 Laws, decrees, epistatai, politarchai and merides  116 Perception of self and the other  121

 . ..

Personalities  125 Philip II  127 Philip’s span of life, length of reign and dates of accession to power and death  128 Philip hostage in Illyria and Thebes  130 Philip’s regency?  135 Philip II’s wives  138 Fathers and sons: Philip and Alexander  142 Philip’s murder  142 Philip II’s last abode  147 Philip’s heritage and Alexander’s plans  161

.. .. .. . .. .. .. 

Envoi: Were the Macedonians visited by heaven-sent madness?  170



Conclusion  177

Abbreviations and Bibliography  179 Index  205 Figures  239

Note on transliterations Unfortunately, there are no generally accepted rules for rendering Greek proper names (ancient and modern) into English. The rules of transliteration followed in the present volume are the following: 1. Names of ancient authors retain their traditional latinised form (Thucydides, not Thoukydides). 2. All other ancient proper names, with the exception of a few which have become part of English vocabulary (e.g. Athens, Corinth, Delphi, Alexander the Great, Philip II) are transliterated into Latin script with each letter retaining its conventional Greek phonetic value (Kassandros, not Cassander; Philippoi, not Philippi; Kalindoia, not Calindoea). 3. Modern Greek place names inherited from antiquity or coined from elements of the ancient vocabulary retain their ancient form, even if the pronounciation has changed (Beroia, not Veria; Thessalonike, not Thessaloniki; Kleidi, not Klidi). 4. Other modern names are rendered phonetically. This category also includes ancient names given to modern settlements, although these do not correspond geographically to the homonymous ancient ones. In such cases they figure between inverted commas. Thus ‘Thermi’ (previously Sedes, not to be confused with ancient Therme), ‘Sindos’ (previously Tekeli, not to be confused with ancient Sindos), etc. 5. In citations the original spelling of the publication is maintained.

https://doi.org/10.1515/9783110718683-204

List of Figures Fig. 1: Map of Macedonia  239 Fig. 2: Defixio from Pella in the Macedonian dialect (by courtesy of the Archaeological Ephorate of Pella)  240 Fig. 3a: The ‘Hunt’ fresco from the frieze of Tomb II (Aigeai/Vergina) (by courtesy of the Archaeological Ephorate of Emathia)   241 Fig. 3b: The ‘Hunt’ fresco (restored) (by courtesy of G. Miltsakakis and the Vergina Archaeological Mission of the Aristotelian University of Thessalonike)   241

https://doi.org/10.1515/9783110718683-205

 Introduction: Why does ancient Macedonia matter? The name of Macedonia inescapably evokes that of Alexander. But the empire he founded disintegrated centuries ago, and the sands of Egypt have effaced even the traces of his last abode. Adepts of political correctness can freely question his achievements and line him up along with Genghis Khan and Timur Lenk as one the major scourges of humanity. It is perhaps another passing fashion, no less than the uncritical admiration Alexander enjoyed in the colonialist era. But does Alexander exhaust by himself the contribution of Macedon to world history? In fact, Alexander was but the brilliant propagator of a Macedonian heritage which dated from before his birth and survived long after his death. The mountain chains of Pindos and of its southern ramifications, from Lake Ochrid and the twin Prespa lakes down to the Corinthian Gulf, cut the Greek peninsula in two: in the east, along the Aegean coast, a Greece of polis-states, in the west a Greece of ethnos-states. For Thucydides (1.5.3–1.6.2) genuine Greece did not extend beyond Delphi. Westwards and northwards the Ozolian Lokroi, Aitolians, Akarnanians and even more so the Epirotes and the other peoples (in Greek ethne) of Upper Macedonia (Lynkestai, Orestai, Elimiotai), even if they were not authentic barbarians, “lived like barbarians” (Thuc. 1.6.1). Thucydides meant that they did not live in walled urban centres which dominated the surrounding countryside and concentrated the main economic and social activities, and practically all the functions of the state, but in unfortified villages of equal status disseminated over extensive territories and having as a single point of reference a common sanctuary, such as Thermos in Aitolia, Dodona in Epirus or Itonos in Thessaly. The survival of kingship in some of these ethnos-states constituted for the Athenians of the classical period an aggravating factor, which (if politically expedient) could be used as a criterion for excluding them from the Hellenic community, although they shared with the other Greeks the same language and the same religion. The Macedonian kingdom under the Temenid dynasty formed a particular case. The central Macedonian plain faced the Aegean sea. In the fifth century Pella, Beroia, Aigeai and Europos were already urban centres. Nevertheless, even for a pro-Macedonian such as Isocrates, who made a clear distinction between Macedonians and barbarians, these elements did not suffice to make him accept the Macedonians as authentic Greeks, because of their monarchical regime. The Macedonian state was founded by the Argeads, one of the numerous ethne or “clans” hailing from the western mountain chains, who conquered https://doi.org/10.1515/9783110718683-001

  Introduction: Why does ancient Macedonia matter? “coastal Macedonia” and settled on the land. In spite of the urban centres they found there, annexed and further developed, they retained from their past the institution of kingship and the primary centrality not only of the royal residences of Aigeai and later Pella, but also of a common sanctuary: that of Zeus at Dion. There, every October, representatives of local communities and ordinary citizens gathered to offer sacrifices, celebrate contests, and take counsel together about the common weal. The originality of the Macedonian kingdom consists in this combination of archaism and modernity, in the appetite of the Macedonian élites for the latest intellectual quests and cultural creations, and also in the parallel conservation of the values of an epic past. No wonder that this heady mixture exerts an irresistible attraction to those who study it. The Macedonian sovereigns imported poets, historians, philosophers, architects and painters from Athens and other city-states renowned as centres of letters and arts, though not to copy them passively. They favoured the appearance of new artistic and intellectual schools and of particular genres which corresponded to local conditions and mentalities. The combination of architectural orders, the emancipation of decoration from structure in monumental buildings, the use of stucco as a complement to masonry and illusionist painting – all these were first developed in Macedonia before being disseminated acrοss the Hellenistic world and later copied by Rome. The same is true of vaulted tombs with monumental façades, which we call “Macedonian”, of the large and sumptuous honorary tholoi (rotundas) and, of course, of the royal palaces. Even the koine, the Greek “common tongue”, which spread across the ancient world as an international language and which is the direct ancestor of Modern Greek, had its cradle in Macedonia, where Philip II (or perhaps Archelaos) adopted Attic Greek, instead of the Macedonian dialect, for the purposes of administration. The vehicle of Buddhism at the gates of India, it later became the language through which Christian religion was propagated in the Mediterranean. It survives today as the liturgical language of the Greek Orthodox Church. These facts have been known for some time and today tend to be generally acknowledged. What has not been yet fully realised is that the kingdom of Macedonia was the harbinger of modern European monarchies, whose direct heirs are modern democracies, be they formally kingdoms or republics, in which an elected prime minister or president wields more power than any king of the past. The combination of a strong executive authority embodied in the king assisted by his Companions/Friends, who formed his Privy Council and General Staff, with extensive local autonomy for the numerous cities of the realm assured an equilibrium between centre and periphery. In effect, each city had its own citizenship,

Introduction: Why does ancient Macedonia matter?  

its own legislation, its own assembly and council, its own magistrates. The cities interacted with the central authorities through the epistatai, the chief magistrates of each city, but also thanks to the continuous renewal of court aristocracy by local magistrates and army officers of the local levy who distinguished themselves. The foremost cohesive element of the Macedonian people was the army, which fell under the direct authority of the king. It comprised both professional units and numerous and well-trained reserves thanks to the institutions of the gymnasium and the ephebeia, a two-year training for youths aged between eighteen and twenty years. This unique political system, which was no less inventive than that of the Athenian democracy, was abolished by the Romans, who, however, ended up adopting many elements of Hellenistic kingship. These elements, were reintroduced into Western European states by jurists of the later MiddleAges, and greatly influenced the evolution of modern monarchies, whose heirs are our contemporary democracies, that combine extensive territories peopled by citizens possessing a vigorous national identity, a strong central government and more or less wide local autonomy, though that autonomy is generally less extensive than in ancient Macedonia. Thus, it would not be out of place to submit that the ancient Greeks not only invented democracy, symbolised by the Parthenon, but also the modern national state embodied in the ‘democratic’ royal palace of Aigeai, with its porticoes wide open to the public.

 The Land: Where was Macedonia? “Macedonia was an historical-geographical term based upon confused historical memories and devoid of geographical significance ... Disappearing at the time of the Turkish conquest (the Turks spoke only of Rumeli) the term was resurrected during the classical revival only to be variously used and later to be deliberately misappropriated”, wrote a British modern historian half a century ago. 1 Until very recently if somebody entered the word “Macedonia” into an internet search engine, the first item that would appear would be the Wikipedia article “Republic of Macedonia”, with the following commentary: “The Republic of Macedonia geographically roughly corresponds to the ancient kingdom of Paeonia, which was located immediately north of the ancient Kingdom of Macedonia. Paeonia was inhabited by the Paeonians, a Thracian people, whilst the northwest was inhabited by the Dardani and the southwest by tribes known historically as the Enchelae, Pelagones and Lyncestae; the latter two are generally regarded as Molossian tribes of the northwestern Greek group”. The second item was another Wikipedia article: “Macedonia-Wikipedia”, itself subdivided into nine articles: 1) Republic of Macedonia, a country in southeastern Europe; 2) Macedonia (Greece), a traditional geographic region, spanning three administrative divisions of northern Greece; 3) Macedonia (region), a region covering all of the above, as well as parts of Bulgaria, Albania, Kosovo and Serbia; 4) Macedonia (ancient kingdom), also known as Macedon, the kingdom that became Alexander the Great’s empire; 5) Macedonia (Roman province), a province of the early Roman Empire; 6) Diocese of Macedonia, a late Roman administrative unit; 7) Macedonia (theme), a province of the Byzantine Empire; 8) Independent State of Macedonia, a proposed puppet state of the Axis powers (1944); 9) Socialist Republic of Macedonia, a part of the former Yugoslavia (1944–1991), predecessor of the Republic of Macedonia. In these articles the reader could follow the protean significance of the geographical term from a small kingdom on the shores of the Thermaic Gulf in the sixth century to Greater Macedonia under Philip II and his successors until 167, to the Roman province established in 146, which constantly changed boundaries during the Republic and the Early Empire, to its division between Macedonia Prima (or Macedonia tout court corresponding roughly to the Greek province of Macedonia) and Macedonia Secunda or Salutaris (territorially corresponding roughly to present-day North-Macedonia) in the late fourth century, to the creation of the Diocese of Macedonia under Constantine the Great, regrouping several

 1 Dakin 1966?, 3, n. 4. https://doi.org/10.1515/9783110718683-002

The Land: Where was Macedonia?  

Balkan provinces of the Roman Empire, to the Byzantine theme of Macedonia situated outside Macedonia in Western Thrace, to the modern Greek province of Macedonia, to the short-lived puppet state created by the Germans in 1944, to the Socialist Republic of Macedonia, which succeeded it in the same year and finally became the Yugoslav Republic of Macedonia. In 1991 this federated state seceded from Yugoslavia and declared its independence dropping the “Yugoslav” part of its name. Situated almost completely outside the borders of the ancient kingdom of that name, it was admitted to the United Nations under the provisional name of Former Yugoslav Republic of Macedonia or FYROM. North-Macedonia, the new official name of this entity, is now added to the list of Macedonian metamorphoses. How did such a paradoxical result come about? The answer is to be sought in the third of the above-mentioned articles, Macedonia as a geographical region including, in addition to the Greek province of Macedonia and to North-Macedonia, parts of Bulgaria, Albania, Serbia and even Kosovo. By the end of the eighth century, when the theme of Macedonia was created in Thrace, the term had lost all connexion with the ancient kingdom, due to successive Roman administrative reforms, though memories of it continued to linger. 2 The Ottoman conquest emptied it of all practical relevance. When early modern geographers rediscovered the works of Strabo and Claudius Ptolemy, and started introducing ancient geographical terms into their maps, they did so with the loosest approximation. See, for instance, the maps of Ortelius (1570), Mercator (1598), Mariette (1645), Blaeu (1650), De Wit (1680), Nolin (1699) and Homann (1740). 3 A new era began with the French Revolutionary and Napoleonic wars, which opened the Ottoman Empire to European political and scientific scrutiny. The ensuing progress of cartography coincided with parallel territorial losses by Turkey. Thus, after the Treaty of Berlin (1878), Ottoman possessions in the Balkans had shrunk to six “vilayets” (provinces): Adrinople, Shkodra, Iannina, Skopje, Monastir and Salonica. The latter three had been lately included in the enlarged state of Bulgaria by the stillborn Treaty of San Stefano (1878), and remained a hotbed of Bulgarian irredentism. For practical reasons European diplomacy arbitrarily regrouped them under the name “Macedonia” reserving (on sounder historical

 2 A twelfth century byzantine text (Timarion) uses the term “Old Macedonia” (παλαιὰ Μακεδονία) to differentiate the area of the ancient kingdom from the homonymous theme. See Τιμαρίων ἢ περὶ τῶν κατ’ αὐτὸν παθημάτων, in Ellisen 1860, 3. 3 Colocotronis 1919, pl. VIII–XIII.

  The Land: Where was Macedonia? grounds) the name “Thrace” for the vilayet of Adrinople. Although this nomenclature (especially concerning the three ‘Macedonian’ vilayets) did not correspond to any administrative, historical or natural entity, it was readily adopted by diplomats and geographers alike, and thus the territorial settlement of the Balkan Wars (1912–13) was presented as a partition of an unimpaired Macedonia, which in fact had never existed within such frontiers, into several parts, but mainly three: Aegean Macedonia, Vardar Macedonia and Pirin Macedonia, annexed respectively by Greece, Serbia and Bulgaria. Other factors contributed to and simultaneously aggravated the confusion regarding the configuration of Macedonia. 4 First, no ancient work comparable to Pausanias’ Description of Greece that includes Macedonia within its scope (instead of stopping at Ozolian Lokris as Pausanias does) had survived from antiquity. Due to additional ill-luck the seventh book of Strabo’s Geography, with the description of Macedonia, has come down to us in a fragmentary state. Modern geographers working with literary sources had to make do only with lists of place names, mostly names of cities, found in the works of Pliny, Claudius Ptolemy, Hierocles, in Roman itineraries and in late lexica. Secondly, Macedonia occupied the northernmost part of the (ancient and modern) Greek world, and due to its location was most exposed to all sorts of invasions and demographic upheavals, which resulted in frequent change of place names. Thus, to mention a couple of instances, besides Mount Olympοs and the city of Beroia, which clung to their antique names, or Thessalonike (Θεσσαλονίκη), which during the Middle-Ages merely lost its first syllable (Σαλονίκη) in the popular idiom, 5 Pella became Hagioi Apostoloi, and Edessa took the Slavic name of Vodena (Βοδενά) in the Middle Ages, retaining its ancient name only in ecclesiastic terminology. Unfortunately, the cases in which the ancient place names were retained or in which their memory persisted were the exception. When Renaissance and early modern scholars became interested in putting on the map place names that they read in ancient authors or in identifying ancient remains that they encountered on their

 4 For what follows, see Hatzopoulos 2006a, 19–33. The two fundamental works on the historical geography of Macedonia are Hammond 1972 and Papazoglou 1988b. For Chalkidike in particular, Zahrnt 1971 remains indispensable. 5 The city was founded by Kassandros, who named it Θεσσαλονίκεια in honour of his wife Θεσσαλονίκη. However, the simplified form Θεσσαλονίκη soon prevailed. During the Middle-Ages the name of the city was further reduced to Σαλονίκη in popular parlance, various adaptations of which were adopted by Western European and Slavic languages (and in Turkish). Nevertheless, the ancient name continued to be used by educated people, and Θεσσαλονίκη has become again the current name of the city.

The Land: Where was Macedonia?  

journeys, they could only propose unverifiable hypotheses, unless their texts provided clear topographical indications. A case in point is the controversy over the two royal capitals of Macedonia. Hellenistic Pella was destroyed by an earthquake in the first years of the first century B.C. 6 A Roman colony was founded by Brutus or Caesar and then by Augustus under the name of Colonia Iulia Augusta Pella. At the time of Dio Chrysostom (first-second century A.D.), Pella (presumably the Hellenistic city) lay in ruins. After further damage sustained during the Gothic raids of the late third century A.D. the city was restored by Diocletian and for a brief period was renamed Diocletianopolis after him. It is last mentioned in a historical narrative in connexion with the presence of Theodoric’s Ostrogoths in Macedonia in A.D. 482, and as a contemporary city of Macedonia Prima in the sixth-century work Synekdemos attributed to Hierocles. A place called Βασιλικὰ Ἀμύντου by Procopius may correspond to the site of Hellenistic Pella, where the remains of a small Early Christian settlement have been discovered. The city was finally destroyed and its site abandoned during the Slav invasions. Its name reappears six centuries later in the title of the bishops Σ(θ)λανίτζης ἤτοι Πέλλης, whose See was probably situated in the town that would be given the name Yenidjé Vardar (Yannitsa in Greek) after the Ottoman conquest. In modern times the first attempt at identifying the site of ancient Pella was made by the eighteenth-century Greek scholar Meletios, who hesitated between Yannitsa and a place “called today ta Palatia” (“The Palaces”). Less than a century later, Pouqueville identified this site, which he calls Palatitzia, with the village Allah Kilissa. In these same years three other travellers, Beaujour, Cousinéry and Leake, confirmed the identification with the same village (variously spelled), the Greek name of which was Hagioi Apostoloi, adding the information that two kilometres to the west a source retained the name of Pella (Πέλλα in Greek, Pel in ‘Bulgarian’). So where was ancient Pella, at Yannitsa, at Hagioi Apostoloi or at a site by the source of that name? Although the distances transmitted by the Roman Itineraries favoured the location of Pella on the site by the source, nineteenth-century travellers unanimously chose the one by Hagioi Apostoloi, because they were impressed by its resemblance to the description of Pella in Livy (44.46.4–11, after Polybius). However, they could offer no valid explanation for the transfer of the place name to the source or to Yannitsa nor could they justify the variety of names under which the site was known. It took the systematic excavations of G. Oikonomos in 1914–15, resumed by Ch. Makaronas and Ph. Petsas in the 1950s, to solve the puzzle. The archaeological  6 On the identification, history, topography and archaeological research of Pella, see now ΕΚΜ II p. 601–636.

  The Land: Where was Macedonia? discoveries made clear that the capital of the Macedonian kings from Amyntas III to Perseus was at the village of Hagioi Apostoloi (consequently renamed “Palaia Pella”), while the Roman colony was at the site near the homonymous source and the modern village of Nea Pella 7. Thanks to subsequent excavations epigraphic evidence is now amply available, vindicating the choice of the nineteenth- century travellers and the resumption of the glorious name of the royal capital by the deme of Hagioi Apostoloi. The question of the location of Aigeai, the first capital and according to Diodorus (22.12.1) the “hearth” (ἑστία) of the Macedonian kingdom, is both more complicated and fraught with political implications 8. Until the 1960s all specialists, with the exception of the nineteenth-century German scholar Th. L. Fr. Tafel, agreed that before its conquest by Karanos (founder of the Temenid dynasty), Aigeai was called Edessa, which they identified with ancient Edessa, the mediaeval Vodena. This conviction was based on evidence from the works of Euphorion of Chalkis, a poet from the third century, and of Justin, a Roman historian who wrote an Epitome of the Historiae Philippicae by Pompeius Trogus, a first-century Latin historian of Gallic extraction. Against this established opinion that Aigeai and Edessa were one and the same city or two parts of the same city Tafel made the following objections: 1) Several ancient authors, namely Pliny the Elder (HN 4.33 and 6.34), Plutarch (Pyrrhus 10.2; 12.6; 26.6) and Claudius Ptolemy (3.13.39) made a clear distinction between Aigeai and Edessa; 2) the meteorological phenomenon observed by Theophrastus (De ventis 27) at Aigeai and attributed by him to the proximity of Mount Olympos could not concern Edessa situated some 90 kilometres to the north, but a distinct city closer to that mountain; 3) no royal tombs had been reported at Edessa, whereas according to ancient authors the royal cemetery was situated at Aigeai. 9 It is not easy to explain why such strong arguments were not taken seriously. Perhaps the reasons were more sentimental than rational. The magnificent natural setting of Edessa with its panoramic Upper City, cataracts and luxurious vegetation deserved to be chosen as the capital of a glorious dynasty more than any other city in Macedonia. Such at least was the feeling of all its visitors who have left us lyrical descriptions of its beauties. As a result, L. Heuzey did not realise that he might have discovered a royal capital, although he was fully aware of the

 7 On the date of the transfer of the usual royal residence from Aigeai to Pella, see Hatzopoulos 1987a, 41–44. 8 On the identification, history topography and archaeological research of Aigeai, see also ΕΚΜ II, p. 60–62. 9 Tafel 1842, 48–50.

The Land: Where was Macedonia?  

paramount significance of the palace he had unearthed at Palatitsia; fully aware, too, that the Great Tumulus, according to him “the most beautiful tumulus of Macedonia”, concealed a funerary monument of the greatest importance; although he did not fail to remark that the ruins of Palatitsia could possibly be identified with Ptolemy’s Aigaia, a city of Emathia. 10 Thus, with all due reservations, he inclined towards identifying Palatitsia with the obscure city of Balla. It took nearly a century before two scholars, the Yugoslav Fanoula Papazoglou and the Englishman N.G.L. Hammond successfully challenged the communis opinio. The former in her doctoral thesis of 1957 alleged literary and epigraphic evidence, which mentioned the names of both cities and attested the use of two distinct ethnics, Edessaios and Aigaios, proving that we are dealing with two different cities. She also drew attention to the absence of royal tombs in Edessa, adding that sumptuous tombs were indeed discovered, but south of Edessa, in the neighbourhood of the modern town of Naoussa 11. Eleven years later, at the first international conference on ancient Macedonia held in Thessalonike in 1968, Hammond added two more arguments to those put forward by Tafel and Papazoglou, namely a) that Diodorus’ narration of the unsuccessful march of the pretender Argaios from Methone to Aigeai and back in 360 is incompatible with the 90 km distance from the former of these two cities to Edessa, and b) that an inscription from Argos dating from the late fourth century and enumerating cities of Macedonia in a geographical south to north order mentioned Aigeai, then another city whose name is now lost, probably Beroia, and after that Edessa. Hammond, however, went beyond the mere rejection of the identification of the two cities. He argued that the ancient site which satisfied all requirements was that of Palatitsia-Vergina. It was close enough to Mount Olympos and its Pierian ramification to satisfy the conditions of the phenomenon observed by Theophrastus; it displayed a “built Macedonian tomb” containing a throne, which might be that of Philip II; being at a mere distance of 25 kilometres from Methone, it was compatible with Diodorus’ narration of the march of Argaios; finally, it lay to the south

 10Heuzey 1876, 183: “Il est vrai que Ptolémée fait d’Αἰγαία une ville émathienne ; mais il n’est pas même bien prouvé que le canton de Palatitza appartînt à l’ancienne Piérie. Aujourd’hui ces villages au-delà du fleuve, loin de se rattacher à l’évêché piérien de Kitros, sont dans la dépendance du métropolitain d Verria. Or on sait à quelle haute autorité remontent parfois les conscriptions ecclésiastiques, calquées sur les anciennes délimitations politiques du pays. Il est de fait que la chaîne épaisse des monts Piériens forme en cet endroit une limite plus effective que le cours guéable de l’Haliacmon, et que cette partie de la rive droite du fleuve paraît se relier naturellement à la grande plaine d’Emathie”. 11 Papazoglou 1957, 111.

