Chapter 18. ΗΕΔΡΟΛΟΣ-ΑΡΡΩΛΟΣ
p. 229-236
Note de l’auteur
My warmest thanks are due to A. Tziafalias, Director of the Larissa Museum, who gave me permission to publish the funerary inscription from Atrax and sent me excellent photographs of that monument, to Professor Claude Brixhe, who was willing to discuss some delicate questions of Greek phonetics, and to my colleague at KERA Garth Fowden, who had the patience to read and to improve my English.
Texte intégral
1The Athenian Tribute Lists name five times in the ἐπὶ Θράικης panel an otherwise unattested city of the ΗΕΔΡΟΛΙΟΙ or ΧΕΔΡΟΛΙΟΙ. The Community appears for the first time in 448/447 under the name ΗΕΔΡΟΛΙΟΙ (IG, I2, 264, III, 2) paying 500 drachmas, then in 447/446 (IG, I3, 265, II, 33). Subsequently it disappears for twelve years. It is notably absent from the fully preserved list of 443/442. It resurfaces in the form ΧΕΔΡΟΛΙΟΙ in the 435/434 list (IG, I3, 277, VI, 25) and is also registered in the 434/433 one (IG, I3, 278, VI, 16). Finally, it makes a last appearance in the 433/432 list (IG, I3, 279, II, 88), perhaps as (ἄτακτοι. In the last two years the Hedrolioi/ ΧΕΔΡΟΛΙΟΙ are listed as πόλεις αὐταί ταχσάμεναι paying 1000 drachmas, double the amount of the initial tribute. The city has been variously located in Mygdonia, near the mouth of the Echedoros3 and in Chalkidike3. The fact that the Hedrolians are named only in the panels of the forties and then again after 437 and before 432, that is to say first when the Athenians were particularly active in the Thraceward area and then again only after the foundation of Amphipolis but before the revolt of Potidaia, the Bottiaians and the Chalkidians, has been taken as an indication that the city was situated in the periphery of the Athenian sphere of influence and accordingly the editors of the Athenian Tribute Lists and later scholars have sought Hedrolos in the hinterland of the Chalkidic peninsula4. I had myself suggested that it might figure as Γ[έδ] I ρωλος in the border settlement by Philip II between small townships in the vicinity of the Thermaic Gulf and the Mygdonian lakes in the northernmost part of the peninsula5. However, that suggestion was superseded by the revelation of the existence of a lake Πυρρωλία situated precisely in that area and obviously owing its name to a city Πύρρωλος, which should be restored in the inscription6. Thus the question of the identification of the site of Hedrolos remained as unsolved as ever.
2A series of at first sight unconnected developments, namely the discovery of an inscription mentioning a grant of land by the city of Berge at Neos Skopos near Serres, the publication of a letter of Galerius and Maximinus to the IIIIviri and the decuriones of the Herakleotai unearthed at Muletarevo, just north of Mount Kerkine and the Greek-Bulgarian border, and, finally, the reconsideration of the geographical setting of Alexander the Great’s answer to the ambassadors of Philippoi, encouraged me to revisit the whole question of the historical geography of the Strymon valley and specifically of its Roman roads perpendicular to the Via Egnatia which are recorded in the Peutinger Table. The first-mentioned document definitely established that Berge was not an “indigenous”–however Hellenized–settlement situated on the western (as everybody, including myself, had believed), but a Thasian foundation on the eastern bank of the Strymon. The second document provided the exact location of the starting point of the two Roman roads, which until then was thought to have been situated at various distances south of Mount Kerkine and the Rupel defile. The reinterpretation of the third document showed that Mount Dysoron was to be identified with Mount Menoikion and Lake Prasias with the now drained Lake of Pravi in the Philippoi basin, and more significantly, it lead to the conclusion that for a certain period after 479 Alexander I of Macedon controlled the mines of Mount Pangaion. In this way there emerged a wholly new network of ephemeral alliances and recurrent conflicts between Thasians, Athenians, Macedonians and Thracians for the possession of the fabulous resources in all sorts of raw materials of the Strymon valley and the Philippoi basin.
