Romes Dacian Wars: Domitian, Trajan, and Strategy on the Danube,
Part I*I
Feature
Everett L. Wheeler
Abstract Two recent major monographs, one on the Dacian wars of
Domitian and Trajan (Stefan) and another on ancient migrations from
the Ukraine into the eastern Balkans (Batty, Rome and the Nomads)
invite discussion and evaluation. A survey of the problematic
literary and archaeological sources (not least Trajan's Column) for
the history of this area in the first and second centuries A.D.
prefaces an evaluation of new archaeological evidence on Dacian
defenses and innovative topographical identifications. The
development of a Geto-Dacian state in Transylvania within the
context of multiple ethnicities on the Lower and Middle Danube is
discussed and use of new archaeological discoveries to clarify
narratives of the wars of 8489, 101102, and 105106 is evaluated.
Interpretations of scenes on Trajan's Column and the metopes of the
Adamklissi monument remain controversial.
he Roman conquest of Dacia (the ancient forerunner of modern
Romania) has not ceased to fascinate, as demonstrates the weighty
tome here discussedon my bathroom scale well over six pounds of
arguments, photographs, maps, and*Part II of Everett Wheelers Romes
Dacian Wars: Domitian, Trajan, and Strategy on the Danube will
appear in the Journal of Military History 75, no. 1 ( January
2011).
T
Everett L. Wheeler, scholar in residence at Duke University
received his A.B. from Indiana University/Bloomington and a Ph.D.
from Duke. He specializes in the history of military theory,
ancient history, and Armenian-Caucasian studies. His extensive
publications in ancient military history include Stratagem and the
Vocabulary of Military Trickery (1988), translation (with Peter
Krentz) of Polyaenus, Stratagems of War, 2 vols. (1994), and (ed.)
The Armies of Classical Greece (2007). Besides regular
participation in the International Congress of Roman Frontier
Studies and the Lyon Congress on the Roman Army, he serves on the
editorial board of Revue des tudes Militaires Anciennes.The Journal
of Military History 74 (October 2010): 11851227. Copyright 2010 by
The Society for Military History, all rights reserved. No part of
this publication may be reproduced, stored, or transmitted in any
form or by any means without the prior permission in writing from
the Editor, Journal of Military History, George C. Marshall
Library, Virginia Military Institute, P.O. Drawer 1600, Lexington,
VA 24450. Authorization to photocopy items for internal and
personal use is granted by the copyright holder for libraries and
other users registered with the Copyright Clearance Center (CCC),
121 Rosewood Drive, Danvers, MA 01923 USA (www.copyright.com),
provided the appropriate fee is paid to the CCC.
1185
EVERETT L. WHEELER
bibliography: Alexandre Simon Stefan, Les guerres daciques de
Domitien et de Trajan: Architecture militaire, topographie, images
et histoire, Collection de lcole Franaise de Rome 353 (Rome: cole
Franaise de Rome, 2005). A trio of general theses undergirds this
mountain of archaeological and topographical detail: first, the
people generally called Getae in Greek and Daci in Latin were not
barbarians like their German or Sarmatian neighbors, but a strongly
Hellenized state of some sophistication; second, the Emperor
Domitian (r. 8196 A.D.), the victim of hostile senatorial
historiography and the propaganda of Trajan (r. 98117), merits
rehabilitation: in reality, Trajan only imitated and continued
Domitians work in both the military and artistic spheres; third,
Stefans massive assemblage and reevaluation of archaeological data
on the Dacian wars, combined with innovative use of aerial
photography, permits new topographical interpretations of Roman
campaigns and a reassertion of the historical accuracy of scenes on
Trajans Column. Proper appreciation of the authors contentions,
however, merits a prolegomenon on the archaeological and
historiographical difficulties of treating Romes Dacian wars,
particularly as Stefans work spans the history of the Geto-Dacians
from the late sixth century B.C. to the Roman annexation in 106
A.D., and a subsequent recent work has much to say on the context
of these conflicts.1
1. R. Batty, Rome and the Nomads: The Pontic-Danubian Realm in
Antiquity (Oxford: Oxford University Press, 2007), attempts to
trace the history of migrations from the Ukraine into Romania and
Bulgaria (fifth century B.C.fourth century A.D.). Although
supplementing Stefans tome, Battys disappointing work suffers inter
alia, as this papers commentary will document, from factual errors
and out-of-date or omitted bibliography (e.g., ignorance of A.