  The Land: Where was Macedonia? of both Edessa and Beroia. 12 M. Andronicos, present at the conference, was convinced, while Ph. Petsas disagreed and commented on Hammond’s theory ironically. From being merely academic, the subject became personal and political after the sensational discovery of the royal tombs during Andronicos’ excavation at the Great Tumulus of Vergina. It was not unexpected that Petsas, who had directed excavations both at Levkadia-Kopanos in the neighbourhood of Naoussa and at Edessa, and had brought to light the imposing Tomb of the Judgement in the former site and the circuit wall of the lower city in the latter, and, in addition, had a long history of personal rivalry with Andronicos from their student years, would contest his colleague’s findings passionately. It was no surprise, either, that public opinion in Edessa, whose inhabitants took pride in being citizens of the first capital of Macedonia, and which displayed the inscription “Ancient site of Aigeai” at the entrance of the archaeological site, would react negatively to the loss of status. Harder to foresee was the reaction of some of Greece’s northern neighbours. Particularly revealing is an article by Hristo Andonovski entitled “Rich Archaeological Discoveries in Vergina” in the first issue of the new English-language periodical Macedonian Review published in 1979. The author challenged Hammond’s theory since “so far neither an ancient city nor an ancient theatre has been found in Vergina”, and interpreted the ready acceptance of the identification of Vergina with ancient Aigeai as politically motivated. “The old illness was, and is, that a portion of the learned Greeks and of the official government seek to wipe out anything that could evidence the fact that on the geographical space of Aegean Macedonia lived, and live today, people who are not of Greek descent ... Voden, up until 1912 and beyond, had pure Slav characteristics. It was not until the colonization of this area by fugitive inhabitants of Asia Minor that Slav characteristics were diluted. Voden falls within the ethnographic borders of the Macedonian people, while Ber and Vergina lie further south, just outside the ethnographic borders of the Macedonian people. If the hypothesis proved to be true, if the capital of the ancient Macedonian kingdom were to be found in Vergina and not in Voden, we ourselves would put forth an argument for the continuity of Greek culture within the Macedonian space, supporting the notion that ancient Macedonia had Greek ethnic origins. To date, however the supporters of such a theory have a very weak argument”. 13

 12 Hammond 1970, 53–67. By coincidence, in September 1970, during a visit at Vergina, the young R. Lane Fox observed the phenomenon described by Theophrastus: Lane Fox 1973, 504. 13 Andonovski 1979, 112.

The archaic period  

Since 1979, both the theatre and the ancient city have come to light in Vergina. It is not known whether the author of the above article has been convinced that the first Macedonian capital lay there, and that, following his reasoning, ancient Macedonia had Greek ethnic origins, but there are still Greek scholars who despite the royal palace, the royal tombs and the theatre deny the identification of Vergina with ancient Aigeai, and prefer to locate it either at Naoussa-Kopanos, where there is a theatre but neither palace nor royal tombs, or at Edessa, where none of these three elements is to be found. 14 There are several other examples of (until recently) contested identifications of Macedonian cities, such as Kalindoia, Morrylos, Arrolos, Berge, Strepsa, Herakleia Sintike etc., which have been settled in the last half-century thanks to the progress of archaeology, epigraphy and numismatics, without triggering similar controversies. Leaving aside the issue of the use – and abuse – of ancient Macedonian history in modern diplomatic and political debates, we must now examine the main relevant question in the context of ancient history: where exactly was Macedonia situated and where precisely its borders were. As I wrote some three decades ago, “there are peoples who owe their identity to a country, and countries that owe their identity to a people. If the French are the product of France, Macedonia, by contrast, is nothing more than the country conquered and populated by the Makedones. Hence the difficulty in defining it precisely in geographical terms. Throughout history, the boundaries of the country have followed the expansion of the Macedonian people, from the Pindos range in the west to the plain of Philippoi in the east, and from Mount Olympos in the south to the Axios gorge between Mounts Barnous (Nidje and Paikon) and Orbelos (Beles) in the north. We shall use here the term “Macedonia” to denote the country in the form it had achieved at the end of the Hellenistic period, before the Roman domination, as a result of conquest, colonisation of conquered lands, and expulsion or/and assimilation of ‘indigenous’ peoples”. 15

. The archaic period The question of the borders of Macedonia in the archaic period is indissolubly linked with that of the origins and the early stages of expansion of the Macedonian kingdom. As long as scholars situated Aigeai, “the hearth” of the kingdom,

 14 Faklaris 1994, 609–616; Touloumakos 2006, and my rebuttal Hatzopoulos 1996d, 264–269 (= Recueil 171–176); cf. Saatsoglou-Paliadeli 1996, 225–235; Hammond 1997a, 177–179. 15 Hatzopoulos 1993a, 19.

  The Land: Where was Macedonia? at or near Edessa, they had to face the difficult challenge of combining the two earliest, most detailed, but apparently contradictory pieces of evidence from the historical works of Herodotus and Thucydides respectively. Thucydides’ narrative is firmly dated, and refers to the kingdom of Macedonia as it stood in 429 or rather in the reign of Archelaos (414–399), after the end of the Peloponnesian War, when Thucydides was composing his historical work 16. The only indication helping to put Herodotus’ legend into a chronological frame is that Alexander I was the sixth descendant of Perdikkas, the founder of the kingdom. Given that Alexander came to power in c. 495, the foundation of the kingdom should be situated in the seventies of the seventh century or in the middle of that century, depending on our reckoning on a basis of 30 or of 25 years per generation respectively. Such a date seems to find confirmation in the archaeological evidence 17. Herodotus (8.137–138) stated that the founder of the kingdom was Perdikkas, the youngest of the three Temenid brothers hailing from Argos who from Illyria crossed over to Upper Macedonia and sought employment at Lebaie, the seat of a local king. Following an omen which announced that Perdikkas would inherit the kingdom, the three brothers had to flee to save their lives. In their flight from the king and his horsemen who were pursuing them, they were protected by a river which they alone were able to cross and to which they later made sacrifices of thanksgiving. They settled in the gardens of Midas under Mount Bermion, “in which roses grow of their own accord”, and issuing forth from it they started conquering the rest of Macedonia. At first sight, there was no contradiction between this tale and the supposed location of Aigeai at or near Edessa. After all, Edessa is located on the pass formed by the junction of the foothills of Mount Bermion and Mount Barnous (Nidje), its gardens and orchards are full of flowers and trees, making it a place that corresponds perfectly to Herodotus’ description of the gardens of Midas. This legend was hardly compatible with Thucydides’ (2.99) more sober narrative, however: “So Sitalces’ army was being mustered at Doberus and preparing to pass over the mountain crest and descend upon lower Macedonia, of which Perdiccas was ruler. For the Macedonian race includes also the Lyncestians, Elimiotes, and other tribes of the upper country, which, though in alliance with the nearer Macedonians and subject to them, have kings of their own; but the country by the sea which is now called Macedonia, was first acquired and made their kingdom by Alexander, the father of Perdiccas, and his forefathers, who were originally Temenidae from Argos. They defeated and expelled from Pieria the Pierians, who  16 Cf. Hornblower 1991, 375–376. 17 Saripanidi 2017, 86.

The archaic period  

afterwards took up their abode in Phagres and other places at the foot of Mount Pangaeus beyond the Strymon (and even to this day the district at the foot of Mount Pangaeus toward the sea is called the Pierian Valley), and also, from the country called Bottia, the Bottiaeans, who now dwell on the borders of the Chalcidians; they acquired, further, a narrow strip of Paeonia extending along the river Axius from the interior to Pella and the sea; and beyond the Axius they possess the district as far as the Strymon which is called Mygdonia, having driven out the Edonians. Moreover, they expelled from the district now called Eordia the Eordians, most of whom were destroyed, but a small portion is settled in the neighbourhood of Physca; and also from Almopia the Almopians. These Macedonians also made themselves masters of certain places, which they still hold, belonging to the other tribes, namely, of Anthemus, Grestonia, Bisaltia, as well as of a large part of Macedonia proper. But the whole is now called Macedonia, and Perdiccas son of Alexander was king when Sitalces made his invasion”. 18 Thucydides’ account of Macedonian expansion looks clear and well-informed. Nevertheless it poses a crucial problem: is the above order geographical, from south to north and from west to east, or chronological? Since accepting Thucydides’ narrative as chronological would entail the inescapable conclusion that the Temenid expansion began with the conquest of Pieria and not with that of Bottia, where Edessa is situated, most authors (Müller, Geyer, Kanatsoulis, Beloch, Cloché, Daskalakis) 19 who situated Aigeai at or near Edessa interpreted, explicitly or implicitly, Thucydides’ text as a geographical description. But such an interpretation was not satisfactory either. How could one explain why Eordaia and Almopia, the westernmost conquests of the Temenids, were mentioned after (and not before) Bottia (Lower) Paionia and Mygdonia? Thus many of them took the liberty of freely rearranging the chronological order of the conquests. 20 Only Abel, Hoffmann and Edson accepted Thucydides’ narrative as following a chronological order. 21 In my view Thucydides did not decide beforehand which order – chronological or geographical – he would follow. He started quite naturally from Pieria, the region closest to the cradle of the kingdom and presumably the first Temenid conquest; he proceeded to contiguous Bottia and next to adjoining to it (Lower) Paio-

 18 Translation by Charles Forster (Loeb Classical Library [1956]), slightly altered. 19 K. Müller 1825, 31; Geyer 1930, 39–41; Kanatsoulis 1964, 13–15; Beloch 19122, 341; Cloché 1960, 29–34; Daskalakis 1960, 235–236. 20 Zahrnt 1984, 332–333, with references. 21 Abel 1847, 143–146; Hoffmann 1906, 257; Edson 1970, 27, and particularly nn. 54, 56.

  The Land: Where was Macedonia? nia, and then to adjacent Mygdonia as far as the Strymon river. At that point, realising that he had omitted the by far westerner and expectedly earlier conquests of Eordaia and Almopia, he added them as an after-thought introducing them by the expression “Moreover, they expelled”, 22 before closing his enumeration by the chronologically latest and geographically easternmost and southernmost conquests of Krestonia, Bisaltia and Anthemous. Eventually, thanks to Hammond’s brilliant intuition, 23 the relocation of Aigeai at Vergina, in the region called Makedonis by Herodotus, seemed to resolve the apparent incompatibility between the texts of the two ancient historians in favour of the supporters of a chronological order in Thucydides’ text. The relative sequence of the different stages of Macedonian expansion could now be considered as nearly established, and the early history of Macedonia was finally taking a clear shape. Perdikkas, proceeding from the gardens of Midas on the foothills of Mount Bermion into the lower Haliakmon valley, seized first Aigeai and the surrounding area, i.e. the area called Makedonis by Herodotus. From there the Macedonians issued forth to conquer first Pieria and then Bottia. Thus after this second conquest, early Macedonia comprised the Haliakmon valley between the confluence with the Loudias and the northern extension of the Pierian Mountains, which Herodotus (7.131) names Makedonikon oros, as well as the foothills of Mount Bermion west of the upper course of the Loudias. The early presence of Macedonians in this area is also attested by the Hesiodic Catalogue of Women (fr. 7 M-W), which locates Magnes and Makedon at Mount Olympos and in Pieria respectively, and by another passage by Herodotus (1.56; cf. 8.43), which mentions Mount Olympos as the abode of the Makednon ethnos, the vast transhumant pastoral community a branch of which were the historical Macedonians. 24 The third stage of Macedonian expansion concerned the possession of the “narrow strip of Paionia extending along the river Axios from the interior to Pella and the sea”. 25 The geographical description of this area by Thucydides (2.99.4), like the corresponding passage of Herodotus (7.123.3: “the river Axios, that delimits Mygdonia and Bottia, a narrow strip of which is occupied by the cities of Ichnai and Pella”), is borrowed from Hecataeus. 26 It is remarkable that Herodotus updated the geographical terminology, attributing this region to Bottia instead of Paionia, a term that had become obsolete after the annexation of this narrow strip

 22 ἀνέστησαν δὲ καὶ ... 23 Hammond 1970, 53–67. 24 Cf. Zahrnt 1984, 348–352. 25 Thuc. 2.100.3. 26 Cf. Hammond 1972, 146–147.

The archaic period  

to the Macedonian kingdom. There are two more, indirect, attestations of this stage of Macedonian expansion. Strabo (7, fr. 20 Baladié) relates that there was a fortified place on the Axios called Abydon, identical with Homer’s Amydon, which was destroyed by the Argead Macedonians. There is no doubt that this event took place when the Macedonians annexed the right bank of the lower Axios. The other indirect attestation is the foundation by the conquering Macedonians of Allante, south of Ichnai. This new Macedonian settlement was destined to replace Abydon on the mouth of the Axios, which had by then been displaced southwards by centuries of alluvial deposits, and perhaps also to block Ichnai from the sea at a time when this settlement was still Paionian 27. There followed the fourth stage of expansion with the conquest of Mygdonia and the expulsion of the Edones, who were its former inhabitants. Finally the Macedonians conquered Krestonia, Bisaltia and Anthemous. It is not possible to situate with any chronological precision the annexation of Eordaia and Almopia and the extermination or the removal of the indigenous population, which Thucydides mentions as an after-thought just after the conquest of Mygdonia. It is, however, most probable that it had preceded it. The relative sequence of the different stages of Macedonian expansion could now be considered as nearly established, but there remained the question of their precise chronology. Fortunately, it was at least clear that Bisaltia and Krestonia were conquered after the retreat of the Persians, since according to Herodotus (8.116.1), the Bisaltians along with the Krestonians were still under the rule of their own king in 479. But does the third stage of the extension of the Macedonian kingdom in Lower Paionia, and also the fourth stage in Mygdonia “as far as the Strymon”, date from the Persian retreat after 479 or already from the reign of Amyntas I in the last years of the sixth century? Further information gleaned from other authors, but also from other passages of Herodotus and Thucydides themselves is in favour of the former alternative. Whereas Hecataeus calls Loidias (Loudias) a river of Macedonia (FGrHist 1 F 145), he considers Chalastra “a city of Thracians”, the Sindonians “a Thracian people” and Therme as a city of “Greek Thracians” (FGrHist 1 F 146–147), which means that, when his Description of the Earth was being written at the end of the sixth century, the Macedonians already possessed Pieria and Bottia, but had not yet crossed the Axios into Mygdonia. This conclusion is not contradicted by other passages of Herodotus. When he (5.17) relates the dispatch of seven Persian ambassadors by Megabyzos to Macedonia in c. 510 and writes that “there is a very short route from lake Prasias (lake of Pravi) to ‘Macedonia’” over Mount Dysoron (Mount Menoikion), he does so in  27 Edson 1955, 187, n. 71; cf. Gounaropoulou/Hatzopoulos 1985, 61.

  The Land: Where was Macedonia? an excursus switching from the ‘historical’ tenses (imperfect and aorist) to the present tense, indicating thereby that he is referring to conditions contemporary to the composition of his work and not of the late sixth century. It is obvious that archaic Macedonia did not extend as far as the Strymon valley. 28 Similarly, when he narrates Xerxes’ march through Macedonia, and, borrowing from Hecataeus’ work, writes (7.123.3) about the “river Axios which delimits the lands of Mygdonia and of Bottia”, and further down (7.127.1) “as far as the rivers Lydias (Loudias) and Haliakmon, which delimit the lands of Bottia and of Makedonis blending their waters in the same bed”, he switches once more from ‘historical’ aorist to present tense, showing again that he is referring to contemporary conditions. Therefore, it is not pertinent to argue from the absence of any mention of a ‘state’ border either on the Axios or on the common riverbed of the Loudias and the Haliakmon that Macedonia had already annexed Mygdonia. 29 Charles Edson held fast to the opinion that the conquest of Lower Paionia, Mygdonia and Athemous did not predate the Persian Wars, a view based on the fact that Ichnai in Lower Paionia and Lete in western Mygdonia continued to strike their own coinage during the first decades of the fifth century, 30 whereas the cities of the Macedonian kingdom did not enjoy this privilege until the second century, if they ever did. 31 New discoveries however have weakened the force of this argument. Selene Psoma has shown that the so called “Lete” coinage was the result of erroneous reading, and has now been reattributed to a city of the Thasian Peraia, probably Berge, 32 while the coinage of Ichnai has been reassigned by the same scholar and A. Zannis to an otherwise unattested homonymous city in Eastern Macedonia. 33 Although the latter theory has not been generally accepted, 34 the acquisition of Lower Paionia by the Macedonians might not be incompatible with the continuation of the Ichnaian coinage for a period of time, and the minting of coins should probably be altogether dissociated from the question of Macedonian expansion. A conquered city or region might not be forthwith incorporated into Macedonia proper. There may have been a time of transition during which Ichnai retained a semi-autonomous status compatible with the minting of

 28 Hatzopoulos 2008a, 13–20. 29 As does Zahrnt 1984, 360. 30 Edson 1970, 27 and n. 54. 31 Kremydi 2018, 286–289, with references. 32 Psoma 2006, 61–86. 33 Psoma/Ζannis 2011, 23–46. 34 ΕΚΜ II, p. 753.

The archaic period  

coinage. 35 The scholars who still insist on an early conquest of Lower Paionia and Mygdonia do so because of Herodotus (5.94.1), according to whom, when the tyrant Hippias failed to recover his former position in Athens in 505, Amyntas I offered him Anthemous to settle. They reason that Amyntas could not possess or even consider conquering Anthemous, if he did not already have Lower Paionia and, at least, western Mygdonia under his control. As Edson has rightly pointed out, 36 there is nothing cogent in this argument. Amyntas’ offer “need mean no more than a project of joint occupation” using as base of operations the Pisistratid possession in nearby Rhaikelos; and moreover nothing came of it, since Hippias declined the offer of the Macedonian king. In addition, the recent identification of Rhaikelos with the ancient site at Peraia on the coast of Anthemous, 37 which Thucydides mentions along with Krestonia and Bisaltia as part of the last stage of expansion, well after Mygdonia, further weakens the value of this dubious argument. In another passage Thucydides (2.100.4–101.5) offers an additional indication in favour of a late acquisition of Anthemous, since this region, along with (Lower) Paionia, Mygdonia and Krestonia, seems to have been part of the “New Territories” which constituted the domain of Philippos, Perdikkas II’s brother. 38 In conclusion, the identification of Aigeai with Vergina did resolve the quarrel between partisans of a geographical and of a chronological interpretation of Thucydides’ crucial passage in favour of the latter (though Nicholas Hammond, the scholar to whom this identification was due, remained paradoxically faithful to the geographical interpretation), but the question of the absolute (rather than relative) chronology of the different stages of Macedonian expansion still remained open. One might reasonably consider that after all it was not a subject susceptible of unleashing passions. This ceased to be true, and the question became flammable in the microcosm of Macedonian archaeology, after the sensational discoveries of sumptuous grave goods in the ancient cemeteries of ‘Sindos’, Hagia Paraskevi and Archontikon. For on the date of the Macedonian expansion in Lower Paionia and Western Mygdonia seemed to depend the ethnic identity of their occupants. Were they Macedonians or Paionians, Bottians and other Edonians?

 35 Hatzopoulos 1996b, 107; EKM II, p. 753. Cf. the case of Amphipolis, which continued minting its own coinage for some time after the Macedonian conquest in 357. See Picard, O. 1994, 207– 210. 36 Edson 1955, 186, n. 47. 37 Nigdelis 2011, 103–117; Apostolou 2016, 134–152; cf. BE 2012, 273 and 2017, 335. 38 Hatzopoulos 1996b, 174–176.

  The Land: Where was Macedonia? The answer to this query, in the absence of explicit written sources, raises fundamental methodological issues concerning the precise content or even the relevance of the term ‘ethnic identity’ in an archaic Greek context, the pertinence of material finds for determining language, religious beliefs and practices, institutions and customs, which are generally considered constituent elements of ethnic identity, as well as the precise content of the term ‘expansion’. Sufficiently formidable by themselves, these scientific questions have been further complicated by personal rivalries among archaeologists who were implicated in the discovery and the celebrity of the respective sites.

.. The meaning of the term ‘Macedonian expansion’ “Expansion” may imply several different realities on the ground, from politicalmilitary control to administrative integration or even mass colonisation. Possession by the Macedonian kings and colonisation by the Macedonian people are two different matters which must be clearly distinguished. For instance, in the first half of the fourth century Makedonis was both part of the Macedonian realm and was mainly inhabited by Macedonians, as we can ascertain from the names preserved in literary works and on the earliest inscriptions. In the same period Mygdonia and Anthemous were also (for most of the time) part of the Macedonian realm, but were they inhabited by Macedonians? To cite a modern example, Tunisia was simply a French protectorate until 1956. On the other hand, Algeria was integrated into France and subdivided into ‘départements’ until 1962. Conversely, New Caledonia as a ‘territoire outre-mer’ has been massively colonised to the point that the indigenous population is today a minority. Which of these realities do we have in mind when we speak of ‘Macedonian expansion’? In the modern example neither the protectorate nor the administrative integration resulted in the massive replacement of the local population. Similarly, the unquestionable Macedonian control of the New Territories beyond the Axios after the Persian Wars does not necessarily imply a concomitant massive colonisation from the cradle of the Macedonian kingdom. We must keep that in mind when discussing the ‘Macedonian identity’ of the dead buried at Archontikon, ‘Sindos’, Hagia Paraskevi, ‘Thermi’ and also Trebenishte on Lake Ochrid. Since those who uphold a ‘Macedonian expansion’ in these places in connexion with the ‘nationality’ of the dead interred there do not simply mean a political-military control of (Lower) Paionia, Mygdonia or Anthemous, but massive expulsion and colonisation, that is to say the substantial replacement of the former inhabitants by Macedonians.

The archaic period  

.. Material finds and ‘national’ identities Before examining some recent attempts at classifying the archaic necropoleis of Central Macedonia according to ethnic groups, it is necessary to present a brief overview of the main Macedonian archaic cemeteries excavated in the last decades. ... ‘Sindos’ The first of the main archaic sites to be excavated and the only one to be fully published is ‘Sindos’, named after the modern village in whose industrial zone an ancient necropolis was accidentally discovered. The results of the excavation, which lasted from 1980 to 1982, were published by the excavator Aikaterini Despoini in three magnificent volumes in 2016. 39 Of the 123 graves brought to light half had been looted in antiquity, but the goods discovered in the other half stunned specialists and general public alike by their sumptuous beauty. Gold and silver jewels, mortuary masks, gloves and shoes, bronze and iron spearheads, swords, helmets, shields, cooking and drinking utensils, miniatures of carriages and of furniture locally produced, coexisted with ceramic vessels of the highest quality imported from southern Greece. The necropolis was in use from around 560 to the middle of the fifth century. Fifty-two graves, among which were the richest ones, belonged to the archaic period (560–480), but from the beginning of the severe style period the wealth of the finds sharply declines, and is reduced to a single golden jewel from one of the few tombs of the classical period. According to the excavator, the necropolis belonged to the nearby ancient settlement of Anchialos, which she tentatively (and justifiably) identified with the ancient city of Chalastra, famous in antiquity for producing nitre from a lake situated further north 40. The question of the ethnicity of the people interred at modern ‘Sindos’ is for obvious reasons indissociably linked with that of the ancient name of the site, Anchialos. Two identifications have been proposed: Sindos and Chalastra. The modern name of the village ‘Sindos’ has not come down from antiquity. It was given to a village previously bearing the Turkish name Tekeli after the annexation  39 Despoini 1916; Despoini 2009, 20–65. See also Despoini 1985, Despoini 1993. 40 Hatzopoulos/Loukopoulou 1989, 87–92. The recent (1990–1995) excavations at Nea Philadelpheia have brought to light an important settlement whose archaic and classical cemetery has a lifespan extending from the sixth to the third century (Misailidou-Despotidou 2017, 308–309; cf. Misailidou-Despotidou 2018, 111–125). This site would make a plausible candidate for the ancient city of Klitai known from the list of theorodokoi from Delphi, Pliny and Ptolemy.