3Leaving to O. Picard the disentanglement of this particularly complex page of political, but also numismatic, history7, I focused on the geographic side of the question. I verified the presence of four settlement sites attested as poleis through epigraphic or numismatic evidence on each side of the lower Strymon valley between Mount Kerkine and Amphipolis. On the western bank from north to south: 1) Assari near Kalokastron (Sakfatsa)8, 2) Palaiochori near modern Vergi (Kopatsi)9, 3) Paliokastron near Terpni (Tserpista)10, 4) Prodromos near Aïdonochori. On the eastern bank from north to south: 1) Siderokastron (Demir Hissar)11, 2) Serres12, 3) Tsifliki near Neos Skopos (Toumbitsa)13, 4) Hagios Athanasios near Gazoros (Porna)14. Of these three had already been securely identified: Serres and Gazoros with the contemporary towns of the same name and Aïdonochori with Tragilos. Neos Skopos had long been unanimously identified with the κώμη Ὀλδηνῶν) named on an inscription discovered at the site15. The site of Siderokastron has been variously equated with ancient Herakleia Sintike and Skotoussa16, the one near Terpne with Berge, the site near Vergi with Herakleia Sintike17, Berge or Euporia18, and the one at Kalokastron with Arrolos/Arason19.
4The localisation of ancient Berge at Neos Skopos upsets the geographic puzzle presented by this region, not only leaving the κώμη Ὀλδηνῶν) without a location, but also nullifying the identification of Berge with the station Graero of the Peutinger Table, suggested by Fanoula Papazoglou, therefore modifying our understanding of the Roman road system of the lower Strymon Valley.
5Having determined, thanks to the letter of Galerius and Maximinus, the exact starting point of the two Roman roads perpendicular to the Via Egnatia and accepting as inevitable Fanoula Papazoglou’s suggestion that their destination was not Philippoi but Amphipolis, it is possible to retrace their course with the help of nineteenth century geographical works describing the Ottoman routes in this area20. Regarding especially the road on the western bank of the river, it appeared that, starting from Herakleia, it first crossed to the eastern bank of the river, in order to follow the only road suitable for vehicles along the Rupel pass, and returned to the western bank at the crossing of Euporia, the first station of this road, south of the pass. We owe to Fanoula Papazoglou the ingenious observation that between Euporia and Graero, the next station on the Peutinger Table, there had been omitted another station, which figures as Arason21 in the corresponding list of the Ravenna Geographer (IV 9). C. Müller, in his edition of the Geography of Ptolemy22, identified the Bisaltian city of Arrolos, listed by the ancient geographer (3.12.32), with this station on the Roman road. Arrolos is also mentioned by Hierocles23 and Constantine Porphyrogenetos24.
6The ethnic Arrolios is attested in a late fourth or early third century BC. inscription from Atrax:
Βΐλκης
Βορτᾶ
Ἀρρώλι-
ος Μακε-
δών25.
7Both the name and the patronymic of the deceased are extremely rare. The first appears only once in a first century AD funerary inscription from Thessalonike for a family consisting of Chaireas son of Hermogenes, his wife Meleïs daughter of Bilkes and their son Hermogenes26. This pre-Greek name may also figure as Βΐλσης on a late first century AD ephebic catalogue from Kalindoia (Βΐλσης Βείθυος) 27and as Βΐλθης on the mid-fourth century BC deed of sale from Strepsa (Basilika)28. In all these cases we are dealing with locations in the “New Territories” between the Thermaic Gulf and the Strymon valley, the population of which was in a significant proportion of pre-Greek origin.
8The patronymic Βορτᾶς directs us to the same ethno-linguistic environment. It is attested only once on a first century BC funerary inscription from Anthemous for Artilo daughter of Bortas. It may also figure as Ortas on two ephebic catalogues of Kalindoia of the second half of the first century AD29.
9The ethnic Arrolios, albeit in a slightly different form (unless the letter delta in the last word of the third line has not merely been accidentally scratched but deliberately erased) is also attested on a funerary stele of the end of the first or the beginning of the second century AD from Kalindoia bearing the following inscription:
Μουκάσης, Ήρακλείδης, Διο-
κλῆς, Σεύθης, (Ἱ)εροκ-
λῆς ο Βείθυος Ἀρδρώλ|ι|-
οι τῷ πατρί επί τῆς ξένης 30.
10The presence of the pre-Greek names Moukases31, Seuthes32, and B (e) ithys33, characteristic of the Strymon region, along with the Greek ones Herakleides and Hierokles, leaves little doubt that Ardrolios is the ethnic of the very same city of Bisaltia34usually spelt Άρ(ρ)ώλιος, for it is highly improbable that two cities with practically identical names existed in the same general area.
11A third attestation of the ethnic Ar(dr) olios occurs on a seal imprint (ΑΡΩΛΙΩΝ/ ΔΗΜOΝΙΚOΣ) from the ancient polis at Assari near Kalokastron35, which lies south of the southern exit of the Rupel pass, where we have located Euporia, and north of Trinlo/Tragilos and, according to Fanoula Papazoglou’s convincing restoration, also of Graero, that is to say on the expected route of the Roman road on the west bank of the Strymon. These converging clues confirm the Yugoslav scholar’s proposition that Arrolos was the name of the anonymous city on the ancient site of Assari36.