  The Land: Where was Macedonia? of most of ancient Macedonia by Greece. The ancient settlement bearing that name is mentioned only by Herodotus (7.123.3) as a coastal city of Mygdonia between Therme and Chalastra. It figures under the name of Sinthos in the Ethnika of Stephanus Byzantius referring to the same passage of Herodotus. We know, however, from another entry of Stephanus Byzantius that Hecataeus in his Description of Europe (FGrHist 1 F 147) mentioned the Sindonaioi, a “Thracian ethnos”. Unfortunately all the above landmarks are controversial. Therme has been variously located at Thessalonike, at Toumba Kalamarias, at Karabournaki, and at modern ‘Thermi’ (formerly Sedes or rather Loutra Sedes). 41 Chalastra’s location is also disputed between those who place it at the site of Gephyra (previously Topsin)/Hagios Athanasios, such as Tiverios 42 and his former students St. Gimatzidis 43 and Vasiliki Saripanidi, 44 and those who place it at ‘Sindos’/Anchialos, such as Aikaterini Despoini 45 and myself. 46 The former identification is excluded since Gephyra/Hagios Athanasios has been definitely identified with a city called Herakleia on a milestone that bears the name of that city and its distance from Dyrrhachion (in miles), which tallies with the location of the Gephyra/Hagios Athanasios archaeological site. Moreover, as we know from Strabo (7, fr. 20 and fr. 24), Chalastra was one of the 26 small cities which were included in the synoecism of Thessalonike. In addition, the onomastics on the funerary monuments which have come to light at Gephyra/Hagios Athanasios do not contain any nonGreek names, such as would be expected in a πόλις Θρηίκων (Hecataeus, FGrHist 1 F 146). On the contrary, almost all the names epigraphically attested 47 have an

 41 Vickers 1981, 327–333 (Thessalonike); Sueref 2009, 351 (Toumba Kalamarias); Tiverios 2009, 394 (Karabournaki); ATL I 398–399 and 546 (Sedes). Romaios 1940, 1–7, argued that Therme was the common name of all these settlements. For an overview and a discussion of the early hypotheses on Therme’s location, see Edson 1947, 100–104. 42 Tiverios 1991–1992, 209–234. 43 Gimatzidis 2010, 50–54, Gimatzidis 2017, 313. 44 Saripanidi 2017, 88, n. 112, with reference to Gimatzidis 2010, 50–54. 45 Despoini 2016, 12; Despoini 1993, 33–35; Despoini 2009, 30 and n. 93. More recent research at Lake Pikrolimni has verified the presence of nitre in its water, adding a decisive argument in favour of its identification with Lake Chalastra, and consequently of the identification of Chalastra with ‘Sindos’ and the nearby trapeza of Anchialos and not with the more distant site of Gephyra. See Ignatiadou 2004, 241–248; Ignatiadou et al. 2003 64–67; Ignatiadou et al. 2005, 311–312. 46 Gounaropoulou/Hatzopoulos 1985, 62–71; Hatzopoulos/Loukopoulou 1989, 87–92; Hatzo– poulos 1996b 107–108 and 171–173; BE 1996, 262. 47 Hatzopoulos/Juhel 2009, 423–437: Hadymos, Nikolaos, Kynnane, Nikanor, Hippotas; Misailidou-Despotidou 2003–2004, 64–70: Menneas (cf. BE 2007, 376); Chrysostomou, P. 1998,

The archaic period  

unmistakable Macedonian flavour, and along with the two ‘Macedonian’ tombs discovered at Hagios Athanasios strongly indicate a Macedonian foundation. As to the connexion of the Axios with Chalastra, it concerns of course its ancient riverbed discovered during the excavation at ‘Sindos’. 48 If now we espouse the most plausible hypothesis concerning the location of Therme, that its urban centre was situated at Toumba Kalamarias and that the nearby site at Karabournaki was its commercial port, then the archaeological site in central Thessalonike with its strong Ionian character 49 and its situation between the twin sites Toumba/Karabournaki and ‘Sindos’/Anchialos, could very likely be attributed to ancient Sindos, located according to Herodotus between Therme and Chalastra. 50 ... Hagia Paraskevi The excavation at the site of Hagia Paraskevi by K. Sismanidis began in 1981, while works at ‘Sindos’ were already under way. It brought to light 435 tombs dating from the second quarter of the sixth century to the second half of the fifth century, most of them inviolate and furnished with rich goods: iron and bronze weapons (swords, spearheads, daggers and helmets) for the male dead, bronze, silver and gold jewels (earrings, bracelets, rings, buckles and pins) for the female ones. In a significant number of graves, for both men and women, golden epistomia covered the mouths of the deceased. In addition to local earthenware, a great number of high-quality ceramic drinking and eating utensils imported from Attica, Corinth and Ionia accompanied the dead. According to the excavator the grave goods discovered at the Hagia Paraskevi necropolis present a “great affinity” with those of the (also archaic) ‘Sindos’ cemetery. 51 ... Aigeai (Vergina) The excavation of the archaic tombs at Aigeai was initiated by M. Andronicos and Angeliki Kottaridou in 1988. 52 It concerned two clusters of tombs. Cluster B to the south-east of the main archaic cemetery was dedicated to high-status female deceased and comprised three archaic (540–480) pit graves of great dimensions,  301–333, pl. 93–96: Menandros, Nikanor; Petsas 1967, 339–340, fig. 21: Philotera; Hatzopoulos 2008c, 246. 48 Despoini 1985, 12. 49 Petsas 1982, 66–67. 50 Hatzopoulos 1996b, 107 and 171–173. 51 Sismanidis 1993, 170. 52 Andronicos 1991, 1–3. For what follows, see Kottaridi 2016, 612–639, and Kottaridi forthcoming.

  The Land: Where was Macedonia? but also earlier, c. 600, tombs which had been destroyed. Only one tomb from c. 500 was found intact. It contained some 70 gold and silver objects (diadem, diverse pieces of jewelry, various ornaments, shoe soles and perhaps a gold-decorated sceptre). Metallic drinking and eating utensils were also deposed there. Judging from the items that grave robbers had neglected, the other two tombs displayed grave goods of a comparable sumptuousness. The earlier of them contained also imported ceramic vessels. Cluster Γ, situated to the north-east of the main archaic cemetery, was practically coeval with the Temenid dynasty, having remained in use from the beginning of the sixth century to the beginning of the third century. Among the twentyone burials, eight dated from the archaic period, six from the sixth and two from the beginning of the fifth century. All the graves were discovered looted, but the rests of the funeral pyres furnished helmets, swords, spearheads, bronze and ceramic utensils, which gave an idea of their original contents. They were most probably reserved to male members of the Temenid dynasty. The main archaic cemetery numbers some 200 burials, in contrast to the male royal cremations, almost exclusively interments in pit graves. (The privilege of cremation will be extended to female members of the royal dynasty in the late fifth century and to commoners even later, during the fourth century). In the main cemetery the deceased were buried with belongings characteristic of their sex, status and occupation. The frugality of the grave goods, when compared to those of ‘Sindos’ or even Hagia Paraskevi, is striking. There is nothing comparable to the ‘Sindos’ panoplies (and even less to the Archontikon ones, as we shall see later). There are no shields, and even the helmets are most rare. The female jewels and other outfits are made of iron, bronze and silver. Golden items, miniatures of carriages as well as spits are to be found almost exclusively in clusters B and Γ. Masks, epistomia, epophthalmia are practically unknown. ... Archontikon Excavations begun in 1992 have brought to light an important urban centre and, starting from 2000, the necropolis of this settlement, in the vicinity of the modern village of Archontikon. 53 Its prosperity was due to the exceptional fertility of its well-watered territory, particularly adapted to cattle breeding, and to its strategic position on the main east to west artery of the southern Balkan area, the future Via Egnatia. At its western necropolis 1001 tombs have been excavated, 734 of

 53 For what follows, see A. Chrysostomou/P. Chrysostomou 2009, 477–489. Cf. P. Chrysostomou 2017, 267–269; P. Chrysostomou 2018, 82–110.

The archaic period  

which belong to the archaic period, starting from c. 650; the rest date between 480 and the early Hellenistic period. There are only 11 cremations and almost all the burials are in pit graves. The absence of looting has enabled the excavators to attempt a social stratigraphy of the burials according to the relative wealth of the grave goods. At the basis of the social pyramid are male burials with only one or two spears and a knife, and female ones, which are poorly furnished. On the next level a sword is added to the male panoply, and in both male and female burials the number of jewels and of utensils increases and comprises bronze items. The third level is characterised, in the male burials, by the additional presence of helmets, gold or silver epophthalmia, gold diadems and gloves, as well as gold ornaments on helmets and swords. The female ones contain gold and silver jewels and a variety of gold ornaments originally sewed on clothing. Numerous bronze utensils and often high-quality terracotta imports accompany the deceased of both sexes. Finally, on the top level of the pyramid, eight male burials contain, in addition to the previously mentioned weapons, a bronze shield, and a golden mask or an embossed golden sheet with which the deceased of five graves had their faces covered. Three female deceased of corresponding status were buried with a gold mask and a fourth one with a silver mask. The utensils and jewels found in this category of graves were more numerous and of an even higher quality than in the burials of the previous level. What is particularly striking is that the burials of the two highest levels represent more than a quarter of the sixth-century graves, attesting by the amount of gold the extraordinary wealth, and by the number of imported wares the extent of ‘international’ connexions of the community. This remarkable prosperity suddenly disappears at the end of the sixth century. Not only are there no more golden masks and ornaments, but also the cemetery goes practically out of use for nearly a century, and is reactivated on a modest scale from the end of the fifth century until c. 275. ... ‘Thermi’ A few kilometres to the north-west of Hagia Paraskevi, near the modern village of ‘Thermi’, 54 lies the urban centre and the cemetery of a hitherto unidentified community whose lifespan extends from the eighth century to the first century A.D., although it was particularly active during the sixth, fifth and fourth centuries. Between 1987 and 2007 some six thousand tombs were excavated. Many of them

 54 Quotation marks are used to dispel the possibility of confusion with ancient Therme. The previous name of the village was Sedes. For what follows, see Skarlatidou 2009, 329–343; cf. Skarlatidou 2017, 342–343.

  The Land: Where was Macedonia? contained no grave goods or had been looted of their contents in precious metals. Enough remained, however, to provide a clear picture of its character. The archaic male burials contained weapons (spears, swords, daggers, helmets) and drinking utensils (bronze or imported ceramics), and the female ones bronze, silver and golden jewels. Both men and women were often buried with golden epistomia and all sorts of golden ornaments. The excavator, Evdokia Skarlatidou, stresses the affinity of the archaic burials of ‘Thermi’ with those of ‘Sindos’ and Nea Philadelpheia. 55 ... Nea Philadelpheia (Klitai?) The archaeological site of Nea Philadelpheia, formerly Naresh, and possibly corresponding to ancient Klitai, 56 has been known since the First World War, when it was described by Léon Rey, a French archaeologist attached to the Armée d’Orient. 57 Field research resumed in 1955 during the renovation of the railway network along the Gallikos/Echedoros valley. The excavation of 168 tombs of the archaic and classical necropolis (sixth century-beginning of the third century) revealed male burials furnished with weapons (bronze helmet, iron sword, two spears) as well as tombs in which the deceased women were buried with their jewels. Golden epistomia, as well as cereamic wares, both imported from Athens, Corinth and Ionia and locally produced, were recovered from graves of both sexes. 58 ... Aiane Excavations at Aiane starting in 1980 and vigorously pursued by Georgia Karamitrou-Mentessidi have brought to light twelve monumental chamber graves from the second half of the sixth down to the beginning of the fifth century, several smaller cist graves and a great number of pit graves. All the chamber graves were looted, but the tomb robbers had discarded magnificent commemorative sculptured markers: marble kouroi and korai, a lion, the statue of a bearded man, Ionian funerary stelae more or less well-preserved, or they had left golden ornaments there which had escaped their attention. Pit tombs which remained unlooted have preserved male burials with all sorts of weapons (bronze helmets,

 55 Skarlatidou 2009, 338. 56 Hatzopoulos/Loukopoulou 1989, 87–92. 57 Hatzopoulos/Hautefeuille 1992, 191–200. 58 Misailidou-Despotidou 2017, 308–309.

The archaic period  

bronze shields, iron swords and spearheads) and female ones with jewels in precious metals. Epistomia in gold and silver as well as metal utensils and minature carriages locally produced and imported ceramic wares from Corinth and Athens were discovered in a significant number of tombs 59. *** There have been several recent attempts to classify the archaic necropoleis of central Macedonia according to ‘ethnic identity’. In 2017 Vasiliki [Vivi] Saripanidi published an article of almost a hundred pages (73–170) dedicated to a comparative analysis of three pairs of burial sites in Northern Greece, each of which she connected with one ethnic group: southern Greek colonists (Abdera and Akanthos), Thracians (Mikro Doukato and the region of later Amphipolis) and Macedonians (Aigeai/Vergina and Archontikon). Two years later she returned to the subject with a much shorter article. 60 This time, leaving aside the cemetaries of Hagia Paraskevi, Nea Philadelpheia and Trebenishte, she concentrated on Archontikon, Vergina and ‘Sindos’, seeking thereby to affirm their fundamental similarity and the correlated early expansion of the Macedonians east of the Axios. 61 More particularly she aimed at countering the arguments of an article published in 2016 in which Angeliki Kottaridou, former assistant of M. Andronicos and director of the excavations at Vergina carried out by the Archaeological Service, had attempted a systematic comparison between the necropoleis of Aigeai/ Vergina and Archontikon, to determine the ‘ethnic idenity’ of their respective occupants. It is a question that Vasiliki Saripanidi had not raised in her earlier article, because she had taken for granted that both Vergina and Archontikon belonged to the cradle of the Temenid kingdom. Therefore, she considered that their cemeteries, taken together, could be used as the yardstick by which the cemeteries of Mygdonia on the left bank of the Axios and beyond (particularly those of ‘Sindos’ and Hagia Paraskevi), should be measured. Reasoning from the premiss that “the funerary practices that were introduced circa 560 in the region around the Thermaic Gulf to the east of the Axios did not simply share similarities with the new Macedonian practices”, but “they involved the use of the same local and imported forms of material culture, in the same configurations, the intra-site variations of which follows the same patterns”, she concluded that “it is extremely

 59 Karamitrou-Mentesidi 2017, 106–111. 60 Saripanidi 2017, 73–170; Saripanidi 2019a, 175–196, Plates 50–52; cf. Saripanidi 2019b, 381– 410 (particularly p. 383, n. 11). 61 Saripanidi 2019a, 180, misunderstands what I have written (Hatzopoulos 1996b, 171–173) about the mixed Greek-“Thracian” settlements of Western Mygdonia.

  The Land: Where was Macedonia? hard to believe that these populations, who lived within different cultural systems and whose interests were largely in conflict with those of the Macedonians, would have happily shared the latters’ cultural ideology, not least because one of the main aims of this ideology was to promote the Macedonian identity. It is even harder to believe that different funerary ideologies would have developed identical material expressions and would have produced the same archaeological patterns ... Therefore, the only possible explanation for the burials on the right [slip for “left”] bank of the Axios is that, by that time, this region formed part of the Macedonian kingdom”. 62 The problem in this – at first sight convincing – demonstration lies in a double presumption: first that Archontikon “formed part of ancient Bottia that ... is believed to be one of the first regions to be annexed to the kingdom sometime before the middle of the 6th century”; 63 second that we are really dealing with “the same local and imported forms of material culture, in the same configurations, the intra-site variations of which follows the same patterns” and with “identical material expressions and would have produced the same archaeological patterns” 64. Regarding the first question Saripanidi believes that in the archaic period Aigeai was located in Pieria and that the site of Archontikon belonged to Bottia. 65 In fact Aigeai lay in what Herodotus (7.127) called “Makedonis” and was subsequently reckoned as part of Bottia, 66 while the site of Archontikon, like Pella, situated only five kilometres to the east, belonged to the region that at the end of the sixth century Hecataeus called “the narrow strip of Paionia”. 67 Concerning now the second question, Angeliki Kottaridou in her 2016 article objected to the presumption that at Aigeai we have (to use Saripanidi’s terminology) “the same configurations”, the same patterns” or “identical material expressions” as at Archontikon, and as “in the region around the Thermaic Gulf to the east of Axios”. 68 She argued that

 62 Saripanidi, 2017, 117. 63 Saripanidi 2017, 82 with n. 65 and p. 87, n. 101. 64 The same problem can be seen in the article by Xydopoulos 2017, 72–98. Its author takes for granted that the cemeteries of Vergina and Archontikon share the same material characteristics, and that, therefore, the area of Archontikon and Pella (Lower Paionia) belonged to the Macedonian state in the Archaic Age. 65 Cf. Saripanidi 2019a, 178 and passim. 66 Hatzopoulos 1996d, 265–266. 67 See p. 14 and n. 26, above. 68 That is to say at ‘Sindos’, Nea Philadelpheia, ‘Thermi’, Hagia Paraskevi.

The archaic period  

“the customs which make their appearence at Archontikon from the middle of the sixth century: the interment of warriors with gold-ornamented panoplies, bronze shields, impressive epistomia and epophthalmia and, above all, golden mortuary masks, golden gloves and miniature furniture, find their exact parallels in nearby Mygdonia (‘Sindos’) and more distant Lychnidos (Trebenishte), probably in Amphaxitis, but not in Makedonis. And if someone argues that in 560 the Macedonians already possessed all these regions, they would then have to explain why the masters of all this wealth in their territory were so poor – with the exception of the royal clusters at Aigeai, which even in the sixth century do not stand out for their magnificence – as made amply clear by the fact that the burials of the heartland of the Temenid kingdom from Pydna to Mieza contain only few grave goods and they hardly correspond to the two ‘lowest’ levels of the west necropolis of Archontikon. And conversely, they would have to explain why at the end of the sixth century the ‘golden burials’ of Archontikon disappear and why during the whole fifth century this site declines, while at the same time the royal clusters at Aigeai are being filled with imposing funerary constructions and treasures and the necropolis of Pydna with rich tombs”. 69

Significantly Angeliki Kottaridou’s arguments were strengthened by Saripanidi herself. For instance, Saripanidi wondered: “Returning to the core area of the Macedonian kingdom, it is quite surprising that to day no mask is known from the royal capital of Vergina”. 70 She also drew attention to the fact that, whereas miniature furniture is common at ‘Sindos’ and Archontikon, “no miniature furniture has been published so far from Vergina” 71. Elsewhere she noted: “In fact both local and imported artecrafts from Sindos graves overlap with the finds from Archontiko in terms of types, provenance and modes of deposition. The same largely holds true also of Agia Paraskevi; but the types of offerings, which tend to characterise the wealthier burials at Vergina, the west cemetery of Archontiko and Sindos, either appear less frequently at this site or they are missing altogether”. 72 It is significant that only the “wealthier burials at Vergina”, that is those of the royal cluster, can compare with the in toto burials of Sindos and Archontikon. It is no accident that Saripanidi prefers to couple Archontikon with ‘Sindos’ rather than with Vergina: “Grave goods other than vases [i.e. from Sindos] display very close affinities with those from Archontiko with regard not only to their type but also to their provenance. There are golden masks and sheets for the covering of eyes and mouths of the deceased; golden and silver gilded foils from clothes, hand gloves, shoes and weapons, dress accessories and jewellery made of metals and other materials; metal cart models; bronze distaffs and clay figurines. Moreover, all men were buried with weapons, which appear in sets  69 Kottaridi 2016, 635–636. 70 Saripanidi 2017, 108. 71 Saripanidi 2017, 102, n. 226 72 Saripanidi 2017, 95.

  The Land: Where was Macedonia? similar to the ones from Archontiko. These sets may include shields, helmets, swords, spears, knives, arrows, and perhaps javelins and cuirasses, while there is also a single greave”. 73 And in the next paragraph: “Clay drinking cups, mostly imported but also local products are represented by a wide variety of types. These come from various production centres and they overlap with those known from Archontiko. The same also holds true of pouring and mixing vessels ... As is the case with Archontiko, all undisturbed furnished graves at Sindos were equipped with ‘sympotic’ shapes, which may range between a single local or imported vase and an assemblage of 22 clay and metal drinking, pouring and mixing vessels”. 74 And two pages later: “Offerings [at Hagia Paraskevi] other than vases largely coincide with those from Archontiko and Sindos, but metal and particularly golden and silver artefacts appear to be less diverse and less abundant”. 75 In her second aricle Saripanidou attempts to deal with Kottaridou’s second argument by highlighting similarities between Vergina’s and Archontikon’s archaic necropoleis, which, according to her, Kottaridou had minimised. 76 A similar attempt to refute Kottaridou’s arguments, by invoking more recent or unpublished finds from Vergina itself, the Pierians, Pydna, Beroia, Mieza, and Edessa was made by P. Chrysostomou in an article specially devoted to this subject and published in 2019. 77 It should be obvious, however, that it is extremely difficult for a non-specialist, who has not personally examined the vast and largely unpublished material from half a dozen archaic sites of Macedonia, to form an opinion about their relative affinity or dissimilarity. Given this basic difficulty, it is impossible to pronounce any verdict about the “cultural ideology” they were supposed to reflect, and even more to accept (or refute) that “one of the main aims of this ideology was to promote the Macedonian identity”. In other words, there is a limit to the questions that material finds can answer, and any attempt to go regardless beyond leads to dead ends. Ancient historians have since several decades abandoned the hope (or the illusion) that, short of written documents, the ‘ethnic’ identification of human groups is feasible by material finds alone. 78 The same observation coming from an archaeologist working for many years in Macedonia (specifically at the archaic

 73 Saripanidi 2017, 89. 74 Saripanidi 2017, 89. 75 Saripanidi 2017, 91. 76 Saripanidi 2019a. For more details, see now Kottaridi forthcoming. 77 P. Chrysostomou 2019a, 387–396; cf. P. Chrysostomou 2018, 82–109. 78 Cf. Derks/Roymans 2009, 5; Whittaker 2009, 202; Xydopoulos 2017, 77; Chemsseddoha 1919, 283–284.

The archaic period  

cemetery of Nea Philadelpheia) is most welcome. 79 “Burial customs”, writes Vasiliki Misailidou-Despotidou, “do not constitute direct evidence of political events”, because “a single cultural koine in material civilisation is attested in the region between the Strymon, the Haliakmon and the Korytsa bassin already in the Iron Age”. According to her, the presence of a similar cultural koine in the archaic period can likewise be attributed to osmoses between neighbouring population groups “who live close to one another for many centuries, share the same cultural goods brought by merchants, and eventually intermarry”. As she aptly adds, the extent of this koine from Lake Ochrid to Aiane and from Aiane to the lake of Langadas exceeds by far “the territorial possessions of the Macedonians before 480” But there is also an ambiguity in the whole previous discussion illustrated by Saripanidi’s conclusion in which she deduces the Macedonian annexation of Mygdonia from the burials on the left bank of the Axios. Burials may tell us something about the demographic composition of a given population, but practically nothing about sovereignty, which is a political phenomenon. If in fact it is not Staatsangehörigkeit (nationality) but Volkstum (ethnicity) which is at stake, then a different approach is necessary. If by Macedonian ‘nationality’, ‘ethnicity’, ‘identity’ or whatever other term is used, the subjective feeling of belonging to a group is meant, then, in the absence of written sources from the archaic period, Macedonian ‘identity’ in that period is for us inaccessible. Under such circumstances, all that we can hope to discover is identity markers usually considered as ‘objective criteria’ through which human goups are individualised. Among the ‘canonical’ ones defined by Herodotus (8.144.2) – common blood, common tongue, common religion and common customs – the first cannot be traced in material finds and the third and the fourth are common to more than one ethnic groups. Only languages are sufficiently individualised and are preserved in documentary sources, often though of later periods. It is remarkable that epigraphic evidence, the only one susceptible of attesting the language of the human groups in question, an eminent and in our case the sole available marker of identity, is conspicuous by its absence from the discussion. Although we do not have any sixth-century inscriptions from (Lower) Paionia, Mygdonia and Anthemous, epigraphic documents of the two following centuries are very informative. For if we do not encounter there any typical Macedonian personal names in the fourth century, it is extremely doubtful that they were present there earlier and subsequently disappeared.  79 Misailidou-Despotidou 2018, 110–125.