12Having arrived at this point, it is difficult to avoid the question whether the Ar(r)olioi/Ardrolioi of the Hellenistic and Roman inscriptions should not be identical with the Hedrolioi/Chedrolioi of the Athenian Tribute Lists. The hesitation between /a/ and /e/ in place names of the Thraco-Macedonian area is a well attested phenomenon37. The hesitation between Ηεδρόλιοι and Χεδρόλιοι in the fifth century Athenian Tribute Lists betrays the presence of a non-Greek name presenting phonemes foreign to the Greek ear. Is it too bold to posit that, the local pronunciation of this place name having perhaps evolved in the course of centuries, so that by the beginning of the Hellenistic period it had attained a stage of partial or total assimilation of its /d/ to the following /r/, its hesitant transcription in a writing system devised for a different language is reflected in the spellings– -ρδρ- or -pp- ?38
13One might still hesitate to connect the Bisaltian Ar(d) rolos with Hedrolos/Chedrolos, supposedly located in the Chalkidic peninsula, if in the list of 450/449 the ethnic Erodioi did not figure in the same place, just after the Stolioi, and with the same amount of ἀπαρχὴ as the Hedrolioi of the years 448/447 and 447/446, leaving little doubt that we are dealing with one and the same community39.
14The hypothesis of the identification of Hedrolos in the Athenian Tribute Lists with the city on the left bank of the lower Strymon valley can find support in the close parallelism of its fate during the pentakontaetia with that of Berge, its neighbour just across the river. Both join the Athenian alliance in close sequence, Berge in 452/451 and Hedrolos in 450/449. They both remain loyal to it until 447/446 and both disappear from the lists between that date and 435/434, when they reappear just a couple of years after the foundation of Amphipolis. Finally, they both defect after 432/431, when they are recorded for the last time40.
15The inescapable conclusion that the Athenians had been consistently active deep in the hinterland of τὰ ἐπί Θράικης and that Berge did not constitute an isolated case should not come as a surprise. The results of recent research in the lower Strymon area, exemplified in the excavations at the necropolis of Gazoros41, which brought to light grave goods from the fifth and fourth century B. C. one would expect to find only in Greek colonial foundations, and now confirmed by recent epigraphic discoveries at Neos Skopos, show that Greek penetration, even before the foundation of Amphipolis, had been early and profound and had made a lasting impact on the cultural profile of the area.
Notes de bas de page
2 Kahrstedt 1936, 419 n. 15, following a suggestion by A. B. West on the map facing p. 173 of CAH V1.
3 See below.
4 ATL, I, 561 (indirectly and very tentatively): Zahrnt 1971, 253-254, cf. 185: Hansen & Nielsen 2004, 826 no 566.
5 Hatzopoulos & Loukopoulou 1992-1996, 126-127.
6 BE, 2000, 453, no 2: Hatzopoulos 2000, and 2005.
7 Picard 2006, 269-283.
8 SEG, 14, 613. The stone had undoubtedly been transported to the church of Ampeloi (Mekes) from Assari.
9 SEG, 34, 613: BE, 203, 382.
10 Samsaris 1980, 1-7.
11 SEG, 24, 612.
12 The polis status of the settlement results from a whole series of inscriptions. See Papazoglou 1988, 379-381, and Hatzopoulos 1996, 1, 60-61, with references.
13 See Bonias 2000, 227-246; Matthaiou 2000-2003, 227-232; Koukouli-Chrysanthaki 2000, 227-232.
14 Hatzopoulos 1996, 2, 92 no 76
15 Note, though, the reservations of Papazoglou 1988, 384.
16 See Papazoglou 1988, 381-382, with references. See also the new epigraphic evidence (including a dedication dated from the end of the fourth or the beginning of the third century) published by P. Samsaris (BE, 1997, 415; 2000, 480), who maintains the identification of this site with Herakleia Sintike.
17 Papazoglou 1988, 359, 363, and 365: Hatzopoulos 1996, 1.214. The discussion was reopened after some new epigraphic discoveries: SEG, 31, 639; BE, 2001, 301.
18 See Hammond 1972, 1, 197; Samsaris 1976, 116; Papazoglou 1988, 356-357, 363, 370, 473; Hatzopoulos 1996, 1.214-15; BE, 1987, 708; 1991, 401; 2003, 382.
19 Samsaris 1976, 118-119; cf. Samsaris 1989, 215.
20 See Hatzopoulos 2008.
21 Papazoglou 1988, 362-363.
22 Müller 1901, 513.
23 Hierocles 640, 7, in the corrupt form Ἂραυρος.
24 Const. Porph., De them., 2,38 (Pertusi), in the corrupt form ραλος.
25 Larissa Museum, Inv. no 62/26. I wish to thank Maria Theochari for her kind permission to use this document and Apostolos and Mary Kontogianni, who so generously first drew my attention to this inscription.