  The Land: Where was Macedonia? It is not by chance that the two earliest inscriptions from Pella concern immigrants from Ionia and from Corinth respectively, 80 and that the names inscribed thereon either retain their Ionic form 81 or are adapted to the Ionic (or Attic) dialect 82. The first typically Macedonian name in its dialectal form does not appear at Pella before the end of the fifth or the beginning of the fourth century. 83 On one mid-fourth-century deed of sale from Strepsa (modern Vasilika) are inscribed the names of Bilthes, Arnias (or Arnies), Nemenios and perhaps Myron. None of them can be considered as typically Macedonian, while Bilthes is not even Greek, but probably Bottic, since it is also found in the form Bilses in Kalindoia, a city of northern Bottike, in the first century A.D. 84 Another mid-fourthcentury deed of sale from Strepsa bears the non-specifically-Macedonian Greek names Herakleides, Herakleodoros, Myron, Lyson, Kallippos, Dionysios and the Bottic (?), non-Greek, names Tokkes, Pottes and Skittas. 85 The first of these nonGreek names is also attested in the Chalkidic city of Aphytis (fourth century), in Edonis (sixth-fifth century) and on later inscriptions from Bisaltia and Sintike; the second on a mid-fourth-century deed of sale from Stolos (?) in Chalkidike, while the third is a hapax. An early third-century royal donation from Kassandreia mentions three persons living in the territory of Strepsa: a Gouras son of Annythes, a Chionides, and a Eualkes sons of Demetrios. 86 None of these is specifically Macedonian, while the first two are not even Greek, but probably Bottic. Annythes is a hapax, while the name Gouras is also attested on a mid-fourth century deed of sale from Spartolos (see below), a third-century dedication, and an early third-century A.D. funerary inscription from Thessalonike, 87 as well as on several other inscriptions from Macedonia and beyond. 88 A similar mixture of non-Macedonian Greek and Bottic (?) names can be read on a mid-fourth-century deed of sale from Spartolos, the capital of the Bottic League: Tauriades, Polemokrates, Peison, on the one hand, and Tarbes, Sedeles, Poris, Tralis, Basses,

 80 EKM II 517 and 544. 81 EKM II 517: Ὀρθαγόρης Ἀριστοκράτεος, Ἀριστοβόλη. 82 EKM II 544: the name of the Corinthian priestess figures as Τιμαρέτη instead of Τιμαρέτα, which is the Corintian as well as the Macedonian dialectal form. 83 EKM II 514: Ἀμαδίκας. On these questions, cf. Hatzopoulos 1996b, 172. 84 Hatzopoulos 1988a, 42, n. 5; SEG 58 (2008) 663; BE 2011, 424; Hatzopoulos/Loukopoulou 1996, 234. 85 SEG 58 (2008) 662. 86 Hatzopoulos 1988a, 17–18. 87 IG X, 2, 1, 94; IG X 2, 1 S. 1407. 88 Dana 2014.

The archaic period  

Gouras, on the other. 89 Among the non-Greek names, besides Gouras (see above), Sedalas is attested in the North Bottic city of Kalindoia, 90 in Mygdonian Lete and other parts of Macedonia and the rest of the Greek world. Poris can be read on inscriptions from nearby Dikaia (first century), from the Chalkidic city of Stolos (?) (mid fourth century), and on an inscription discovered at Gyrbea in Bottia, but on behalf of a citizen hailing from Kyrrhos, also in Bottia (beginning of third century A.D.). It is also attested in Aineia by Livy (from Polybius; second century). 91 Tralis is also known from inscriptions found in the city of Anthemous (second century), and in the North-Bottic city of Kalindoia (first century A.D.), and from later inscriptions discovered elsewhere in Macedonia; there are also occurrences in other parts of the Greek world. 92 The two deeds of sale from Strepsa (Vasilika) and the royal donation from Olynthos are particularly informative, because they can be precisely dated to the middle of the fourth century (351/350?) and the beginning of the third century (285/4) respectively, and because they provide us with random specimens of personal names from the Anthemous valley. What we note on these documents is the complete absence of typical Macedonian personal names and the presence of a mixture of non-Greek names and Greek names with an Ionic flavour. Macedonian names in this region appear under particular conditions in the cities of Anthemous and Kalindoia, both of which received colonists from Macedonia in the reign of Philip II and in the beginning of the reign of Alexander III (336–334) respectively. 93 The deed of sale from Spartolos provides a useful point of comparison, because it shows that the onomastic situation which we observe at Strepsa represented the usual pattern in the western part of the Chalkidike peninsula. Now, Anthemous with Strepsa had been Macedonian since 479 at the latest. In the reign of Perdikkas II it was part of the arche that the king had entrusted to his brother Philippos. They must have become Chalkidian for a brief period (383– 379), but were recovered by Amyntas III after the defeat and the dissolution of the Chalkidian League. In 370–368 they were captured along with Therme by the pretender Pausanias, who was soon driven back by Iphikrates, 94 but managed to maintain himself at Kalindoia until at least 360. Not long after Philip II reannexed Anthemous and most probably Strepsa, but ceded them both to the Chalkidians

 89 BE 1997, 402; SEG 46 (1996) 804. 90 Dana 2014, 298–299. 91 Dana 2014, 275. For the inscription from Gyrbea in particular, see now EKM II 428. 92 Dana 2014, 378–379. 93 See Hatzopoulos 1996b, 193–195, with references. 94 For the date, see Hatzopoulos 1985–1986, 37–58.

  The Land: Where was Macedonia? in 356. Finally, he recovered the two cities after the defeat and dissolution of the Chalkidic League in 348. 95 The decisive proof that Anthemous was not inhabited by Macedonians until then is provided by the fact that Philip II ceded this region to the Chalkidians in 356, which would be unthinkable, if a Macedonian population had been settled there. 96 Unfortunately, we do not dispose of equivalent pre-Hellenistic epigraphic documents from western Mygdonia. Nonetheless, in one early-fourth-century letter painted in Attic or in Ionic dialect on a pyxis from Chalastra (modern Anchialos) we read the names of Eudikos and probably Menestratos, which are not typically Macedonian. 97 Another fourth-century funerary inscription from Nea Philadelpheia bears the names of Hegesippos and of his father Epikerdes, neither of which can be considered as typically Macedonian. 98 To conclude, the close affinities between the archaic necropoleis of Archontikon, ‘Sindos’, Nea Philadelpheia, ‘Thermi’ and Hagia Paraskevi, on the one hand, and the divergences they display in comparison with the corresponding necropolis of Vergina, on the other, can be verified, and they do not argue in favour of considering the latter as belonging to the same undifferentiated human group. The rapid decline in wealth of the former five necropoleis, in contrast with the soaring fortunes of the necropolis of Vergina from the end of the archaic period onwards, tells a story of transfer of wealth usually connected with conquest. Whereas most of the fourth-century inscriptions from Vergina bear Greek names that are typically Macedonian, inscriptions from the same period from the Anthemous region display a mixed Ionic Greek and non-Greek Bottic (?) anthroponymy. Such evidence combined with the fact that Philip II – and probably Amyntas III as well before him – ceded this region to the Chalkidians leaves no doubt that at that time (and presumably also previously in the archaic period) the population of the Anthemous valley, where the necropoleis of Hagia Paraskevi and ‘Thermi’ are situated, was not Macedonian. All in all, a late (late sixth century at the earliest) date for the expansion of the Temenid kingdom in (Lower) Paionia, Mygdonia and beyond seems more likely than an earlier one, while the inception of population transfers from the Old Kingdom towards regions beyond the head of the

 95 For the tribulations of Anthemous, see Zahrnt 1971, 152–154. 96 Cf. Hatzopoulos 1996b, 174. 97 Tiverios 1991–1992, 209–234. 98 Despotidou-Missailidou 2003–2004, 61–70.

The classical and Hellenistic periods  

Thermaic Gulf (Anthemous and eastern Mygdonia) should be dated even later (after 348). 99

. The classical and Hellenistic periods The earliest information concerning the borders of the Macedonian kingdom after 479 is found in Herodotus (5.17.2) concerning the Persian embassy to Amyntas I in c. 510: “There is a very short route from Lake Prasias to Macedonia. First, next to the lake lies the mine from which later one silver talent accrued to Alexander every day; after the mine, if one crosses over the mountain called Dysoron, he is already in Macedonia”. 100 As Louisa Loukopoulou and I explained some years ago, we have to distinguish between three chronological moments in this narrative: that of the event described c. 510, that of an indeterminate later point of time, when Amyntas’ successor, Alexander I, exploited the mine near Lake Prasias, and an even later date when Herodotus was writing. By that time, Macedonia had lost control of the basin of Lake Prasias and of the silver mine, and its border had been moved some distance westwards on Mount Dysoron. 101 What bedevilled the understanding of this passage for decades was the identification of Mount Dysoron with Mount Kerdylion (former Krousia) and of Lake Prasias with Lake Kerkine (former Boutkovou). 102 The first to challenge it was L. Missitzis in a revised edition of the famous report of the Philippian ambassadors to their home city concerning Alexander the Great’s decisions. 103 He suggested that Mount Dysoron should be identified with Mount Menoikion, and Lake Prasias with the early modern Lake of Pravi, the ancient Lake of Datos, that is to say the marshes of Philippoi. His suggestion was mentioned with a question mark by Ε. Ν. Borza at the end of an article published four years later; 104 rejected by Hammond in two

 99 The only certainly attested exception is Amphipolis, which, because of its crucial startegic importance, received Macedonian colonists immediately after its conquest by Philip II in 357: Hatzopoulos 1991, 80–86. 100 Ἔστι δὲ ἐκ τῆς Πρασιάδος λίμνης σύντομος κάρτα ἐς τὴν Μακεδονίην. πρῶτα μὲν γὰρ ἔχεται τῆς λίμνης τὸ μέταλλον ἐξ οὗ ὕστερον τούτων τάλαντον ἀργυρίου Ἀλεξάνδρῳ ἡμέρας ἐφοίτα, μετὰ δὲ τὸ μέταλλον Δύσωρον καλεόμενον ὄρος ὑπερβάντα εἶναι ἐν Μακεδονίῃ. 101 Hatzopoulos/Loukopoulou 1992, 20–21. 102 Cf. Cassson 1926, 62–63, with earlier bibliography; Geyer 1930, 46; Kanatsoulis 1964, 86; Hammond 1972, 194; Hammond in Hammond/Griffith, 1979, 58; Borza 19922, 46–48; 53–55; Hatzopoulos/Loukopoulou 1992, 15–25, but see now Hatzopoulos 2008a 13–54. 103 Missitzis 1985, 1–14. 104 Borza 1989, 60–67.

  The Land: Where was Macedonia? articles published in 1988 and 1997 respectively; 105 and, ill-advisedly, criticised indirectly by myself. 106 It was only after M. Faraguna’s thorough article of 1998 107 that I was convinced of the soundness of Missitzis’s suggestion. 108 The next glimpse we get on the situation at the eastern marches of the Temenid kingdom under Alexander I dates from 456 and comes from Plutarch’s Life of Kimon (14). According to that, although it would have been possible for Kimon after the conquest of the Thasian Peraia to attack Macedonia and to seize a significant part of the country, he abstained, and was accused of having been bought off by Alexander. 109 Alexander was succeeded by Perdikkas II in c. 454. If at that time the Macedonian kings still had the Strymon valley under their control, this favourable situation was not going to last long. In 451 Perikles was able to settle one thousand Athenian klerouchoi in Bisaltia (Plutarch, Life of Pericles 11). The result is clearly reflected in the aparchai catalogues during the immediately following years: Argilos and Hedrolos/Arrolos on the right bank of the Strymon, and Berge on the left bank become tribute paying members of the Athenian empire: coastal Argilos in 454/3, 110 continental Berge 111 and Hedrolos/Arrolos in 452/1 and 450/49 respectively. 112 The whole lower Strymon valley was lost for Macedonia. Perdikkas reacted by appointing his brother Philippos ‘Lieutenant General’ in the New Territories from the Axios valley eastwards, with the mission to defend the Macedonian positions on the eastern marches of the kingdom, and, if possible, to recover what had been lost. By 432 it was clear that Philippos had utterly failed. Not only was he unable to retrieve the former Macedonian possessions in Bisaltia, but he was not even able to stop the Athenians from shutting completely the Macedonian kingdom from the mouth of the Strymon by the foundation of Amphipolis  105 Hammond 1988, 382–391 (= Collected Studies III 211–220); Hammond 1997b, 41–45. 106 Hatzopoulos/Loukopoulou, 1992, 19–20. 107 Faraguna 1998, 349–395, who, however, on p. 389 ascribes to me the unfounded theory that when royal land was made over to Macedonians for the foundation or refoundation of a city, it ‘was transferred to an administration distinct and separate from the king’s authority under the authority of the assembly of Macedonians’. 108 Hatzopoulos, BE 2000, 436; Hatzopoulos 2008a 13–54. The objections of Vasilev 2015, 91– 107, who prefers to identify Lake Prasias with the unattested (and probably non existant) in antiquity Lake of Butkovo, do not carry conviction. 109 On Persians, Parians (Thasians), Athenians and Macedonians in the Strymon valley before the foundation of Amphipolis, see the excellent in-depth study by Mari 2014, 53–114. For what follows, see now Hatzopoulos 2020. 110 Flensted-Jensen 2004, 820–821, no 554 111 Loukopoulou 2004, 858–859, no 628. 112 Hatzopoulos 2010a, 229–236 (= Recueil 251–258).

The classical and Hellenistic periods  

opposite Argilos in 437/6. 113 It would not be long before the Athenians would cut off the Macedonians completely from the Stymonic Gulf by the annexation of Bormiskos, probably in 425/4 114 and of Tragilos before 422/1. 115 The situation was not much better on the western marches of the kingdom. From the three “allied and subordinate” kinglets of Upper Macedonia, Derdas of Elimeia, Arrhabaios of Lynkos and Antiochos of Orestis, mentioned in the prewar treaty of the Athenians with Perdikkas (IG I3 89), the first two turned against the king of Lower Macedonia (Thuc. 1.57.3; 4.83.1) and the third one seemed absorbed by events taking place on the western side of the Pindos chain (Thuc. 2.80.6). In fact, Temenid rulers did not hold sway beyond ancient Arnisa, modern Vegora, on the southern shore of Lake Begorritis. 116 During the Peloponnesian War, Macedonia suffered further losses when the Athenians stormed Therme in 432 (Thuc. 1.61.2), and annexed Herakleion in Pieria to the Athenian empire, probably in 425/4. 117 But they held neither of them for long. Therme was returned to Perdikkas in 431 (Thuc 2.29.6), and Herakleion appears for the last time in the Athenian assessment of 421. The panorama of the Temenid realm offered by Thucydides (2.99) reflects conditions in the reign of Archelaos, probably after the end of the Peloponnesian War. The eclipse of Athens allowed Macedonians to recover the whole western coast of the Thermaic Gulf, with the exception of Methone, Mygdonia at least as far as the Strymon river, and Bisaltia; but Argilos, Tragilos, Berge, and, most importantly, Amphipolis remained independent. On the western front, Archelaos improved relations with the Elimiotai by a dynastic marriage of his elder daughter to the son of king Derdas, which enabled him to acquire an ally against “Sirrhas and Arrhabaios”, apparently the guardian of the king of Lynkos and his underage ward. 118 Archelaos’ demise in 399 opened a forty-year period of instability during which Macedonian kings had the greatest trouble in preserving the integrity of their realm. On the western front Amyntas III’s second marriage with Eurydika, daughter of Sirrhas, put an end to the desires of Lynkos for independence, but from then on, along with Lynkos, the Temenid kingdom had to assume the burden of defending Macedonia against Illyrian aggression. Amyntas was soon so

 113 Flensted-Jensen 2004, 819–820, no 553. 114 Flensted-Jensen 2004, 817, no 547. 115 Flensted-Jensen 2004, 821, no 555. 116 Hammond/Hatzopoulos 1982, 128–149. 117 Hatzopoulos/Paschidis 2004, 802, no 537. 118 Arist. Pol. 1311b with Hammond in Hammond/Griffith 1979, 139.

  The Land: Where was Macedonia? hard pressed by the Illyrians that he had to consent to an unequal treaty with the Chalkidians, according to which he temporarily ceded borderland (Anthemous?) to the League. Their help was not sufficient though to save his throne, and he had to be reinstated by the Thessalians a couple of years later (Diod. 14.92.3-4; 15.19.2–3). The Illyrians would become a recurrent threat for Macedonia under Amyntas III’s sons and successors Alexander II, Perdikkas III and Philip II. Elimeia, on the other hand, remained independent, and its king Derdas II (grandson? of Derdas I, active during the Peloponnesian War) was able to deal as an equal with the Temenid king Amyntas III, when the latter had to face the aggressive Chalkidians in 382–379 (Xen. Hell. 5.2–3). He was also the only Elimiote king to issue his own coinage between 390 and 375. 119 There is however reason to believe that soon after that date Amyntas III acted as the sovereign of both litigants in the notorious boundary settlement between Elimeia and Doliche (AE 1923, pp. 161–162, no 386f). 120 The bond of Elimeia with the Argead kingdom was probably strengthened in the sixties of the fourth century by another dynastic marriage, that of Amyntas’ son Perdikkas III with “Phila, the sister of Derdas and Machatas” (Athen. Deipn.13.557c). 121 The main danger for the integrity of the realm came from the East and the expansionism of the Chalkidic League. Not only did the Chalkidians refuse to return the temporarily ceded territories, but also began annexing cities of Mygdonia, and even Pella beyond the Axios (Xen. Hell. 5.2.11–24; 37–39; 3.18–20; 3.26; Diod. 15.23.3). Eventually they were driven back and their League was dissolved in 379 thanks to the decisive intervention of Sparta, but Amyntas never recovered parts of his former possessions: Apollonia by Lake Bolbe and Arethousa on the Strymonic Gulf, which had taken advantage of Amyntas’ troubles in order to secede from the kingdom. They both retained their independence, while Kalindoia was captured by Pausanias, probably a son of Archelaos. 122 In fact, Mygdonia east of Mount Kissos was lost for Macedonia. Soon after Amyntas’ death in 370 Anthemous and the greater part of Western Mygdonia was also temporarily lost when Pausanias, probably with Chalkidian support, captured Anthemous, Strepsa and Therme, and threatened Pella. The Athenian general Iphikrates was able to push him back and to save Alexander II’s mother and brothers,  119 Liampi 1998, 5–11. 120 For a detailed commentary, see Lucas 1997, 101–108 and 211–219, who argues that Amyntas was able to annex Tripolis of Perrhaibia soon after 379. Contra, Liampi 1990, 11–22, who believes that at roughly that time Tripolis issued its own autonomous coinage. Lucas, on the contrary, assigns this coinage to the Molossian Tripolis. 121 See below in the discussion of Philip II’s wives. 122 Zahrnt 1971, 155–158; 160–161; 191–193.

The classical and Hellenistic periods  

but Pausanias survived entrenched at Kalindoia, 123 while Methone and Pydna were seized by the Athenian general Timotheos in 364. 124 When in 360, after the assassination of Alexander II in 368/7 and the death of Perdikkas III in battle against the Illyrians, Philip acceded to the Macedonian throne, the kingdom was in a shambolic state. The Illyrians were occupying Upper Macedonia, the Paionians were sweeping down the Axios Valley, Pausanias from Kalindoia threatened the eastern marches of the kingdom and claimed once again the Temenid throne with Thracian support, and, finally, Argaios, another pretender, marched to Aigeai with a mercenary force supplied by the Athenians. Philip played the only good card that he still retained, a Macedonian garrison introduced by his brother Perdikkas III into Amphipolis, which he removed as a good-will gesture to Athens and started the brilliant ‘reconquista’ of his kingdom. He defeated Argaios, got rid of Pausanias and the Paionians, recovered Upper Macedonia from the Illyrians, recaptured Amphipolis, and tricked the Athenians out of Pydna. For the first time in a hundred years Macedonia had completely recovered the Strymon valley. However, he did not stop there. He occupied Krenides and founded a new city, Philippoi, naming it after himself, and thus regained possession of the mines of Mount Pangaion, which his ancestor Alexander I had lost a century before. Finally, he took Methone, an old Eretrian colony which had become an Athenian possession. Thus, in only six years he had fully restored the Macedonian realm at its greatest extent. It stretched from Lake Ochrid and the twin Lakes Prespa in Upper Macedonia to the valley of the Strymon, and from the Tempe pass to the Gorges of the Axios (Demir Kapija). The western coast of the Thermaic Gulf, from Herakleion to Therme and beyond, was exclusively Macedonian. Nevertheless, Philip did not hesitate to alienate momentarily Anthemous to obtain the favour and alliance of the Chalkidic League (Diod. 16.8.4–5; Dem. 6.20; Liban. Hypoth. Dem. 1.2). He was probably already planning to take it back. The opportunity arose in 349–348, when he defeated the Chalkidian League, and also conquered not only Anthemous, but also the whole of the Chalkidike peninsula. Macedonia proper (i.e. excluding the external dependencies) had practically acquired its definitive contour that would stay essentially unaltered until the Roman conquest. There remained the formidable task of transforming this mosaic of peoples and heterogeneous political units into a single people and a single state.

 123 Hatzopoulos 1985–1986, 37–58. 124 Dinarch. 1.14; cf. Demosth. 4.4. Cf. Hammond in Hammond/Griffith 1979, 186.

  The Land: Where was Macedonia?

. National territory, royal land and allied cities So far we have been dealing with the expansion of the Macedonian realm. But this chronicle of conquest leaves unanswered the question of the legal status of the conquered lands and the legal status of their inhabitants. The first question has not attracted much interest, while the second one, which has been diversely answered, will be discussed later. I have devoted to the question of the legal status of the conquered lands a chapter of some fifty pages in another work. 125 It dealt in particular with the incorporation into Macedonia of Southern Paionia, Western Mygdonia and Krestonia, Pydna and Methone, a matter that has either been discussed above 126 or does not raise any serious problems in the present context, since nobody doubts that these regions and cities became integral parts of Macedonia proper before or during Philip’s reign. It also dealt with Philip’s annexation of Amphipolis and Eastern Macedonia. My conclusion based on literary, epigraphic and numismatic evidence was that the Strymon valley as far east as Mount Dysoron (Menoikion) and Mount Pangaion, along with the Pierian coast up to Oisyme, were also incorporated into Macedonia proper, but that the basin of Philippoi with its port Neapolis (modern Kavala) remained outside it as independent entities allied to the Macedonian king, at least until the very last years of Macedonian independence. The same may have been true of some parts of the Strymon valley where a few Hellenised cities (Gazoros, Berge, Sirrhai, Arrolos, Skotoussa, Herakleia Sintike) coexisted with indigenous villages of Bisaltians, Odomantes and other Sintians. 127 Finally, it examined the Macedonian annexation of Eastern Mygdonia and the Chalkidike Peninsula, which was a gradual affair, particularly under the last Temenids, Kassandros and Antigonos Gonatas. In Eastern Mygdonia Argilos was annexed by Amphipolis, while Apollonia and Arethousa became Macedonian cities before the end of the fourth century, and lesser cities in that area were incorporated into Apollonia or into Thessalonike, Kassandros’ new foundation. 128 In Anthemous the homonymous urban centre became a Macedonian city, while Strepsa lost its polis status, and its territory was converted into royal land and eventually into rural estates, some of which were awarded to royal Friends and were later incorporated into the civic territory of Kassandreia. Further east, the  125 Hatzopoulos 1996b, 167–216. What I wrote about the Pentapolis on pages 58–63 and 216 is partly obsolete after the new readings of an important inscription by Zannis 2014, 159–160 and 507–512. 126 See pp. 11–32, above. 127 Hatzopoulos 1996b, 184; 247, n. 8; cf. Livy 44.45.8; 45.30.3. 128 Hatzopoulos 2005, 201–212 (= Recueil 239–250).

National territory, royal land and allied cities  

Northern Bottike cities of Kalindoia, Kamakai, Tripoai and Thamiskos (?) ceased to be poleis, and their territories were converted into royal land, which eventually was donated to the Macedonians for the foundation of the new Macedonian city of Kalindoia. In Krousis, although Aineia survived as a settlement, it was incorporated into the territory of Thessalonike, while most of the rest of Krousis became the territory of the new city Antigoneia (Psaphara) probably founded by Antigonos Gonatas. 129 Further south in Bottike, at Nea Kallikrateia, the Eretrian colony of Dikaia ceased to exist as a polis and must have been eventually incorporated into Kassandros’ new foundation of Kassandreia. 130 Indeed, southern Bottike initially shared the fate of Strepsa in Anthemous. Its cities such as Sinos and Spartolos were reduced to royal land and in their territories estates were carved out for the benefit of royal Friends. These estates too were eventually included in the civic territory of Kassandreia. The same fate befell the Chalkidian cities of Olynthos and Sermylia and the ancient Corinthian colony of Poteidaia, on whose site the new city of Kassandreia was founded. We do not know much about what happened farther east, except that Stageira was destroyed by Philip and then rebuilt either by Philip himself or his son Alexander. Later it was incorporated into the new Antigonid foundation of Stratonikeia. The extreme south and south-eastern regions of the Chalkidic peninsula seem to have been spared in terms of both destruction and Macedonian colonisation. This appears to be the case of Akanthos, and of the cities of Akte (Athos), which had not been part of the Chalkidic League, and of Mende, Skione, Aphytis, and even Torone, which had resisted Philip. Gradually, they were included into Macedonia proper. We get a glimpse of the process in a deed of sale from Torone. The legal formulation of the act is that of the old Chalkidic League, the dialect and the name of the month are still Ionic, while the names of the contracting parties, the neighbour, the guarantor and the witnesses, do not betray the presence of Macedonian colonists from the Old Kingdom; but the eponymous priest is no longer that of the (by then extinct) Chalkidic League, but the priest of Asklepios, as in all the other Macedonian cities. 131

 129 Papazoglou 1988b, 419–421. 130 Biliouka/Graikos 2005, 381–389. 131 SEG 62 (2012) 479, with BE 2017, 337. Cf. the parallel case of a deed of sale from Amphipolis displaying the same characteristics of a gradual transition from a colonial Greek city to a Macedonian one (traditional legal formulary, but with invocation of God and of Good Fortune, Ionic names, Ionic month, but with the addition of the priest of Asklepios): Hatzopoulos 1991, 24–28, no III.

  The Land: Where was Macedonia? It is clear from the above that the foundation of Kassandreia decisively simplified the map of most of the Chalkidike peninsula. 132 It prepared but did not actually effect its incorporation into Macedonia proper, since Kassandreia, like Philippoi before, in contrast to the other foundation of Kassandros, Thessalonike, was not founded as a Macedonian city, but as an independent Greek polis, allied to Macedonia. Its magistrates had a different nomenclature, and it had its own eponymous priest and its own calendar. It even sometimes followed its own policies, and that proved fatal to its particular status. When in 276 Antigonos Gonatas conquered the rebellious city, he incorporated it into Macedonia proper. The nomenclature of the magistrates was maintained, but the Macedonian calendar replaced the former one. Moreover, as already stated, Antigonos Gonatas founded two new cities Antigoneia in western and Stratonikeia in eastern Chalkidike. The first comprised the territory of Krousis between the territories of Thessalonike and Kassandreia; the second, the lands north of Akanthos, probably including Stageira. The foundation was accompanied by the settlement of Macedonian colonists and the adoption of a Macedonian ‘constitution’. Thus from the beginning of our relevant documentation in the fifth century down to the Roman conquest of Macedonia the possessions of the Macedonian kings exceeded the lands inhabited by Macedonian citizens: they also included subject or ‘allied’ cities and peoples, and spear-won royal land. Royal land was partly retained (but in trust) 133 and exploited directly by the kings, partly rented to entrepreneurs or granted to individuals on a more or less permanent basis. When the grant was expressly hereditary and the grantee was (or happened to become) the citizen of a city whose territory was contiguous with the land granted, royal land could be converted into privately owned civic land. Nevertheless, the Crown remained the ultimate owner in default of heirs. 134 The arche, the regnum of a Macedonian king 135 included both Macedonia and ‘allied’ (in fact subject) cities, regions and peoples, such as Demetrias, the Dolopes or Thrace beyond the Strymon basin, but the civitates Macedonum or he chora he Makedonon designated Macedonia proper composed of cities belonging to Macedonian citizens. This Macedonia proper did not include Thessaly of the four tetrades nor Thrace between the Nestos and the Hebros, nor Macedonian Illyria with cities  132 The contemporary foundation of Ouranopolis by Kassandros’ brother Anaxarchos on the territory of the Andrian colony of Sane did not have any comparable effect. 133 Cf. Arr. Anab. 7.9.9: ὅσα ἕνεκα ὑμῶν φυλάττεται. 134 See Tziafalias/Helly 2010, 86–93, with my comments BE 2011, 399. 135 Cf. IG I3 89: ἐπὶ πόλιν οὐδεμ[ία]ν hōν Π[ερδίκκας κρατεῖ]; Hatzopoulos 1996c, 20, no 1: [καὶ ἐάν τις ἴηι ἐπ’ Ἀμ]ύνταν [ἐς | τὴν χώρην ἐπὶ πολέμοι ...]; SEG 57 (2007) 576, l. 26–27: ἐκ τῆς χώρης πάσης ἧς ἐπάρχει Π[ερ]δίκκας.

Some particular cases  

such as Lychnidos or Antipatreia, not even Paionia, although this, as well as some other cases, require further clarification.

. Some particular cases .. Paionia We know the names of at least seven Paionian kings who reigned from 359/8 down to c. 230. During that period Paionia, whether completely independent (and even hostile) or allied and in fact subject to Macedonia, was a clearly distinct entity. It was only in c. 227 that Antigonos Doson conquered the southern part of Paionia and founded the city of Antigoneia in the middle Axios valley near the gorge of the Axios (Demir Kapija). Ten years later, Philip V completed the conquest of Paionia capturing Bylazora, the ancient seat of the Paionian kings, to ward off Dardanian incursions. For the next fifty years Paionia remained Macedonian, but not part of Macedonia proper. It was in these years that Perseis and most probably a third Macedonian city, Astraion, were founded. Nevertheless, Paionia remained a distinct strategia, Paionians served in distinct army units, 136 and the citizens of Paionian cities were not Macedonians but Paionians. 137 It is only after the Roman settlement of Macedonia in 167 that Paionia was partitioned between the Second and the Fourth Meris, and was integrated into Macedonia (Livy 45.29). 138 .. Perrhaibian Tripolis Perrhaibian Tripolis, consisting of the cities Pythion, Doliche and Azoros, was attached to Elimeia, but retained nevertheless its particular status. 139 The exact date of its annexation is not known, but it had taken place before the middle of the third century at the latest, since the Delphians appointed Philarchos son of Hellanion, a “Macedonian Elimiote from Pythion” as their proxenos in c. 254– 253. 140 In fact, it must have taken place much earlier. The catalogue of the theorodokoi of Epidauros mentions in its Macedonian section Bouplagos and Ek-

 136 Merker 1965, 35–54. 137 Cf. Petrakos 1968, 110, no 77: Νικάνωρ Βιαίου Εὐδαρισταῖος Παίων; cf. BE 1969, 456. 138 Papazoglou 1988b, 68–71; Sakellariou 1983, map on p. 198. 139 Cf. note 120, above. 140 FD III 4, 417 III L. 14–15 et pour la date Lefèvre 1995, 190; cf. Lucas 1997, 217.

  The Land: Where was Macedonia? phantos as the two theorodokoi at Pythion. The original compilation of the catalogue dates from c. 360, but Pythion figures in an addition to it probably dating from the end of the fourth century. 141 Azoros, Doliche and Pythion are absent from a list of Perrhaibian cities of the first half of the fourth century, because apparently they were already attached to Macedonia. 142 The settlement of the territorial dispute between Doliche and Elimeia by Amyntas III confirms the annextion of the Perrhaibian Tripolis to Macedonia before 370/369 at the latest. 143 The particular status of Tripolis within Elimeia is confirmed by the letters dated to October 222 of Antigonos Doson “to Megalokles and to the koinon of the Tripolitai”. 144 The title of Megalokles is not mentioned, but judging from the evidence of a late third century inscription from Samothrace, it was most probably strategos. 145 Tripolis remained Macedonian until 196, when Flamininus restored the Perrhaibian koinon (Pol. 18.47.6; Livy 33.32.5; Plut. Flam. 10). 146

.. Magnesia An inscription from Kos reproducing a decree from 243 of Homolion which granted inviolability to the Asklepieion of Kos and which is dated by the current priest of Asklepios and the current epistates shows that this city and probably the major part of Magnesia (with the possible or probable exception of Demetrias, which uses a particular calendar, different from the Macedonian one) 147 had been annexed to Macedonia. That would explain the extent of royal land in this perioikic territory. 148 Magnesia remained attached to Macedonia until 196, when it became an independent federation. It was recovered by Philip V in 191 (Livy 39.23.12) and was definitely lost in 168 after the defeat of Perseus at the battle of Pydna and the reconstitution of the Magnesian federation, for which there is ample epigraphic evidence. 149

 141 IG IV 1, 2, 94 Ib; cf. Lucas 1997, 81–82, no 33. 142 Helly 1979, 165–200. 143 See p. 36 and n. 120, above. 144 SEG 60 (2010) 586; cf. Hatzopoulos, BE 2011, 399. 145 IG XII 8, 178; cf. Lucas 1997, 84–85, no 37. 146 Cf. Lucas 1997, 218–219. 147 Hatzopoulos 1996b, 155; 158. Contra Intzesiloglou 2006, 69, n. 3. 148 Batziou-Efstathiou/Pikoulas 2006, 79–89. 149 Cf. IG IX 2 1101; 1102; 1103; 1104; 1109.

Some particular cases  

.. Tymphaia, Parauaia, Atintania The long-drawn controversy on the inclusion or not of Tymphaia, Parauaia and Atintania within Macedonia has stumbled on the problem of situating these regions on a map, which remained for long unsolved. Fortunately, there is some degree of consensus on one of the three ancient toponyms, which can provide a stable point of reference for locating the other two. Most people agree that Tymphaia must be sought on the highlands “southwest of the bend of the river Haliacmon”. 150 Its clearest geographical definition is given by Hammond: “In the south Tymphaea reached the headwaters of the Penaeus (Str. 7.7.9) and bordered Aethicia, a region immediately south of the river; in the east it came up to and perhaps included Grevena, the modern capital of this region; in the west it marched with Parauaea; and in the north its limit is uncertain”. 151 Its inclusion within Macedonia proper, at least since the reign of Philip II, is evidenced by the presence of a Tymphaian taxis [brigade] under Polyperchon in the expeditionary corps of Alexander III (Diod. 17.57.2) and by the inclusion of Attalos son of Andromenes, expressly identified as Tymphaian, among the Macedonians (Arr. Ind. 18.6–7). The fact that Tymphaia was annexed by Pyrrhos (Plut. Pyrrhos 6.4) was of no consequence, because this province was soon recovered by Macedonia. The location of Parauaia is rendered difficult by the dearth of relative sources. Besides the above passage from Plutarch, it is only mentioned by Thucydides, probably Strabo (after Theopompus) and Arrian. Thucydides (2.80.6–7) enumerating the forces which in Epirus sided with the Lakedaimonian expeditionary corps under Knemos in 430/29, names the Molossoi, the Atintanes and the Parauaioi together in one and the same sentence, and specifies that the Orestai campaigned together with the Parauaioi under Oroidos, the king of the latter. Strabo (7.7.8 Baladié) presents a list of Epirotic peoples in the following order: Amphilochoi, Molossoi, Athamanes, Aithikes, Tymphaioi, Orestai, Paroraioi (an error for Parauaioi according to Papazoglou) 152 and Atintanes. Finally, Arrian (Anab. 1.7.5) describes the route of Alexander III from Pelion in Illyria to Pelinna in Thessaly in the following manner: “He marched, therefore, past Eordaea and

 150 Woodward in Wace/Woodward 1911–1912, 181–182; cf. E. Oberhummer in RE VII 1754–1756, s.v. Tymphe; Lévêque 1957, 127, n. 5; Hammond 1967, 680–682; Hammond 1979, 25; Hammond 1980, 175 (= Collected Studies 49); Bosworth 1981, 91, n. 22; Papazoglou 1988b, 230, with references. 151 Hammond 1980, 175 (= Collected Studies 49). 152 Cabanes 1976, 130; Papazoglou 1988b, 230.

  The Land: Where was Macedonia? Elimiotis, and the borders of Stymphaea and Paravaea, and on the seventh day reached Pelinna in Thessaly” (translation by P.A. Brunt in the Loeb series, modified). Wace placed Parauaia in the upper valley of the Haliakmon, 153 E. Kirsten between Grevena and the sources of the Aoos; 154 Lévêque sided with Wace in the note he devoted to the question, 155 but put it in the middle valley of the Aoos on his map; 156 Hammond in “the area of Permet and Leskoviq (Danglli and Kolonië)”, that is to say also in the middle valley of the Aoos, 157 which obliged him to change Parauaia into Perrhaibia in the passage of Arrian’s Anabasis, 158 but in his 1980 article on the march of Alexander the Great on Thebes he retained the reading Parauaia at the price of locating a second Eordaia in the area around Lake Little Prespa; 159 Bosworth suggested that we should treat Tymphaia and Parauaia as a single georaphical entity; 160 finally Papazoglou located Parauaia “on the territories on both slopes of Mount Pindos with its centre in the upper valley of the Aoos, to which it owned its name”. 161 Taking into consideration that in the passage from Thucydides the Parauaioi are directly coupled with the Orestai and connected more loosely with the Molossoi and the Atintanes, that in Strabo they are mentioned between the Orestai and the Atintanes, and not far from the Tymphaioi, and that in Plutarch and Arrian Parauaia and Tymphaia are closely linked, it is reasonable to place Parauaia between Tymphaia and Orestis on the east, Molossis on the south-west, and Atintania on the north-west, that is to say on the upper valley of the Aoos and astride Mount Lynkon. Now, Alexander’s march to Thessaly becomes perfectly understandable: “From the Lyncestian plain he crossed the Kirli Derven pass to Eordaia and moved southwards to the modern town of Kozani. From there he probably passed through the Siatista gap into the Haliacmon valley and the ancient canton of Elimiotis. Alexander then penetrated Tymphaea somewhere near the modern crossroads at Grevena”. 162 “It is better to suppose that Tymphaea and Parauaea are regarded as a single geographical entity. If Alexander drove south,  153 Wace in Wace/Woodward 1911–1912, 181–182. 154 Kirsten in Philippson/Kirsten 1956, 76, n. 1. 155 Lévêque 1957 127, n. 5–6. 156 Lévêque 1957, pl. II. 157 Hammond 1967, 680. 158 Hammond 1967, 680 n. 4; Hammond 1972, 109, n.2; 118. 159 Hammond 1980, 174 (= Collected Studies III 48). 160 Bosworth 1981, 91, n. 22. 161 Papazoglou 1988b, 231–232. 162 Bosworth 1981, 96.

Some particular cases  

as Hammond supposes, from the area of modern Grevena, his route to the headwaters of the Peneus would have skirted the spine of the Pindus dividing Tymphaea from Parauaea. That is all that is implied by Arrian’s vague wording”. 163 I would add that in my view Tymphaia and Parauaia did not form only a single geographical, but also a single administrative unit. That would explain why Parauaia is often subsumed in Tymphaia in the works of Alexander the Great’s historians, as well as in works on Macedonia during the Hellenistic period, and appears so scarcely as such in ancient texts. 164 The Atintanes have bedevilled Macedonian and Epirote scholars for decades. They are mentioned by Thucydides (2.80.6), Pseudo-Skylax (26), Pseudo-Aristotle (Mir. 833a 9), Lycophron (Alex. 1042–1046), Polybius (2.5.8; 11.11; 7.9.13), Strabo (7.7.8 Baladié), Livy (27.30.13; 29.12.13; 45.30.7), Appian (Ill. 7–8), Polyaenus (4.11.4), Stephanus Byzantius (s.v. Ἀτιντανία) and on an inscription from Dodona (SGDI 1336). The historical contexts of the above references are the Peloponnesian War, the Illyrian Wars of the Romans, the First Macedonian War and the Roman settlement of Macedonia in 167. 165 In spite of the relatively numerous citations, there has been no consensus on the location of Atintania. The obvious and after all roughly correct solution was sketched by M. Holleaux 166 and adopted by P. Lévêque, 167 namely that Atintania was the middle valley of the Aoos. 168 Since then several scholars have submitted different interpretations of the ancient evidence. There have been roughly three theories contesting the above location of Atintania (or Atintanis): 1) Hammond has contended in a series of articles stretching from 1966 to 1989 169 and in his monumental monograph on Epirus 170 that there are two different countries and two different ethne: the Illyrian Atintanoi in Atintania to the north-west of Lake Lychnitis and the Epirotic Atintanes in the upper valley of the Drynos in Central Epirus. 2) Hasan and Neritan Ceka have proposed in a series of articles that the Atintanes were an Illyrian ethnos whose territory extended originally between the territories of Orikos, Amantia and Byllis to

 163 Bosworth 1981, 91, n.22. 164 M. B. Hatzopoulos 1993c, 187–188 (= Recueil 529–530); Hatzopoulos 1996b, 77; cf. Cabanes 1976, 130. 165 Fot what follows, see Hatzopoulos 1993c, 183–190. 166 Holleaux 1921, 109–110 et n. 1. 167 Lévêque 1957, 184 and n. 7. 168 Lévêque added the valley of Drynos for no good reason. 169 Hammond 1966, 53–54 (= Collected Studies III 279–280), with fig. 3; Hammond 1968, 8 (= Collected Studies III 288), with fig. 1–2; Hammond 1989a, 11–25 (= Collected Studies III 245–259), with fig. 1. 170 Hammond 1967, 599–601, and map 15.

  The Land: Where was Macedonia? the north, of Chaonia to the west and of Molossia to the south, corresponding thus to the middle valley of the Aoos, but comprising also the valley of the Drynos as far as Dodona. The cities of Amantia and Byllis would have eventually seceded from this ethnos and formed separate states. 171 3) P. Cabanes espoused the view of his Albanian colleagues, but left the Drynos valley to the Chaones, reducing thus Atintania to a small area between the city Byllis and Dassaretis. 172 In a variant of this theory, Fanoula Papazoglou, would somewhat enlarge the territory of Atintania from the gorges of the Aoos at Kleisoura to the East to Selenice to the west. 173 In a 1993 paper I tried to show a) that Hammond’s theory of two homonymous countries and ethne in the same region, although understandable, was both unacceptable and unnecessary. It was simply the result of the incompatibility between two pieces of evidence: on the one hand, Kassandros’ stratagem in Polyaenus, which required a close proximity between Epidamnos and Atintania so that the fires in Atintanian villages would be visible from Epidamnos, and, on the other, the retreat of Epirote forces from Phoinike to Atintania related by Polybius, which presupposes proximity between Phoinike and Atintania, and in the same time is incompatible with the latter’s location anywhere close to Epidamnos, some 150 km distant from Phoinike. In fact, Polybius’ authority is confirmed by Thucydides, who couples the Atintanes with the Molossoi going to battle under the same commander, by Strabo (7.7.8 Baladié), who cites the Atintanes along with the Epirotic ethne of the Molossoi, the Athamanes, the Aithikes, the Tymphaioi, the Orestai and the Parauaioi, and by Livy (45.30.6), who includes Atintania into the Fourth Macedonian Meris along with Eordaia, Lynkos, Pelagonia, Tymphaia and Elimeia. The rest of the evidence has nothing much to contribute concerning the location of Atintania, while Appian’s (Ill. 7–8) use of the ethnic “Illyrian” for the Atintanes has been explained away as referring to their political situation after their annexation by Illyrian rulers. 174 The real crux of the matter is the passage from Polyaenus, which must result either from a misunderstanding of his source, probably Hieronymus of Κardia, or from an error of the latter due to his lack of familiarity with the geography of that region. The political history of Atintania was complicated. An independent principality allied to the Molossoi (Thuc. 2.80.6; cf. SGDI 1336), it was briefly annexed

 171 Ceka 1987, 135–149 (with earlier Albanian bibliography). See also Hatzopoulos 1993c, 190. 172 Cf. Cabanes 1976, m aps 1 and 8; cf. Cabanes 1986, 80–85. See also Hatzopoulos 1993c, 190. 173 Papazoglou 1986, 444, n. 27. 174 Cabanes 1986, 82.

Some particular cases  

by the Illyrians (230), but the Atintanes took advantage of the Roman intervention of the following year in order to put themselves under Roman protection (App. Ill. 3.7, 229). A few years later they sided with Demetrios of Pharos (App. Ill. 3.8, 222), but after the latter’s defeat Atintania was recovered by the Romans. In the following years Atintania was disputed between the Romans and the Macedonians (Livy 27.30.13), and was finally awarded to Macedonia by the peace of Phoinike (29.12.13). It must have remained Macedonian thereafter, since it was included into the Fourth Macedonian Meris (Livy 45.30.7).

.. Derriopos Derriopos is described by Livy (39.53.14) as a region of Paionia (Paeoniae ea regio est), but in a different passage the same author (31.39.4) locates the main Derriopian city Styberra in Pelagonia (Stuberram deinde petit atque ex Pelagonia frumentum quod in agris erat convexit), while Peithon son of Krateuas, from Alkomena, a town in Derriopos (Πείθων Κρατεύα Ἀλκομενεύς), is listed among the Macedonians by Arrian (Ind. 18.6), and one of Philip V’s courtiers ‘signs’ a dedication as “Antigonos son of Herakleitos, citizen of Styberra, Macedonian” (Ἀντίγονος Ἡρακλείτου Στυβερραῖος Μακεδών). 175 Moreover, as Papazoglou rightly stresses, onomastic material from Styberra 176 verifies the Greek and more precisely Macedonian character of the population. 177 It is possible that the Derriopes were an outshot of the Upper Macedonian ethne, the foothills of Mount Peristeri assuring the territorial continuity with the Macedonian Lynkestai. What can one deduce from these contradictory pieces of evidence? First, if Derriopos could be considered a region of Paionia and at the same time the Derriopian city Styberra could be described as a city in Pelagonia, then either Pelagonia itself was a part of Paionia or Pelagonia and Paionia were interchangeable geographical terms. Both propositions may be true, since Strabo (7 fr. 38 Baladié) indicates that Paionia included Pelagonia, while an Athenian inscription of 365/4 honours a Π[......] king of the Pelagonians. The name should most probably be restored as Π[άτραον], perhaps grandfather of another Patraos, king of Paionia in the second

 175 SEG 35 (1985) 821. 176 But also from the other sites of Derriopos. 177 Papazoglou 1988a, 250: “L’anthroponymie des stèles éphébiques de Stuberra présente une prépondérance absolue de noms grecs (gréco-macédoniens) ... Le tout donne l’impression d’un milieu purement grec”.

  The Land: Where was Macedonia? half of the fourth century. 178 The name of Pelagonia was used for the westernmost part of Paionia which was incorporated into the Fourth Meris by the Romans in 167, but the north-westernmost part of Pelagonia continued to be referred to as Derriopos. 179 To conclude, although there remained always a distinction between Macedonia and the vaster arche of the Macedonian kings, there was a constant pressure to open up new lands for Macedonian colonisation, and to assimilate the ‘allies’. Consequently, the national territory tended to become coextensive with the Macedonian possessions. This process eventually made the national Macedonian territory extend from the Pindos mountains practically to the Strymon valley and perhaps beyond. Epigraphic evidence displays the extraordinary legal homogeneity of the Macedonian land and the Macedonian people. By the end of the Antigonid period all the free inhabitants (with the notable exception of freedmen and foreign tenant farmers) of the national territory, regardless of their ultimate ethnic origins were citizens of one or the other civic communities (polis, ethnos or sympolity) 180 and together formed the Macedonian ethnos, one of the two constituent parts of the Macedonian state.

 178 Cf. Papazoglou 1988b, 276, n. 2. 179 Cf. Papazoglou 1988b, 68–71, and the map in Sakellariou 1983, 198. 180 On the different types of Macedonian civic communities, see Hatzopoulos 1996b, 47–123.