26 IG, X, 2, 1, 404.
27 SEG, 42, 582, 33, discussed in Hatzopoulos & Loukopoulou 1992-1996, 234.
28 SEG, 37, 583.
29 SEG, 42, 580, 22 and 582, 10, discussed in Hatzopoulos & Loukopoulou 1992-1996, 234-236.
30 Hatzopoulos & Loukopoulou 1992-1996, 86-87, no. K8.
31 It is attested at Gazoros, on the east bank of the lower Strymon valley, and six times in the middle Strymon valley, north of the Rupel pass (LGPN).
32 It is attested four times in the region between the Thermaic Gulf and the Strymon valley, twice at Thessalonike and twice at Kalindoia (LGPN).
33 It is attested five times at Kalindoia, twice at Lete, three times at Thessalonike, once at Gazoros, seven times at Serres and seven times in the middle Strymon valley, north of the Rupel pass (LGPN).
34 Cf. Hatzopoulos & Loukopoulou 1992-1996, 86-87. Following a regrettable misapprehension which made me attribute to Fanoula Papazoglou the opinion that Assari was situated in Sintike, I unduly challenged the localisation of Assari in Bisaltia, proposed by Kaphtantzis.
35 Kaphtantzis 1967, 312, no 520.
36 Papazoglou 1988, 362-363.
37 Cf. Ἀρεθούσιος (Syll.3, 268 G) and Έρεθούσιος (Syll.3, 269 K), Κεμακαί in IG, I3, Καμακαί in SEG, 36, 626 etc. The same phenomenon can be observed in personal names (cf. Ἀρρίδαῖος - Έρρίδαιος, Ἀρραβαῖος - Έρρεβαῖος) and has been studied by Threatte 1980, 129.
38 My friend and colleague Claude Brixhe, in whose expert knowledge I took refuge, tentatively–and very kindly – suggested per litteras electronicas the following: “L’hypothèse la moins coûteuse consisterait à partir de dr et de supposer une spirantisation du !d! :
<d spirantisé (symbolisé par le delta) + r> -rr- (évolution banale; cf. latin quadratu > français “carré”).
< rr, géminé ou long, se maintient ou se réduit à –r-.
< Si maintien de la séquence comme géminée (deux phonèmes succésifs), on peut imaginer une surarticulation avec épenthèse consonantique, d’où –rdr- (cf. latin sur(ge)re > français “sourdre”). Dans ces évolutions, il ne faut pas oublier les affinités entre d et un r qui est apical”.
39 IG, I3, 263, IV, 18. This spelling variation can be explained either as a common error, if we suppose that the stone cutter was copying from a manuscript in the Ionian alphabet, in which lambda and delta have closely similar forms, or as an authentic variant, since, as Claude Brixhe kindly reminds me, the phonemes d and apical l have a strong affinity.
40 For the references concerning Berge, see Hansen & Nielsen 2004, 858-59. The references regarding the Hedrolioi have been cited at the beginning of this paper.
41 Poulios 1994, 598-99, 1995 631-33, 1998, 411-22.
Auteur
Research Centre for Greek and Roman Antiquity (KERA), National Hellenic Research Foundation, Athens, Greece
Le texte seul est utilisable sous licence Licence OpenEdition Books. Les autres éléments (illustrations, fichiers annexes importés) sont « Tous droits réservés », sauf mention contraire.
La transmission de l’idéologie impériale dans l’Occident romain
Milagros Navarro Caballero et Jean-Michel Roddaz (dir.)
2006
La circulation de l’information dans les états antiques
Laurent Capdetrey et Jocelyne Nelis-Clément (dir.)
2006
Villes et territoires dans le bassin du Douro à l’Époque romaine
Milagros Navarro Caballero, Juan José Palao Vicente et Maria Ángeles Magallón Botaya (dir.)
2007
Le monde de l’itinérance
En Méditerranée de l’Antiquité à l’époque moderne
Claudia Moatti, Wolfgang Kaiser et Christophe Pébarthe (dir.)
2009
Les seigneuries dans l’espace Plantagenêt
(c. 1150-c. 1250)
Martin Aurell et Frédéric Boutoulle (dir.)
2009
Studies in Greek epigraphy and history in honor of Stefen V. Tracy
Gary Reger, Francis X. Ryan et Timothy Francis Winters (dir.)
2010
Ombres de Thucydide
La réception de l’historien depuis l’Antiquité jusqu’au début du xxe siècle
Valérie Fromentin, Sophie Gotteland et Pascal Payen (dir.)
2010