 Who were the Macedonians? . The evidence of the ethnics The definition of the Macedonian identity can be read in the formulas used by the Macedonians when ‘presenting’ themselves: “Agelochos son of Philippos, Macedonian from Aigeiai” (Delphi, FD III 1 112: Ἀγέλοχος Φιλίππου Μακεδὼν ἐξ Αἰγειῶν), Antigonos son of Herakleitos, Styberraian, Macedonian (Kaliakra, SEG 35 [1985] 821: Ἀντίγονος Ἡρακλείτου, Στυβερραῖος Μακεδών), Antipatros son of Theodoros, Amphipolitan (Athens, SEG 32 [1982] 290: Ἀντίπατρος Θεοδώρου Ἀμφιπολίτης), Antigonos son of Amyntas, Beroian (Demetrias, A. S. Arvanitopoulos, Θεσσαλικὰ Μνημεῖα [Athens 1909] 295−296, no 82: Ἀντίγονος Ἀμύντου Βεροιαῖος), Sosikrateia daughter of Artemon, Macedonian (Demetrias, op. cit. 134, no 11: Σωσικράτεια Ἀρτέμωνος Μακέτα), Aristodemos son of Artemon, Kassandreian (Demetrias, op. cit. 417, no 167: Ἀριστόδημος Ἀρτέμωνος Κασσανδρεύς), Philoxenos son of Simmias, Tyrissaian (Marinia, EKM II 117: Φιλόξενος Σιμμία Τυρισσαῖος). What do these ‘self-presentations’ tell us? Firstly, that Greek onomastic formulae did not display the same rigidity as Roman ones. Secondly, that Macedonians behaved exactly like the citizens of other Greek ‘federal’ states. For instance, an Aitolian could be identified at Delphi as Teisandros son of Mikkinas, Aitolian from Bottos (CID IV 23: Τείσανδρος Μικκίνα Αἰτωλὸς ἐγ Βοττοῦ) or simply as Aitolian Bouttios (CID IV 101: Αἰτωλῶν ... Λεοντομένεος Βο[υττ]ίου) or even more simply as Lanikos Aitolian (CID IV 33: Λανίκου Αἰτωλοῦ); an Akarnanian as Satyros son of Nikomachos, Akarnanian from Tyrbeion (Delphi, ibid. Σάτυρος Νικομάχου Ἀκαρνὰν ἐκ Τυρβείου) or a Thessalian as Nikostratos son of Anaxippos, Thessalian from Larissa (CID IV 106: Νικόστρατος Ἀναξίππου Θεσσαλὸς ἐγ Λαρίσσης). The reasons behind the use of the one or the other formula were practical. The formal identification with name, patronym, ‘federal’ ethnic, and name of the city introduced by a preposition or the ethnic of the city followed by the ‘federal’ ethnic was mainly used abroad, outside the ‘federal’ state. When the city from which the person hailed was important and well known in the Greek world, the ‘federal’ ethnic could be omitted, or, conversely, when the city was insignificant or for some other reason the person concerned set greater store by his or her Macedonian identity than by his or her civic one, as did Sosikrateia in Demetrias (Arvanitopoulos, Μνημεῖα134 no 11: Σωσικράτεια Ἀρτέμωνος Μακέτα), while her brother Aristodemos preferred to be remembered as a citizen of Kassandreia (op. cit. 417, no 167: Ἀριστόδημος Ἀρτέμωνος Κασσανδρεύς). On the other hand, the omission of the ‘federal’ https://doi.org/10.1515/9783110718683-003

  Who were the Macedonians? identity within the boundaries of the ‘federal’ state, as in the case of Philοxenos, was the regular practice. As stressed in Part I, these variants should in no case be used as evidence of diverse categories of Macedonians, such as “Mitglieder der Hochadel” or “schlichte Bewohner der χώρα βασιλική”. 1 Macedonia, just as pre-republican Molossia, were ‘ethnic’ (in modern terms ‘federal’) states, the citizens of which had multiple overlapping political identities. A local ‘federal’ magistrate (damiourgos) under the Molossian king Neoptolemos in the beginning of the fourth century, besides being Molossian, was also Arktan from Eurymenai (Dodona, Cabanes, Epire 534−535: Ἀνδοκάδεος Ἀρκτᾶνος Εὐρυμεναίων), just as under Antigonid rule Philarchos son of Hellanion, a Macedonian proxenos of Delphi, was “Macedonian, Elimiote from Pythion (Φιλάρχωι Ἑλλανίωνος Μακεδόνι Ἐ[λ]ειμιώτ[ηι] ἐκ Πυθείου). 2

. Weder Blut noch Boden: How does one become Macedonian? Of course, most Macedonian citizens were children of Macedonian citizens. There is no reason to believe that the Macedonians conformed to Perikles’ law on illegitimate children 3 or that Macedonian citizenship of both parents was a prerequisite for inscription in the citizens’ roll as in Athens. 4 At any rate, it did not apply to the royal family, in which dynastic unions with foreign princesses were a current practice, without raising questions of legitimacy. 5 Most probably, in a society with a conquering past like that of the Macedonians, marriage with the womenfolk of the conquered populations must have been too frequent to allow for the exclusion of the offspring thereof. Conversely, the very conquering bent of the Macedonian state was incompatible with conferment of citizenship exclusively by birth.

 1 Rosen, unpublished dissertation without indication of place or date, 12; cf. Hatzopoulos 1996b, 168. 2 Cf. Cabanes 1980, 323−351. 3 Plut. Per. 37. 4 As Badian 1963, 244: “At the wedding feast, Attalus, the bride’s uncle, insulted Alexander, implying—what was probably true, and indeed obvious enough—that this marriage was intended to give the country a legitimate (i.e. fully Macedonian) successor”. 5 Cf. the vain discussion about Alexander’s legitimacy because of the Molossian origin of his mother. Cf. Ellis 1981, 120−121; Hatzopoulos 2018b, 69−71.

Weder Blut noch Boden: How does one become Macedonian?  

A theoretical schema, both complete and convincing within the limits of the early formation of the Macedonian state, was put forward by Paola Zancan in her 1934 thesis, but unfortunately went unnoticed by practically all English-speaking scholars. 6 Having firstly stressed “that the term Μακεδών did not imply a concept purely or essentially ethnic”, but that by the fourth century at least “it included a political concept”, she proposed to interpret the legend which exemplified the Macedonian aversion to build trophies (Paus. 9.40.8−9) as an allegory of generosity towards the vanquished enemy. The decision to let a Macedonian victory fall into oblivion was intended to pave the way for a future reconciliation with a former enemy and for their eventual inclusion within the Macedonian community. Yesterday’s enemies would become tomorrow’s ὑπήκοοι and the day after tomorrow’s σύμμαχοι, fighting along the Macedonian citizens in the army, before finally becoming full Macedonian citizens themselves. Hammond independently propounded an apparently similar but in fact different thesis. For him the Makedones proper were “an élite group of men, trained for war and chosen by the king for their prowess in war and reliability in military service, a small proportion only of all his subjects”. 7 They were fully enfranchised, took part in the ethnos assembly and possessed both Macedonian and local citizenship in the city where they resided 8. The other free inhabitants of the country had only local citizenship. An opposite view was submitted by Kl. Rosen in his unpublished doctoral dissertation: most Macedonians possessed both citizenships, but a few of them, the highest nobles and the poorest inhabitants of royal land had only Macedonian citizenship. 9 Hammond’s theory was forthwith challenged by E. M. Anson, R. M. Errington, Fanoula Papazoglou and myself. 10 More recently Papazoglou returned to this question suggesting that the term Makedon never lost its ethnic connotation, but that the military reforms by Philip II and Alexander III incorporated into the army recruits from other ethnic groups conferring them thereby Macedonian citizenship. 11 She drew an insuperable line between the “Argead” (meaning Temenid) 12  6 Zancan 1934, 124−136. Ellis 1976 is a notable exception. 7 Hammond in Hammond/Griffith 1979, 163−164. 8 Hammond in Hammond/Griffith 1979, 163−164; 647−652. 9 Rosen, unpublished dissertation (without indication of place or date), 9−13. 10 Anson 1984, 67−68, and in Errington’s 1980, 78−80 and 1989, 288−300, Reviews of Hammond/Griffith 1979, and of Hammond/Walbank 1988, and by Papazoglou 1990, 229−233; Hatzopoulos 1996b, 167−171. 11 Papazoglou 1998, 25−41. 12 Argead is in principle the name of the population group (“clan”?) who founded the kingdom of Lower Macedonia, whereas Temenid is the name of the reigning dynasty until the death of

  Who were the Macedonians? and the Antigonid period, because she underestimated the development of civic institutions already in fourth-century Macedonia. In fact, neither Philip’s nor Alexander’s soldiers were all Macedonians, but on the other hand all Macedonian soldiers were citizens of Macedonian cities. Neither Nearchos nor Eumenes hailed from Macedonia, but had their origins in Crete and Kardia respectively. Nevertheless, the former is listed among the Macedonians while the latter among the Greeks (Arr. Ind. 18.3−11). The distinction had nothing to do with ethnic origins or with military service, but with the fact that whereas Nearchos was a citizen of a Macedonian city, namely Amphipolis, Eumenes was not. This is neither a unique nor a recent phenomenon. Laomedon and Androsthenes were also listed among the Macedonians, although they were from Mytilene and Thasos respectively, because they were both citizens of Amphipolis. Already in the second quarter of the fourth century, that is in the very early years of Philip’s reign, we find persons bearing typical Macedonian and non-Macedonian names sharing magistracies or buying and selling landed properties in Amphipolis, something that would not have been possible, had they not shared a common citizenship. 13 Unambiguously, Philip V’s diagramma on military service stipulates that incorporation of non-citizens in the regular Macedonian army was prohibited under heavy penalties, unless there was a written dispensation from the king. 14 Hammond in his later works 15 reiterated his theory in a modified form, and I must admit now that there is some truth in this later version. The diagramma on military service ulteriorly published shows indeed that the king could in particular cases authorise local authorities in writing to insert non-citizens in the civic elementary recruitment units, making them thus eligible for service in the armed forces. 16 Moreover, as Hammond rightly pointed out, there were circumstances  Alexander IV in 309. Cf. Hammond in Hammond/Griffith 1979, 26−28. The stele of the Kytenians (Bousquet, 1988, 12−53) made me momentarily waver (Hatzopoulos 2003a, 216 and n.74 [= Recueil 206 and n. 74]), but since the Kytenians included also Antiochos within the Heraklid/Argead kings (although the Seleucids claimed a distinct Apollinian ancestry), and since Philip V not only claimed an Argead origin for himself (Pol. 5.10.10; Livy 27.30.9; cf. Livy 32.22.11) but also had it ascribed to him by hostile sources (Paus. 7.8.9; see Edson 1934, 213−246), the term Argeadai must usually have referred to a wider cercle of families than the Temenid dynasty (cf. Just. 7.1.10: populum Aegeadas vocavit). Nevertheless, cases of confusion between Temenids and Argeads cannot be excluded (cf. Bousquet 1988. 40, n. 55). 13 Hatzopoulos 1991, 75−82; Hatzopoulos 1996b, 184; Hatzopoulos 2007a, 59−63 (= Recueil 401−415). 14 Hatzopoulos 2001a, 96−98 and 154, L. 21−34. 15 Papazoglou 1998, 25−41. Hammond 1989b, 49−64; 152−165 and 382−395; Hammond in Hammond/Walbank 1988, 12; 86; 484. 16 Hatzopoulos 2001a, 96−98 and 154, ll. 21−34.

Weder Blut noch Boden: How does one become Macedonian?  

where royal decisions resulted indirectly in the conferment of Macedonian citizenship. When, for instance, Macedonians from the Old Kingdom were transplanted into the New Territories beyond the Axios and were given land there, the previous population did not simply disappear. Onomastic evidence shows that some were awarded full Macedonian citizenship right from the start, 17 while others, especially if they lived in the countryside and were not of Greek origin, may have occupied in their villages an inferior position as cultivators attached to citizens of the Macedonian city. 18 Consequently, from the middle of the fourth century at least, inclusion in the Macedonian community does not seem to be the result of service in the army; conversely, military service seems to have been a consequence of acquiring Macedonian citizenship through enrolment in a politeuma. In sum, it seems that the question of civic rights had less to do with ethnic origin or with military service and more with social stratification. If we except a passing remark by K. J. Beloch, the first to turn his attention to the social stratification of the Macedonian population was Franz Hampl. 19 To the question “who belonged to the class of the hetairoi in Macedonia?” he answered that they were neither the narrow circle of the king nor those serving in the heavy cavalry, as G. Plaumann had asserted, 20 but a group of people who had a feudal relationship with the person of the current king, offering service in exchange for revenues from landed property in the royal land. Hampl’s feudal thesis was attacked with various arguments, some of which were pertinent, while others were off the mark, by P. de Francisci. 21 A generation later, J. R. Ellis distinguished three social groups: the hetairoi (Companions), who manned the heavy cavalry, 22 the ‘soldiers’, who served both as heavy infantry and in the élite corps of the hypaspists, 23 and in the light cavalry, and the “sub-citizens”, serfs like the Spartan helots and the Thessalian penestai, who served in the army in auxiliary units (groomsmen, servants, wagondrivers etc.). In the same period, Hammond, besides distinguishing the “Macedo-

 17 See Hatzopoulos 1991, 80−86; Hatzopoulos/Loukopoulou 1992, 110−122. 18 Such may have been the condition of the λαοί mentioned in EKM II 401, l. 31. 19 Hampl 1934, 66−67. 20 Plaumann 1913, 1374−1379. 21 de Francisci 1947, 360−380. 22 He does not mention the infantry guard units, the pezhetairoi under Philip (see Griffith in Hammond/Griffith 1979, 705−713). 23 It is an anachronism. See previous note.

  Who were the Macedonians? nes proper” from the rest of the free population who enjoyed only a city citizenship and served in local militias, 24 asserted that there were no domestic slaves in Macedonia. 25 Fanoula Papazoglou, a Marxist, put the stress on class distinctions. Contrary to Hammond, she refused to believe that there were no domestic slaves, a matter in which she was proven right, since records of manumissions 26 and tombstones most probably belonging to slaves 27 from pre-Roman Macedonia have been coming to light lately. Nonetheless, the phenomenon did not seem to have been as widespread as in the major polis-states of Southern Greece. Conversely, Papazoglou disbelieved Ellis’s assertion that the estates of hetairoi in newly conquered regions populated by non-Macedonians were cultivated by serfs. 28 Consequently, she interpreted the term metoikos, which makes a single appearance in pre-Roman Macedonia, as designating a free tenant of royal land. 29 In fact, non-Macedonian tenant farmers are precisely attested in Bottia, 30 where Gauls are evidenced and where the word λαούς has been read on a badly preserved inscription from Kyrrhos. 31 Many authors have stressed the challenge of unifying a geographically and culturally diverse country that Philip II had to face after the extension of his dominions from the Pindos chain to the Strymon valley and from the Tempe gorge to Lake Ochrid. As we have seen, both Ellis and Papazoglou related this achievement to the reform of the army. The slight difference between their interpretations lies in that, whereas Ellis assumed that Philip’s aim was to expand and unify the army in order to unify the state, 32 Papazoglou found more appropriate to say that the unification of the army had as side-effect the unification of the state. She argued that Philip, by choosing the best and strongest young men irrespectively of their Macedonian or non-Macedonian origin and by subjecting them to the same military discipline, made Macedonians out of the non-Macedonian recruits. 33 I

 24 Hammond in Hammond/Griffith 1979, 647−648. 25 Hammond in Hammond/Walbank 1988, 12. 26 Papazoglou 1998, 33 and 35. See EKM I 45−46; ΕΚΜ II 174; 402−404. 27 Cf. EKM II 44; 470; 473. 28 Papazoglou 1998, 34. 29 EAM 87; Papazoglou 1997, 239, n. 466. 30 Livy 45.30.5: tertia regio nobilis urbes Edessam et Beroeam et Pellam habet et Vettiorum bellicosam gentem, incolas quoque permultos Gallos et Illyrios, inpigros cultores. 31 ΕΚΜ ΙΙ 401, l. 31. For details, see Hatzopoulos 2011a, 62−68 (= Recueil 500−506); Hatzopoulos 2013a, 1376−1377. 32 Particularly Ellis 1969, 9−17; Ellis 1977, 103−111; Ellis 1981, 36−47. 33 Papazoglou 1998, 36−37.

Weder Blut noch Boden: How does one become Macedonian?  

believe that both authors undervalued the importance of civic institutions, which is natural, since at the time they were writing the importance of new epigraphic evidence which highlighted the early integration of non-Macedonians in the new Macedonian cities that Philip founded (e.g. Herakleia Lynkestis), re-founded (e.g. Kalindoia) or reformed (e.g. Amphipolis, Torone) had not yet been fully grasped. 34 Indeed, from the time when epigraphic sources become available, it is inclusion in the citizen corps that leads to enrolment in the army and not the other way round. Therefore, the formation of Macedonian civic units was a prerequisite for the creation of a stable system of recruitement in a citizen army (πολιτικοὶ στρατιῶται). 35 This social and political revolution epitomised in one sentence in the speech attributed to Alexander by Arrian (An. 7.8.2: “He made you dwellers of cities, and graced you with good laws and customs”) implied a complex demographic, economic and institutional operation. In recent times its demographic facet has been studied by Ellis in an article especially devoted to population transplants by Philip II, 36 by Hammond in a whole chapter of his monograph on the Macedonian state, in which demography, economy, urbanisation and military developments are holistically examined, 37 and it has been detected ‘on the ground’ in the Macedonian cities founded or refounded in Krestonia, Anthemous and Northern Bottike by Louisa Loukopoulou and myself. 38 In my book on Macedonian institutions I emphasised the unfailing progressivity of Macedonian policies attested by Thucydides (2.99.3−6), Aristotle (Pol. 1310 b), and Theopompus in Justin (7.1.1−12; 8.5.7−6.2), and leading from expansion to annexation and to assimilation within an urbanisation context. 39 More recently, E. M. Anson, in a useful reappraisal of Philip II’s policies, stressed the connexion between conquest, land distribution, urbanisation, recruitment in the phalanx and the unweavering loyalty of the poorer Macedonians to the king who turned “a dependent population of herdsmen and tenants into a nation containing tens of

 34 For my point of view, see Hatzopoulos 2012, 37−53. 35 Alexander inherited from Philip II an army organised on a civic basis; cf. Diod. 18.12.2: ἐσπάνιζε γὰρ ἡ Μακεδονία στρατιωτῶν πολιτικῶν διὰ τὸ πλῆθος τῶν ἀπεσταλμένων εἰς τὴν Ἀσίαν διὰ τὴν διαδοχὴν τῆς στρατείας. 36 Ellis 1969, 9−17. 37 Hammond 1989b, 152−165; 49−53. 38 Hatzopoulos/Loukopoulou 1989, Hatzopoulos/Loukopoulou 1992, Hatzopoulos/Loukopoulou 1996. See also Hatzopoulos 1991, 62−86 for the transformation of Amphipolis into a Macedonian city. 39 Hatzopoulos 1996b, 169; 179 and 473−474.

  Who were the Macedonians? thousands of loyal landowners”. He rightly questioned Ellis’s and Billows’s assumption 40 that these poorer Macedonians belonged to a conquerd subject population, and had a status comparable to that of the Laconian helots or the Thessalian penestai. 41 On the other hand, there is no reason to assume that all beneficiaries of land distribution in the conquered territories were herdsmen and landless tenants. Small landholders might be willing (or compelled) to abandon a small holding in the Old Kingdom in favour of a bigger estate in the New Territories. 42 Blood was certainly not the sole means of acquiring Macedonian citizenship, but neither was birth on Macedonian soil. Children of non-Macedonian tenant farmers, descendants of earlier populations – no more than children of manumitted slaves – were not automatically awarded Macedonian citizenship solely because they were born on the territory of a Macedonian city. Nevertheless, as previously asserted, in the long run most were aggregated to the citizen body. The same is true of ‘allies’ of Macedonian kings (in fact subject populations) and former mercenaries (Thracians, Illyrians, Gauls etc.) who were settled among Macedonians in depopulated areas. Epigraphic evidence leaves no doubt that they were assimilated by the indigenous Macedonian population, becoming Greekspeakers and sharing in local cults and beliefs, 43 because there were no insuperable barriers for those who were disposed to learn Greek and to go through Greek educational institutions, mainly the gymnasium, to ascend socially. 44 Perhaps some of them were to be found among the syntrophoi of young Macedonian citizens. 45 The question still to be answered is “had all Macedonian citizens equal rights?” Though he did not mention it, Ellis gave serious consideration to a passage by Curtius Rufus (6.8.23) referring to the trial of Philotas: VI milia fere militum venerant, praeterea turba lixarum calonumque impleverant regiam. Who were

 40 Ellis 1976, 27; Billows 1995, 10. 41 But see below for the exceptional situation in the royal land. 42 More recently Mari 2019, 213−239, returned to the same subject and to a conception of the connexion between military service, Macedonian citizenship (and land ownership) somewhat closer to Hammond’s. For her (p. 2015 and passim), land ownership was a privilege derived from service in the army, and “the acquisition or maintenance of land ownership was the prerequisite for gaining the full ‘citizenship’ in the Macedonian state and the right to participate regularly in the assemblies of the Μακεδόνες”. 43 Cf. the dedications made to Athena of Kyrrhos by slaves manumitted by persons bearing Celtic names: ΕΚΜ II 403−405. 44 Hammond 1989b, 388−391. 45 Hatzopoulos 2015−2016a, 57−70.

Weder Blut noch Boden: How does one become Macedonian?  

those valets, those grooms? Ellis’s answer was, as we have already mentioned, that they belonged to a class of “sub-citizens” comparable to Spartan helots and to Thessalian penestai, that is to say that they were agricultural labourers attached to the soil, who manned the army commissariat. This hypothesis reflected the complete ignorance of social conditions in pre-Roman Macedonia prevailing nearly half a century ago. In the light of the new epigraphic discoveries there is nothing that can substantiate a theory of Macedonian helots. To begin with, a distinction should be made between conditions prevailing in royal land or ex-royal land devolved on an ‘allied’ city such as Kassandreia, on the one hand, and in the “land of the Macedonians”, on the other. 46 For instance, among the estates granted to Limnaios son of Harpalos by king Lysimachos the one situated in the territory of Sermylia bordered the lands belonging to Lysimachos’ eldest son Agathokles and to one of the king’s courtiers (Bithys son of Kleon) respectively; the one in the territory of Olynthos bordered the lands of two citizens of Kassandreia (Menon son of Sosikles, Pylon son of Epiteles), judging by their Ionic names, probably descendants of Olynthians who had betrayed the city to Philip II; finally, the one in the territory of Strepsa bordered the lands of one person with a Bottic, that is a non-Greek, personal name and patronym (Gouras son of Annythes) and of two brothers, who according to their personal names and to their patronym (Chionides and Eualkes sons of Demetrios) were probably not of Macedonian but of southern Greek origin. 47 It is obvious that neither Lysimachos’ son nor his courtier cultivated the land themselves. They probably farmed out their estates, which were perhaps as vast, or even vaster, than that of Limnaios, to tenant farmers who might possess or not possess Kassandreian citizenship. If the more modest estates of Menon and Pylon in the ‘select’ neighbourhood of Trapezous were of a size comparable to that of Limnaios, the two Kassandreians could either exploit them themselves with the help of hired hands or farm them out to tenants. It is more difficult to visualise the status of Gouras and the brothers Chionides and Eualkes. If the size of Limnaios’ estate was a valid indication of the size of his neighbours’ estates, it would give ground for inferring that the spoliation of the former citizens of Strepsa had not been that severe. In any case, estates of that scale (920 plethra), even if they were mainly olive groves, could not be exploited without the help of tenant farmers (or chattel slaves, of whom there is no trace). In conclusion, on royal or former royal land we find substantial land owners with ‘extraterritorial’ (Agathokles, Bithys, Limnaios) or local  46 On the status of Kassandreia before the reign of Antigonos Gonatas, see Hatzopoulos 1993b, 575−584. 47 See p. 30, above.

  Who were the Macedonians? citizen status, and, by inference, tenants and agricultural workers of indeterminate status. The situation was different in the “land of the Macedonians”. 48 There were Macedonian citizens of various social positions: king’s “Companions”, later “Friends” (ἑταῖροι, ἀμφ’ αὐτὸν ἑταῖροι, φίλοι, πρῶτοι τῶν φίλων, in Latin amici, principes amicorum, purpurati), who formed the innermost circle; bodyguards, cavalry and phalanx officers, commanders of garrisons (σωματοφύλακες, later ὑπασπισταί, ἴλαρχοι, ἡγεμόνες τῶν τάξεων, φρούραρχοι, custodes corporis, satellites, duces copiarum, praefecti praesidiorum), generally called in Latin regii or purpurati, formed a second circle. And then of course came the other Macedonians, who served under the orders of the aforementioned in the guard cavalry (βασιλικαὶ ἶλαι) and infantry (ὑπασπισταί, later πελτασταί) regiments, in the heavy or light cavalry, and in the phalanx. All the above were both Macedonian citizens and citizens each of his own civic unit. 49 Recent epigraphic discoveries have revealed differentiations of financial nature within the body of Macedonian citizens. The diagramma on military service prescribed that the wealthier citizens should serve in the agema and in the peltast regiments, while the less wealthy should serve in the phalanx, and that the bodyguards should be chosen in consideration of each person’s aptitudes and of his real estate and movable property. 50 Troopers are equated to infantry junior officers and receive the same treatment. 51 Moreover the fragmentary gymnasiarchical (?) law from Neapolis (Kavala) seems to link somehow the recruitment of fifteenyears old boys to the local gymnasium with the sum of 200 drachmae, probably of revenue per year, 52 while the ephebarchical law from Amphipolis compels the ‘boys’ who have registered as ephebes and possess a property of 3,000 drachmae in real estate and in cattle to follow the ephebic training every day unfailingly. 53 It must be stressed however that a property of 3,000 was not a prerequisite for

 48 Cf. Hatzopoulos 1996c, 74−75, no 58, l. 6: τὴν λοιπὴν χώραν τὴν Μακεδόνων. On the distinction between “royal land” (cf. SEG 60 [2010] 606: τὸ βασιλικόν) and “land of the Macedonians” (τὴν λοιπὴν χώραν τὴν Μακεδόνων), see also Faraguna 2019, 50−51. Dependent populations did exist in Hellenistic Macedonia, not only in the New Territories, but also in the Old Kingdom (See p. 54, above). What we do not know is whether they assured the exploitation of royal land only or were also active on privately owned civic territory. Cf. Hatzopoulos 2011a, 62−69 (= Recueil 500−506). 49 Cf. Hatzopoulos 1996b, 332−333, and Hatzopoulos 2001a, passim. 50 Hatzopoulos 2001a, 103−107, and 155, no 2 I, B, L. 1−8; 158, no 2 II, L. 14−19. 51 Tziafalias/Helly 2010, 104−105. 52 Hatzopoulos 2001a, 123−127, and 164, no 4. 53 Lazaridou 2015, 3, L. 14−19; Hatzopoulos 2015−2016b, 155−156.

Weder Blut noch Boden: How does one become Macedonian?  

following the ephebic training. It rather meant that the more fortunate young men were obliged to observe the strictest assiduity, while there was a more tolerant attitude towards the less fortunate ones, which is characteristic of oligarchic regimes. 54 It is probable that a similar rule applied to ‘boys’ with an income of 200 drachmae or more per year. On the other hand, the gymnasiarchical law from Beroia provided for the exclusion from the gymnasium of ‘boys’ and young men on the basis of a number of social criteria. 55 The list of those excluded comprised the slaves, the manumitted slaves and their offspring, the physically unfit (apalaistroi), the (male) prostitutes, the shopkeepers and craftsmen, the drunks and the mentally insane. The slaves and the manumitted slaves and their offspring were clearly not citizens. Manumission, provided that it did not comprise a paramone clause, granted effective freedom, but not citizenship. The manumitted slave obtained at best the status of a metic. The physically unfit could obviously neither follow the gymnastic and military training nor could they serve in the army. Their physical handicap made them moreover unfit for most civic offices which required a minimum of physical fitness. Even if they were not de iure deprived of their citizenship, they were de facto second-rate citizens. The exclusion of prostitutes, drunks and mentally insane aimed at keeping good order in the gymnasium, but that of the shop keepers and the craftsmen raises difficult issues. It recalls the prohibition made to the ephebes to stop at workshops or to enter a marketplace, 56 which illustrates the aversion of Macedonian mentality for manual labour and money-making. Ideally the citizen ought to be ‘a gentleman of independent means’, hence the special duty of persons disposing of such means to train as ephebes and to serve in élite army corps. Nevertheless, according to Philip V’s diagramma on military service, even the ἀπορώτατοι and οἱ ἐλαχίστην οὐσίαν ἔχοντες are recruited in the phalanx infantry (πεζοί). 57 Therefore, it seems that, in Antigonid Macedonia at least, only physically unfit citizens, or citizens who were for some other reason untrained – probably along with non- citizens– served as valets and grooms in the army. We simply do not know whether their inferior status had as consequence the curtailing of their political rights as Macedonian and local citizens. A last question is whether the Macedonian state had made provisions so as to increase the number of trained youths, in order not to be deprived of possible sources of military manpower. It is certain that most Macedonian kings (Amyntas

 54 Arist. Pol. 1297a. 55 Gauthier/Hatzopoulos 1993, 20−21, L. 27−39, and 78−87. 56 Lazaridou 2015, 5, L. 42. 57 Hatzopoulos 2001a, 103−104.

  Who were the Macedonians? I, Alexander I, Perdikkas II, Philip II, Alexander III, Kassandros, Antigonos Gonatas, Philip V) pursued a vigorous demographic policy. 58 Besides the measures taken by Philip II, which are known to some detail and were examined above, a heretofore unsuspected one has now been suggested by the same, unfortunately fragmentary, diagramma on military service, which mentions syntrophoi replacing young Macedonian recruits by serving in the light cavalry. 59 Apparently these, like the Spartan mothakes, were boys of inferior social or financial status who were raised in the families of well-to-do boys of the same age and were thus enabled to receive the same education in the gymnasia, and could, in case of need, replace their foster brothers in their military obligations. 60

. The origin of the Macedonians The word “origin” can have two different meanings: a spatial one: “the starting point”, or a chronological/logical one: “that from which anything first proceeds”. In a previous chapter I skimmed over the first issue when discussing the different stages of Macedonian expansion. 61 Fortunately, right from the start there was a rough agreement concerning the geographical starting point of the Macedonians: they came from the west. 62 Herodotus (8.137.1) clearly stated that the three brothers from Argos went to Illyria and thence, crossing over the mountains, to Upper Macedonia. Appian (Syr. 63) confirmed the western origin of the Argead Macedonians who founded the kingdom of Lower Macedonia, although he erroneously connected the name of that Macedonian clan, Argeadai, with the city of Argos in Orestis. 63 Nevertheless, before the identification of Aigeai with Vergina, there was a problem with this theory. It was partly contradicted by the obvious chronological understanding of Thucydides’ account of Macedonian expansion, which placed the conquest of Eordaia after that of Pieria, Bottia, Lower Paionia and Mygdonia. It was unthinkable that all these conquests could have taken place before the Macedonians had set up their capital at Aigeai. But, coming from the west,

 58 Hatzopoulos 2015−2016a, 65−68. 59 Hatzopoulos 2001a, 159, no II 2, L. 52−55. 60 Hatzopoulos 2015−2016a, 57−70. 61 See p. 11–15, above. 62 Müller, K. O. 1825, 42−43; Abel 1874, 97; Cloché 1960, 26−27. Rosen 1978, 1−27. 63 As Hoffmann 1906, 121−122, has shown that Ἀργεάδαι cannot be an ethnic. It can only designate the descendants of a person named Ἀργέας. Cf. Geyer 1930, 37. See p. 51, n. 12, above.

The origin of the Macedonians  

how could they have captured Aigeai, at or near Edessa, without having first conquered Eordaia? 64 As we have seen (pp. 12–14, above), Macedonian scholars were faced with the uncomfortable choice of either taxing Thucydides with chronological inconsequence 65 or of arguing that he did not follow a chronological, but a (rather complicated) geographical order. 66 Only Kl. Rosen, in his Chiron article on the foundation of Macedonian power, managed to get around this difficulty, but at the price of two errors: he arbitrarily put Lebaie in Pieria instead of Upper Macedonia, and he did the same with Vergina, although it is in fact situated in Makedonis (later comprised into Bottia). 67 Moreover, he discussed the excavations at the site of Vergina without any reference to Hammond and to his identification of this site with ancient Aigeai. One wonders why he did not take advantage of the original and convincing reconstruction of Macedonian expansion elaborated by Paola Zancan, 68 although he must have read it judging from the inclusion of her work in the bibliography of his earlier dissertation. 69 Starting from Herodotus’ passage (8.43; cf. 1.56.3) on the Δωρικὸν καὶ Μακεδνὸν ἔθνος ἐξ Ἐρινεοῦ τε καὶ Πίνδου and continuing with the Hesiodic Catalogue of Women (fragment 7 M-W) on Zeus’ sons, Μάγνητα Μακηδόνα θ’ ἱππιοχάρμην οἳ περὶ Πιερίην καὶ Ὄλυμπον δώματ’ ἔναιον, Zancan follows the peregrinations of the early Macedonians from Pindos to the western slopes of the Pierian mountains and thence to the Μακεδονὶς γῆ on the northern foothills of these mountains called by Herodotus (7.131) ὄρος τὸ Μακεδονικόν. Her bold and innovative conclusion, anticipating Hammond’s identification of Aigeai with Vergina by thirtyfive years, was that “in this region the (Macedonian) conquerors had their first capital ... that this region was the first step of the conquest, the nucleus of the future kingdom”. She also deduced that it was for good reason that Thucydides named Pieria, to the south of the Μακεδονὶς γῆ, as the first Macedonian conquest. Zancan did not fail to highlight another consequence deriving from the verification of Herodotus’ notices: for him the Macedonians, just like the Dorians belonged to the ἑλληνικὸν ἔθνος (cf. Herod. 1.56.2−3). Since the extent and hence the chronology of Macedonian expansion has been already discussed in a previous chapter, there only remains to examine to  64 For what follows, see Zahrnt 1984, 330−334. 65 Müller, K. O. 1825, 31; Geyer 1930, 39−40; Kanatsoulis 1964, 13−15; Beloch 1912, 1, 341; Kaerst 1927, 166; Cloché 1960, 26−27 and 29−31; Daskalakis 1960, 29−34. 66 Hammond 1972, 432−440. 67 Rosen 1978, 12−13 and 22−27; on the location of Aigeai, see Hatzopoulos 1996d, 264−269 (= Recueil 171−176). 68 Zancan 1934, 125−131. 69 Rosen (no date) 310.

  Who were the Macedonians? what degree Herodotus’ Macedonian logos enables us to reconstruct the precise itinerary of the Argead Macedonians from Upper Macedonia to the site of Vergina. 70 From this point of view the publication of inscriptions from the sanctuary of the Mother of Gods Autochthonous at Leukopetra mentioning Aleb(a)ia “a village of Elemia” 71 was decisive for definitively placing Herodotus’ Lebaie on the western slopes of the Pierian mountains. The main archaeological site in this area is at Palaiokastron near Velvendos, but its important remains indicate a city rather than a mere kome. The site of Bravas down by the Haliakmon, or even better that of Sphikia, with its rich Late Archaic cemetery, 72 would make a more likely candidate. Starting from there it is possible to reconstruct the itinerary of the three Temenid brothers: they could have forded the Haliakmon somewhere near Polymylos (ancient Euia) and could have crossed over Mount Bermion through the Kastania pass and Leukopetra to the “Gardens of Midas” at or near Beroia. This would be “the other land of Macedonia” 73 which they acquired and from which “they issued forth and began to subdue the rest of Macedonia as well”. 74 Herodotus left the story incomplete, but there is no doubt that the next step was the founding of Aigeai, by Perdikkas, the youngest brother, who named his capital after the goats that he had tended at Lebaie. A further question concerns the profile of those who became masters of Lower Macedonia. Ethnography has provided precious clues for understanding what sort of people they were socio-economically. 75 Until a century ago in Ottoman Macedonia economic and social conditions had hardly changed since antiquity. On the foothills of Mount Titarion the main industry was transhumant breeding of sheep and goats. “The shepherds moved thousands of stock in late April or early May from the plains of Lower Macedonia and Thessaly to the summer pastures of Upper Macedonia and drove them back to the plains in late October, sometimes covering scores of kilometres. For instance, the shepherds of Karitsa in Pieria had their summer pastures at Phteri on Mt. Titarion, those of Kokkinoplos, near ancient Pythion in Thessaly, at Kalyvia Kokkinoplou on Mt. Olympus, while those of Kalyvia Kokavas in Pieria had theirs in Polydendron in

 70 For what follows, see Hatzopoulos 2003a, 171−218. 71 Petsas et al. 2001, nos 12 and 106. 72 Kottaridi/Brekoulaki 1999, 111−114. 73 Herod. 8.138.2: οἱ δὲ ἀπικόμενοι ἐς ἄλλην γῆν τῆς Μακεδονίης ... 74 Herod. 8.138.3: ἐνθεῦτεν δὲ ὁρμώμενοι ὡς ταύτην ἔσχον, κατεστρέφοντο καὶ τὴν ἄλλην Μακεδονίην. 75 Wace/Thompson 1910.

The speech of the ancient Macedonians  

Emathia”. 76 Based on such data first E. Ivanka and then Hammond argued that in Geometric and early Archaic period transhumant pastoralists led a similar life moving back and forth between Mount Olympos and the Pierian Mountains and the plains of Thessaly, Pieria and Emathia, until under the leadership of a new dynasty they took the bold step of settling permanently at the site of Vergina on the edge of the great Macedonian plain. 77 The “wandering Perrhaibians”, 78 who roamed between Mount Olympus (or Mount Titarion) 79 and the Pindos chain, 80 offer a comparable ancient paradigm, which legitimises Hammond’s approach. A recent epigraphic discovery, the royal diagramma on military service, which uses the term πυρόκαυσις (“camp fire”) in the sense of “household”, provides an indirect confirmation of the transhumant pastoral past of the ancient Macedonians. 81 The above ethnographic model has been further elaborated 82 and used for the interpretation of linguistic phenomena. 83

. The speech of the ancient Macedonians The speech of the ancient Macedonians has been a problem which has bedevilled scholars for more than two hundred years, because until very recently not one sentence (with one uncertain exception) 84 of the local tongue had come down to us. The reasons were a) that Macedonian, like many other Greek dialects, never attained the dignity of a literary language, with the result that Macedonian authors, like their Thessalian- or Arcadian-speaking colleagues, wrote their works in Attic koine; b) the Temenid kings, when they set up a chancery, they adopted the Attic dialect, which from the fourth century (if not earlier) was gaining ground all around the Aegean basin, as their administrative language. Thus, the only sources at the disposal of scholars for acceding to Macedonian were the

 76 Hatzopoulos 2003a, 213−214, based on information from G. Kontogonis’ 1/200,000 map, Larisa sheet. 77 Hammond 1989b, 1−4. Cf. Ivanka 1950, 349−361. 78 Str. 9.5.12: Περραιβοὺς μετανάστας ἀνθρώπους. 79 Str. 9.5.20. 80 Cf. Str. 9.5.22. 81 Hatzopoulos 2001a, 91−98. 82 Hatzopoulos 2003a, 213−218; cf. Hatzopoulos 2006a 47−49. 83 Hatzopoulos 2007c, 173−176; Hatzopoulos 2013b, 204−221; Hatzopoulos 2018a, 314−316; Reservations in Helly, 2007, 177−222. Brixhe 2018, 21−22. 84 Strattis, Macedonians (Kock i. 719) in Athen. Deipn. 7.323b: κέστραν μὲν ὔμμες, ὡττικοί, κικλήσκετε.

  Who were the Macedonians? “glosses”, i.e. terms collected by lexicographers for their rarity or their strangeness in literary works, 85 and proper names, which in Greek are formed from appellatives (for instance Θεόφιλος, “he whom god loves”, is composed from the appellatives θεός and φίλος). The glosses, words rare and strange (often borrowed from a non-Greek language) by definition, had the major flaw of being easily corrupted when transmitted through the centuries by copyists who did not recognise them. Moreover, because of the criteria by which they were collected, they give a very slanted image of a tongue. As for the proper names, in the absence of modern corpora of Macedonian inscriptions, there existed no reliable collections until a few years ago.

.. The ‘Prehistory’ J. N. Kalléris, in his pioneer work on the ancient Macedonians, has written a panorama of the early history of the Macedonian linguistic question. 86 It will suffice here to recall the main points. The first modern study on the speech of the ancient Macedonians was written by F. G. Sturz and published under the title De dialecto macedonica et alexandrina liber in Leipzig, in 1808. It was essentially an alphabetical list of genuine or putative Macedonian words, without any attempt at etymological analysis. In spite of the scarcity of the evidence and the unreliability of the method, Sturz concluded that Macedonian was a Greek dialect akin to Dorian. A few years later K. O. Müller published a short study Über die Wohnsitze, die Abstammung und die ältere Geschichte des makedonischen Volks. Eine ethnographische Untersuchung (Berlin 1825) as a byproduct of his two-volume major work on the Dorians published the previous year. 87 His conclusion was that the Macedonians were of Illyrian descent. His approach was the following: The peoples that lived in or near the Macedonian area were the Thracians, the Paionians, the Illyrians and the Greeks. A Greek origin of the Macedonians is excluded, since the Greeks considered the Macedonians as barbarians. The Paionians are also excluded, since the Macedonians were never designated as members of the Paionian group of peoples in our sources. Therefore, the choice is only between the Illyrians and the

 85 It is significant that Hesychius’, Lexicon in which most of them are preserved, survives in a single, seriously corrupt manuscript. 86 Kalléris 1954, 20−47. 87 K. O. Müller, Die Dorier (1824). It appears as “Appendix I” in the English translation of this work, published in London in 1839.

The speech of the ancient Macedonians  

Thracians. Now, the cradle of the Macedonians was situated far from Thrace, but close to the Illyrians (within which Müller subsumed the Epirotes). Besides, Strabo (7.7.8) states that “some persons, moreover, give to the whole country as far as Corcyra the name of Macedonia, assigning, as their reason, that the inhabitants nearly resemble one another in the mode of wearing the hair, in their dialect, in the use of the chlamys, and in other points of this kind: some of them likewise speak two languages”. 88 Müller mistakenly deduced from this passage that Macedonians and Illyrians wore the same costume, had the same hair style, and spoke the same language. What Strabo actually meant was that Macedonians and Epirotes had much in common and spoke similar Greek dialects, and that on the linguistic frontier between Greek and Illyrian some people were bilingual. 89 Finally, in the linguistic field, Müller asserted that Macedonian was unintelligible to Greeks, that the little we know of Macedonian vocabulary can be certainly identified as Illyrian, and that the Macedonians substituted B for Φ and Δ for Θ, because they could not pronounce the Greek aspirates, displaying thereby an Illyrian trait. This basic Illyrian element was eventually ‘sprinkled’ with Greek borrowings under the influence of Athens. This is in a nutshell the theory of a ‘mongrel’ Macedonian tongue, which has haunted the issue in various versions until our days. The controversy continued indecisively with alternate pro and contra contributions. A. Fick, taking into consideration not only words but also personal names, asserted the purely Greek character of Macedonian. 90 G. Meyer responded by an analysis of Macedonian words listing them in three categories: Greek, barbarian and doubtful, in view of which he concluded that the Greek character of the Macedonian could not be ascertained. 91 P. Kretschmer followed a middle road, considering that the Macedonian tongue had remained at a stage of evolution which the other Greek dialects had left behind and which is patent in the substitution of (unvoiced) ‘aspirates’ by mediae (voiced) stops (Β for Φ, Δ for Θ), adding that its isolation had been aggravated by intermixture with Illyrians and Thracians. 92 In fact, Kretschmer took over from Müller a) the concept of the “mongrel language” (Mischsprache) and b) the decisive importance of the treatment of the Greek ‘aspirates’ as a criterion of the character of the Macedonian idiom. 93 He

 88 Translation by H. L. Jones in the Loeb Classical Library (1924). 89 Hatzopoulos 1997b, 140−145 (= Recueil 533−540); cf. Hatzopoulos 2007a, 55, n.18 (= Recueil 405, n. 18); Hammond in Hammond/Griffith 1979, 43. 90 Fick 1864, 718−729; Fick 1874, 193−235. 91 Meyer 1875, 185−192 (non vidi). 92 Kretschmer 1896, 283−288. 93 The so-called ‘aspirates’ were actually ‘expirates’.

  Who were the Macedonians? attributed the allegedly divergent treatment of the Indo-European mediae aspiratae (voiced ‘aspirate’ stops) to an early separation of Macedonian from the other Greek dialects. G. Hatzidakis contested Müller’s, Meyer’s and Kretschmer’s conclusions in a series of studies insisting that the Macedonian speech had gone through the same evolution as the other Greek dialects, and that the presence of mediae instead of ‘aspirates’ was a secondary late phenomenon resulting from the voicing of the former ‘aspirates’ which had already become spirants and were pronounced as in Modern Greek. 94 A counterattack soon followed. In 1905, in the first volume of his study on the Indo-Europeans, H. Hirt too argued that Macedonian was not a Greek dialect, because it was characterised by the evolution of Indo-European mediae aspiratae into simple mediae instead of unvoiced ‘aspirates’ as was the case with all the other Greek dialects. Consequently Macedonian was either a separate Indo-European dialect or an Illyrian subdialect, but incomprehensible to other Illyrians because of its mixed character. 95 Two years later however, in the second volume of his study, Hirt admitted the possibility of a second Macedonian tongue, which could only have been Greek. 96 The theory of two Macedonian tongues was destined to experience a second coming in our days. The year between Hirt’s two volumes saw the publication of the first in-depth study of all the available evidence on the Macedonian language issue: O. Hoffmann’s monograph on the speech and the nationality of the Macedonians. 97 The systematic examination not only of stray words but also of all proper names then available led him to the conclusion that Macedonian was a Greek dialect akin to Thessalian. Hoffmann’s arguments convinced several of his colleagues who reviewed his book. 98 Others (A. Thumb, U. von Wilamowitz-Moellendorf, A. Debrunner, A. Meillet) expressed either their dissent (the first three), stressing “en passant” the unreliability of proper names as evidence of nationality and/or the notorious treatment of the Indo-European mediae aspiratae 99 or (the last one) the futility of the discussion given the lack of sufficient data, whereas the problem would be immediately solved if we obtained by chance a continuous text of

 94 Hatzidakis 1897, 5−57; Hatzidakis 1899, 131−157; Hatzidakis 1900a, 313−320; Hatzidakis 1900b, 15−154; Hatzidakis 1901, 33−114; Hatzidakis 1910−1911, 87−109. Cf. Hatzidakis 1928, 390−415. 95 Hirt 1905−1907, 149−150. 96 Hirt 1905−1907, 603. 97 Hoffmann 1906; Buck 1910, 288. 98 Lesny 1909, 297−302; cf. V. Lesny, in Christ’s Geschichte der griechischen Litteratur II 1, (Munich 19206) 2, n. 3. 99 Thumb 1909, 8−10, § 9; Wilamowitz-Moellendorf 18952, 11, n.23; Debrunner 1926, 516 § 20−21.

The speech of the ancient Macedonians  

ten lines. 100 Little did Meillet expect that his wish would come true 85 years later. Similar objections were subsequently recycled by other linguists. V. Pisani imagined an Indo-European dialect akin to Osco-Umbrian and to Greek, which was eventually contaminated by the languages of peoples living in the same geographical area as the Macedonians and which produced a mixed language distinct from Greek, Thracian and Illyrian. 101 Two years later Ed. Schwyzer returned more or less to Kretschmer’s hypothesis of the separation of Macedonian from the other Greek dialects at a time when they all retained the Indo-European mediae aspiratae and of the subsequent divergent evolutions, the Macedonian converting them into simple mediae instead of unvoiced ‘aspirates’. The distance between Macedonian and Greek was allegedly further increased by the admixture of phonetic and morphological traits originating from the conquered populations. 102 Actually, Kalléris’ magisterial ‘summa macedonica’ partakes both of the ‘prehistoric’ (of which it is the swan song) and of the modern phase of Macedonian linguistics. It is prehistoric, because it uses the same, limited, non-updated corpus of often unreliable glossae and proper names, and devotes to them, and particularly to an inextricable discussion on the Macedonian voiced mediae, a large number of pages (57−461), justifying thus Meillet’s cynical remark on the futility of discussing at length without sufficient data a problem that ten lines of continuous text would immediately solve. 103 It is modern because it is at the same time holistic, in examining linguistic phenomena non in the abstract, but within their geographic, historical and institutional context, and in making the necessary selections, by putting the existing data in order, and particularly by freeing them from most of the extraneous elements which had accrued to them over the years. J. N. Kalléris died in 1992 without finishing his magnum opus. His second volume was published in 1976. A third volume had been planned to include the second part of the third chapter devoted to “Ways and Customs”, including “Social customs” and “The private life of the Macedonians”, a fourth chapter devoted to “Historical Evidence” and a “Conclusion”. He had only time to publish privately a very full and precious Index of the first two volumes. Thus Kalléris was unable to exploit the revolution in Macedonian studies initiated in the 1970s. Had he been able to take advantage of it fully, he might have avoided suggesting the untenable theory that the substitution of common Greek ‘aspirates’ by mediae in

 100 Meillet 19303, 57−58. 101 Pisani 1937 8−31. 102 Schwyzer 1939, 69−71. 103 Cf. Kalléris 1954, 53.

  Who were the Macedonians? Macedonian was a fiction invented by grammarians, since, according to him, it concerned only five words or proper names (Ξανδικός, ἀβροῦτες, ἀδραία, Βερενίκη and Κεβαλῖνος). 104 So Kretschmer’s and Schwyzer’s hypothesis about the retention of the Indo-European mediae aspiratae in Macedonian was recycled again down to the end of the last century. To the same twilight period of Macedonian studies, when the exploitation of existing linguistic data and the emergence of new ones had not yet been assimilated by scholars working on Macedonia, belongs the case of two American ancient historians, Ernst Badian and Eugene N. Borza. 105 Harvard professor Ernst Badian was the doyen of ancient historians (both Greek and Roman) in America and possessed incomparable influence and patronage. Outraged by Andronicos’ reaction to the publication by Phylis Williams Lehmann (widow of the excavator of Samothrace Karl Lehmann) of an article challenging the Greek archaeologist’s dating and attribution of Tomb II at Vergina (see below), he decided to write a paper which would teach a lesson to the Greek-American backers from Time-Life and to the archaeologists from the Greek Ministry of Culture who organised the National Gallery exhibition The Search of Alexander the Great, since, according to him, they had sacrificed scholarly ethics to further their personal or political agendas. In effect, it was rumoured that Andronicos had demanded and obtained Phylis Lehmann’s replacement from her position as consultant in the forthcoming exhibition. Badian’s brilliant paper, “Greeks and Macedonians” was the first to be read at the symposium which took place on the occasion of the exhibition at Washington in November 1980 and was duly published in the volume of the acta. 106 It was aimed at the tender spot where it would most hurt the Greek delegation, officials as well scholars: the unjustified (according to him) Greek presumption to appropriate the Macedonian heritage. The attack was masterly conceived and executed. Its three main axes were the following: a) Badian disqualified a priori Greek scholarly works on account of the alleged bias shown by Greek historians and archaeologists, who, contrary to their foreign colleagues, cannot help being distracted by politics and nationalism. Even worse, Greeks scholars would readily indict their foreign colleagues for the

 104 Kalléris 1967, 457−461. 105 Cf. Crossland 1982, 843−847, who erroneously asserts that “there are no substantial inscriptions in Greek from Macedonia earlier than the third century”; Bonfante 1987, 83−85; Borza 1995, 17−24; Borza 1999, 42, n. 27, who writes “Thus it would seem that no progress has been made on the issue of the native language of the Macedonians”. For my appraisal of Borza’s recent scholarly production, see below and also Hatzopoulos 1999a, 226−239; Hatzopoulos in BE 2000, 431. 106 Badian 1982, 33−49.

The speech of the ancient Macedonians  

very failings of which they are themselves guilty. 107 For instance, Badian felt free to accuse a reputed Greek scholar (Kalléris) of knowingly concealing a historical fact (according to Badian), namely that a certain Xennias figuring on a papyrus 108 was a Macedonian, given that his name was written with a double consonant. However, Badian should have known that the form with the double consonant is not encountered in Macedonia, whereas it is at least three times attested in Boiotia. 109 Moreover, he inflated out of proportion a trivial typographic mistake in a date (359 instead of 399) in a catalogue of exhibits, by way of excusing it as such, while simultaneously insinuating that it was not just an innocent oversight! 110 b) Badian himself affected incompetence to judge and indifference to find out whether the Macedonians were Greek or spoke a Greek dialect. Nonetheless, he marshalled every possible (relevant as well as irrelevant) 111 piece of evidence, which might be used to undermine the Hellenic character of the Macedonians, while underplaying all testimonies to the contrary, and also advanced arbitrary interpretations of evidence that did not suit him. 112 c) He dissembled his strong bias by repeatedly pretending that he was not discussing historical facts but “perceptions” 113 and “sentiments”, 114 which incidentally present the incomparable advantage of not being susceptible of confrontation with hard facts, archaeological or linguistic, but are amenable to subjective interpretations. All in all, Badian’s contribution to the symposium was a most effective piece of sophistry in the service of what he surely considered a noble cause. 115  107 Badian 1982, 33 and n. 2. 108 PSI 1284, 19. 109 Badian 1982, 41 and n. 66. The only Xennias registered as Macedonian in LGPN IV –and that with a question mark– is precisely the one figuring on the papyrus PSI 1284, 19, and who is in fact of unknown origin. For the attestations in Boiotia, see LGPN III.B 315. The American scholar, adding insult to injury, did not hesitate to give an arrogant lesson of scholarly proprieties to his Greek colleague: n. 68: “Political sentiments based on the survival of nineteenth-century nationalism ought to have no place in academic discussion”. 110 Badian 1982 47, n. 24: “Rarely has a misprint conveyed so much disinformation!”. 111 Badian 1982, 35: “We find him (Alexander I) described in the lexicographers, who go back to fourth century sources, as ‘Philhellen’ – surely not an appellation that could be given to an actual Greek”. Had he only opened his Liddell & Scott, he could have read: “φιλέλλην ... also of Hellenic tyrants as Jason of Pherae, Isoc. 5.122: generally of Hellenic patriots, Pl. R. 470e; of Hipocrates, Sor. Vita Hippocr. 8; καλὸν Ἕλληνα ὄντα φιλέλληνα εἶναι X. Ages. 7.4”. 112 Badian 1982, 37 and n. 28. 113 Badian 1982, 33. 114 Badian 1982, 42. 115 It may seem that such a detailed analysis of Badian’s article is out of proportion. However, it is not gratuitous, for it aims at highlighting a pattern of reaction based on prejudice, which is typical of some of us when we are confronted with new facts unsettling conventional wisdom.

  Who were the Macedonians? E. N. Borza, Professor of Ancient History at Pennsylvania State University, was in the early eighties of the last century a rising star in Macedonian studies. He had been a historical consultant of the Washington exhibition and had written a very well-documented and optimistically forward-looking introductory paper to the National Gallery symposium under the title “The History and Archaeology of Macedonia: Retrospect and Prospect”. 116 Badian discovered in Borza a kindred spirit and a worthy successor, at least in things Macedonian, and extended to him his patronage, thus helping him to become president of the Association of Ancient Historians, which he had himself founded. 117 As he recalls: “About thirtyfive years ago I started on what at the time was a formidable task: an interpretation of Alexander the Great that should as far as possible be devoid of mythology, ancient and modern. I had no thought of extending this any further. It is chiefly Gene’s merit that a recognizable historical interpretation of the history of classical Macedonia has not only become possible, but is now accepted by all ancient historians who have no vested interest in the mythology superseded by Gene’s work. Needless to say I welcome and agree with that approach and have never disagreed with him except on relatively trivial details of interpretation”. 118 Borza had indeed shown a great interest in Macedonia, which he visited several times, while he kept abreast of archaeological discoveries and wrote about its past and its present in a simple but elegant and attractive style. Actually, Borza also shared Badian’s preference for immaterial perceptions and sentiments instead of facts, 119  We shall encounter it again and again in the following chapters in questions not only of language but also of religion, institutions or customs of the ancient Macedonians, even in the “unending controversy” about the Vergina tombs. At its origin, I suspect a misplaced Athenocentric philhellenism that could but abhorr the Macedonians who put an end to ‘Greek Liberty’ (or Athenian imperialism) at Charoneia in 338. Having espoused the Athenian point of view, many scholars cannot conceive the Macedonians otherwise than as crass barbarians living in hovels and ignorant of any law besides that of naked force. When gradual but accumulating discoveries reveal a different reality, it is easier to ignore it or brush it aside taxing those who bring it forth with ulterior motives (preferably nationalistic) rather than imagine (and even less adopt) a different paradigm. Thus, the Badian-Green-Borza dogma (for the association of the three historians, see Borza 1999b, 264, n. 23) of the “ungreekness” of the Macedonians based on ignorance of new data has spread across American academia (cf. Pomeroy et al. 1999, 373−374). 116 Borza 1982b, 17−30. 117 Borza was elected (1984−187) and reelected (1987−1990) president of the Association of Ancient Historians, a key position for wielding influence over his younger colleagues. 118 Thomas 1995, 2. The aspersion that all scholars who do not agree with Badian and Borza have vested interests in mythology is typical of Badian’s odium historicum. 119 Cf. Borza 1982b, 11 (= Makedonika 118): “a comment which, regardless of Demosthenes’ intent, was surely designed to play upon existing scepticism”. Demosthenes’ unquestionable intent is minimised in view of underlining the imagined scepticism of the Athenians.

The speech of the ancient Macedonians  

his method of imperceptible shifts from opinions to certitudes, 120 and his cursory dismissal of modern Greeks as scientifically unreliable nationalists. 121 In linguistic matters, however, Borza, who was not a philologist, initially confined himself to a prudent agnosticism. His negative (but important) conclusion was that, in the absence of adequate evidence, one could not legitimately affirm that Macedonian was a Greek dialect. Nevertheless, from 1990 onwards he ventured into linguistics. 122 This new interest went hand in hand with a growing involvement in the political and diplomatic aspects of the ‘Macedonian question’. 123 Unfortunately, enthusiasm could not compensate for lack of the necessary technical background and for the difficulty of keeping abreast of developments recorded in specialist studies published in a variety of languages, as it is particularly obvious in the terminology he uses 124 and in his amazement and perplexities. 125 Having no prior experience in the process of acculturation of ancient peoples, he thought that his hypothesis of “Hellenization” could offer a satisfactory explanation for the use of Greek by the Macedonians, 126 and he alleged the example of the Thracian silver vessels on which “Thracian personal and place names are inscribed in Greek”, 127 failing to see the capital difference from the contemporary Macedonian inscriptions in which the names written in Greek are Greek names and not names  120 Borza 1982a, 12 (= Makedonika 119): “Given the historical milieu in which these tales were spawned and then adorned, a denial of myth seems prudent. The Temenidae in Macedon must disappear from history”. 121 Cf. Borza 1999a, 34: “One needs read only a representative sampling of modern Greek archaeological literature about Macedonia to see that some claims of Hellenic origin are dangerously close to what has been described as “nationalistic archaeology”; cf. Borza 1982a, 13, n. 28 (= Makedonika 123, n.28). 122 Borza 19922, 92−94 and 305−306; Borza 1995, 18−21; Borza 1999a, 41−43. For what follows, see Hatzopoulos 1999a, 227−233. 123 Cf. Borza 1999b, 249−266. 124 For instance he employs the term “standard Greek” in order to designate the Attic koine, as well as the totality of Greek dialects as opposed to “Macedonian”, or even the Macedonian dialect itself. Cf. Borza 19922, 93: “those few inscriptions that have been recovered are written in standard Greek” (meaning Attic koine); Borza 1995a, 20: “standard Greek in several dialects was used for formal expression in speech and writing”; Borza 1995a, 21: “the tablet was written in standard Greek” (meaning the tablet in Macedonian dialect; see below). 125 For instance Borza 19922, 306: “As a question of method: why would an area three hundred miles north of Athens – not colonized by Athens – use an Attic dialect, unless it were imported?” Apparently Borza was not aware of the phenomenon of ‘Atticisation’ which affected the whole Aegean basin in the fifth and even more the fourth century. See Brixhe/Panayotou 1988, 245−260; Panayotou 1990, 191−228. 126 Borza 1995a, 18−19. 127 Borza 1995a, 20.

  Who were the Macedonians? of some foreign tongue. Had he studied contemporary cases of Hellenisation, he would have known that in such environments non-Greek names massively survive long after the adoption of Greek as a communication medium by a non-Greek population group. Moreover, in the case of ancient Thracians, as with that of most ancient Hellenised peoples (Phrygians, Carians, Lycians etc.), the existence of the indigenous language is also attested by texts (however few or short) written in that language, whereas in the cradle of the Macedonian kingdom there is no trace of any other language and no foreign names, except some stray Illyrian or Thracian ones. 128 East of the Axios the situation of course was very different. The use of local names, both non-Greek and typically Ionic Greek, is perpetuated centuries after the annexation of these areas by the Macedonian kings. 129 The parallel phenomenon of the survival of Atticised Ionic as lingua franca in these same areas –and not the radically foreign nature of Macedonian, as Borza thinks 130– explains the choice of Attic, instead of Macedonian, by Philotas in his defence, so that persons not hailing from the Macedonian heartland, but from the ‘New Territories’, “will understand more easily”. 131 What is also patent in Borza’s later works is the sumptio quaerendi, 132 which reflects a strong bias, the very bias of which Borza accuses his Greek colleagues. 133 It is true that one cannot a priori exclude that extra-scientific considerations may influence the scholars who have expressed opinions or convictions regarding the origin of the idiom used by the ancient Macedonians. Greeks, however, are not

 128 Hatzopoulos 2000b, 99−117. 129 Cf. Hatzopoulos/Loukopoulou 1992, 117−122. 130 Borza, 1995a, 19. 131 Curtius 6.9.35, my emphasis. As Hatzidakis 1897, 28, and Beloch 1897, 200 pointed out long ago, “more easily” means that non-Macedonian Greeks could understand Macedonian, but less easily than Attic Greek, which implies that Macedonian was a Greek dialect and not a completely different language, as Borza would have it. For a description of the process by which the Macedonian dialect was progressively abandoned in favour of Attic Greek, see Brixhe 2018, 21−23. Were Macedonian not a Greek dialect, but a completely foreign tongue, the Attic (or Ionic) speakers in Alexander’s army would not understand it at all. 132 Borza 1999a, 42, n. 27: “There is insufficient amount of ‘Macedonian’ –by that I mean a nonHellenic language– surviving to know what language it was”. I wonder what he means by “insufficient amount”. There is no trace at all of another language current in the Macedonian kingdom besides Greek before the Roman conquest. As it is clear from the same note, Borza was aware of the existence of the Pella tablet, but he a priori brushed it aside, because it is written in a Greek dialect. 133 Borza 19922, 90−91.

The speech of the ancient Macedonians  

exclusively concerned by such failings. After all, other Balkan nations, even European Great Powers, have been also implicated in the Macedonian imbroglio. 134 One might see in this insistance on the lack of objectivity of Greek scholars a defense mechanism for avoiding recognising one’s own prejudices and reluctance to face new data.

.. The ‘pentekontaetia’ revolution The revolution in Macedonian studies was due to four independent but concomitant and interacting developments: a) the institution by the Thessalonike Society for Macedonian Studies of the Ancient Macedonia quinquennial international meetings of scholars from all over the world, the proceedings of which started being published from 1970 onwards; b) the coming to fruition of long-term works by reputed scholars, such as the publication in 1972, after 36 years of labour, of the first volume of the inscriptions of Macedonia by Charles Edson in the Inscriptiones Graecae, and the ongoing publication from that same year onwards of N. G. L. Hammond’s monumental, three-volume History of Macedonia; c) the sensational discovery of royal tombs during the excavation of the Great Tumulus at Vergina by Manolis Andronicos starting by 1977; d) the inception in 1979 of research activities aiming at the systematic publication of the epigraphic corpora of Macedonia and Thrace at the Centre for Greek and Roman Antiquity (KERA) of the National Hellenic Research Foundation (NHRF). 135 I had the privilege of collaborating both with Andronicos for the publication of the collective volume Philip of Macedon (Athens 1980), and with Hammond for the exploration of the Via Egnatia, and at the same time to be in charge of the Macedonian epigraphic project at KERA. Andronicos, besides being the excavator at Vergina, was also a member of the Governing Board of the NHRF. It was the collection of thousands of inscriptions from Macedonia that renewed, indeed revolutionised, Macedonian philology and linguistics. They were published by researchers of KERA, usually in collaboration with archaeologists from the Archaeological Service, in a series of corpora 136 and in the monograph

 134 Cf. Kalléris 1954, 43, n. 1. 135 Cf. Hatzopoulos 1981, 91−108. 136 EAM; EKM I; EKM II.

  Who were the Macedonians? series ΜΕΛΕΤΗΜΑΤΑ. 137 Meanwhile Argyro Tataki produced a succession of onomastic and prosopographic studies, 138 which greatly contributed to the publication by P. M. Fraser and Elaine Matthews of Volume IV of the Lexicon of Greek Personal Names (Oxford 2005) devoted to Macedonia, Thrace and the Northern Regions of the Black Sea. The importance of these contributions for the study of the tongue of the ancient Macedonians cannot be overestimated. For the first time scholars interested in Macedonian linguistics had access to a reliable collection of primary epigraphic evidence and of personal names from Macedonia. Many, such as O. Masson, Cl. Brixhe, L. Dubois, Anna Panayotou, J. Méndez Dosuna, E. Crespo, and others, made full use of them. 139 The access to such a vast sample of written documents from Macedonia, notwithstanding that almost all were written in koine, proved valuable, because it enabled the identification of cases where the author (or the stone cutter) of an inscription let his native tongue come to the surface, unwittingly substituting a koine form by a dialectal one. It all began with a passage from Philip V’s late third (or early second) century diagramma on service in the field army from Amphipolis, 140 where we read ΤΩΙ ΔΕ ΧΕΙΡΙΣΤΑ ΜΗΔΕΝ ΔΙΔΟΣΘΑΙ. It was obvious that the dative ΧΕΙΡΙΣΤΑ was a lapsus for the expected koine form ΧΕΙΡΙΣΤΗ(Ι). This slip confirmed that Macedonian, like most Greek dialects, retained the /a/ sound of the long /a/ of Common Greek, which in the Attic and Ionic dialects had evolved to long /e/, before ending up in /i/. Moreover it enabled us to reconstitute the paradigm of the first masculine declension. Subsequently other similar cases

 137 Gounaropoulou/Hatzopoulos 1985; Hatzopoulos 1988a; Hatzopoulos 1988b; Hatzopoulos/ Loukopoulou 1989; Hatzopoulos 1991; Hatzopoulos/Loukopoulou 1992; Hatzopoulos/ Loukopoulou 1996; Hatzopoulos 1994c; Hatzopoulos 1996b; Hatzopoulos 1996c; Petsas et al. 2000; Hatzopoulos 2001a. 138 Tataki 1988; Tataki 1994; Tataki 1998; Tataki 2006. 139 E.g. Brixhe, 1999, 41−71; Brixhe 2017−2018, 1862−1867; Brixhe 2018, 9−16; Brixhe/Panayotou, 1988, 245−260; Brixhe/Panayotou 1992, 129−135; Brixhe/Panayotou 1994, 206−222; Crespo 2018, 329−348; Hatzopoulos 1987b, 397−412; Hatzopoulos 1994a, 249−254; Hatzopoulos 1998, 1189−1218 (= Recueil 261−279); Hatzopoulos 1999a, 225−239; Hatzopoulos 2000b, 99−117 (= Recueil 51−69); Hatzopoulos 2000c, 177−182; Hatzopoulos 2006a, 35−51; Hatzopoulos 2010b, 356−365 (= Recueil 417− 425); Hatzopoulos 2017a, 203−209; Hatzopoulos 2018a, 299−328; Masson 1996, 905−906; Masson 1998, 879−880 (= OGS III 292−295); Méndez Dosuna 2012, 133−145; Panayotou 1992a, 13−31; Panayotou 1992b, 181−194; Panayotou 1992−1993, 5−32; Panayotou 1996, 124−163; Voutiras 1998). 140 Hatzopoulos 2001a, 162, 3 A, Col. 3, L. 3.

The speech of the ancient Macedonians  

(κόρας=κόρης), 141 (κόραν=κόρην), 142 (παιδίσκαν=παιδίσκην), 143 (Καλᾷ θεᾷ=Καλῇ θεᾷ) 144 allowed us to do the same for the first feminine declension. A little later an idiomatic infinitive was recognised in a 232 A.D. manumission from Skydra: ἠξίωσαν οἱ θρέψαντες καλῶς δουλευθέντες ὑπὸ θρεπταρίου εἰδίου ὀνόματι Ὀνησίμαν, περὶ ἔτη ΙΗ, ἀνατιθήμειν θεᾷ Ἀρτέμιδι Γαζωρίᾳ ταύτη[ν]. 145 It was again obvious that ἀνατιθήμειν was the dialectal Macedonian form which had unwittingly replaced the koine infinitive ἀνατιθέναι. Thus, such idiomatic forms, wherever they could be spotted, became, along with glosses and proper names, a third source for the knowledge of the Macedonian tongue. Their great merit was that they offered the possibility of solving once and for all the issue of the voiced mediae which appear in certain Macedonian glosses and proper names instead of unvoiced ‘aspirates’ of the other Greek dialects. In a series of votive or manumission inscriptions from the territory of Aigeai the local deity Artemis appeared with the epithets Digaia (Διγαία) and Blaganitis (Βλαγανῖτις), from Blaganoi (Βλάγανοι), the place name of their location. 146 The first epithet was in fact the local form of the adjective δικαία, “the Just”, but what about Blaganoi and Blaganitis? A gloss in Hesychius, βλαχάν· ὁ βάτραχος, provided a first clue, which was confirmed by another inscription naming the deity to whom the dedication was made as τῇ [θεᾷ τῶν β]ατράχων. 147 Blaganoi was the place name “The Frogs”, probably because of its location near a marsh, a type of landscape dear to Artemis. 148 The forms Διγαία instead of Δικαία and Βλάγανοι instead of *Βλάχανοι showed clearly that the sporadic appearence of voiced mediae instead of unvoiced ‘aspirates’ had nothing to do with an early divergent evolution of the Indo-European mediae aspirates in Macedonian in relation to the other Greek dialects, but was a secondary and late phenomenon of voicing within Greek, which in Macedonian could affect all unvoiced consonants, both ‘aspirates’ and non- ‘aspirates’, and which occurred after the unvoiced ‘aspirates’ (/ph/, /th/, /kh/) had become unvoiced spirants (/φ/, /θ/, /χ/) and the voiced mediae (/b/, /d/, /g/) voiced spirants (/ᵬ/, /ᵭ/, /ǥ/), as G. Hatzidakis had stressed

 141 ΕΚΜ II 398. 142 ΕΚΜ I 56. 143 ΕΚΜ II 168. 144 Hatzopoulos 1994c, 44. 145 ΕΚΜ II 121. 146 ΕΚΜ II 62−78. 147 ΕΚΜ II 65. 148 Cf. Artemis Limnaia (Paus. 2.7.6; 3.14.2); Artemis Limnatis (Paus. 3.23.10; 4.4.2); etc.

  Who were the Macedonians? more than a century ago. 149 It was just a banal phenomenon of voicing affecting unvoiced consonants known from many other languages and dialects. Meanwhile Macedonian vocabulary was enriched by other technical terms of indisputable Greek etymology but unknown in the rest of Greece. They were often attested in a specific Macedonian phonetic form in inscriptions otherwise written in koine, such as the epithets of Dionysos Ἐρίκρυπτος and Ψευδάνωρ or the verbs νεύω and ἀρχινεύω. 150 Another sequence of epigraphic discoveries confirmed several glosses of Hesychius concerning religious vocabulary (Δάρρων = Θάρρων, a healing deity close to Asklepios; Περίτας = Φύλαξ, an epithet of Herakles), or institutional vocabulary of the Macedonians (πελιγάν, πυρόκαυσις) 151 exclusively Macedonian and of impeachable Greek etymology. The etymology of some Macedonian months (Xandikos, Gorpiaios, Audnaios/Aidonaios), the Greek origin of which had been unduly contested, was incontrovertibly established. 152 New personal names (e.g. Ἱκκότας), place names and ethnics (e.g. Κραννέστης) revealed dialectal forms of Greek words which Macedonian shared with other Greek dialects, such as ἵκκος (= ἵππος) and κράννα (= κρήνη). 153 The above examples taken together constituted a significant progress which allowed to sketch out an outline of the Macedonian dialect. What was missing was a continuous and somewhat substantial dialectal text, which A. Meillet demanded nearly a century before. In fact shorter texts did exist. Queen Eurydika’s mid-fourth century votive inscriptions from Aigeai did not follow Attic but Macedonian phonetics and morphology: Εὐρυδίκα Σίρρα Εὐκλείαι. 154 The same is true of a late fifth or early fourth century consecration fom Dion: Πραξιδίκαι καὶ [Ἑρ]μᾶι Τύχωνι, 155 and of a third century funerary inscription from a kome of Thesalonike: Πισταρέτα Θρασιππεία κόρα. 156 Finally, Meillet’s wish was granted in 1998 with the publication of an early fourth-century dialectal defixio of fairly well

 149 Many similar cases have been collected since then. Cf. Φάλακρος/Βάλακρος/Βάλαγρος, Διονύσιος/Διονύζιος etc. 150 Hatzopoulos 1994c, 44; 65−71. 151 Hatzopoulos 1998, 1195−1196 (= Recueil 267−268); Hatzopoulos 2003c, 54−55 (= Recueil 354−355). 152 Hatzopoulos 1999a, 237−239. Today I would not hesitate to connect the month Δύστρος with the verb θύω