Suceveanu and A. Barnea, La Dobroudja Romaine [Bucharest: Editura
Enciclopedica, 1991]; and A. Alemany, Sources on the Alans: A
Critical Compilation [Leiden: E. J. Brill, 2000]). His curious
pronouncements about Roman policy on the Lower Danube derive
exclusively from a very limited (cherrypicked?) knowledge of the
literature on Roman strategy (discussed in Part II of this
article). For other (and less tendentious) recent surveys of Roman
archaeology on the Danube, although not comprehensive and somewhat
disappointing from a military historians perspective, see J.
Wilkes, Recent Work along the Middle and Lower Danube, Journal of
Roman Archaeology 11 (1998): 23197; and Wilkes, The Roman Danube:
An Archaeological Survey, Journal of Roman Studies 95 (2005):
124225. Footnotes to this discussion, modestly updating Stefans
fifty-nine double-columned pages (70563) of bibliography through
2003, alert readers to important work available in North American
and Western European libraries without attempting to be
comprehensive. Stefan attests the prolific production of Romanian
scholars, including many works generally inaccessible outside
Romania, and often adds his own twist to other excavators ideas,
properly cited in his footnotes. His fuller documentation will not
be reproduced. Full bibliographical citations for all ancient
sources will not be given; English translations of most are
available in the Loeb Classical Library series. The following
abbreviations appear: AE=LAnne pigraphique (Paris: Presses
universitaires de France, 1888-); ILS=H. Dessau, Inscriptiones
Latinae Selectae, 3 vols. in 5 (Berlin: Weidmann, 18921916).
Apologies are owed to the Editor, whose patience in awaiting this
paper and toleration of its length are exemplary.
1186
THE JOURNAL OF
The Dacian Wars of Domitian and Trajan
The skills of siegecraft, engineering, and logistics required to
penetrate the sophisticated and extensive Dacian defenses of the
Carpathians probably exceeded those of the more famous sieges of
the Jewish War (6670 A.D.) and certainly involved much larger
forces on both sides. Although the Dacian conflicts of Domitian
(8489) and Trajan (101102, 105106) lack a Josephuss detailed
narrative, the spades of Romanian archaeologists, active for over a
century, have compensated for sparse literary sources by unearthing
much of the Dacian fortification system and providing clues to the
campaigns. Above all (quite literally), Roman victory required
capture of the Dacian capital, Sarmizegethusa Regia (modern
Gradishtea Muchelelui)no small feat for operations at an elevation
of nearly 1,000 meters in the heart of the southern Carpathians
Orashtie Mountains. The capital lies on a narrow ridge, which peaks
at Muncel (elevation 1,563.5 meters) and whose sheer slopes plunge
into the Alb and Godeannul Rivers on its northern and southern
sides respectively. Built on fourteen man-made terraces with
additional habitation extending along the ridge for 2 kilometers to
the west and about 1 kilometer to the north, Sarmizegethusas
massive fortifications, exploiting every topographical advantage,
enclosed an urban area of over three acres.2 But only half of the
capitals urban space has even been explored, much less dug.
Extensive forestation, greater today than in Antiquity, has impeded
understanding this site besides many others of these wars.
Moreover, the potholes of treasure-hunters as early as the
Napoleonic era, seeking fabled Dacian gold, and the discontinuity
of Romanian excavations, often with different working assumptions,
have complicated discerning the sites pristine state. Not least,
the Romans, true masters of wiping cities off the face of the
earth, as modern investigators of Hellenistic Corinth and Carthage
(both destroyed in 146 B.C.) can verify, left little behind. These
factors render Sarmizegethusa Regia an archaeological nightmare.
Dacian accomplishments and Romanias Roman heritage play a
significant role in Romanian national prideperhaps even more so
than with the popular notions of Roman Britain or Roman Germany,
spawning antiquarianism and costumed wargamers acting out their
fantasies. In 1980, when Romania hosted the International Congress
of Historical Sciences in Bucharest, the regime of Nicolae
Ceaucescu simultaneously celebrated 2,050 years of a Romanian
national state, taking 70 B.C. as a firm date (the real date in the
first or second quarter of the first century B.C. is hazy) for
Burebistas creation of a Dacian empire, extending from Ukrainian
Olbia on the Black Sea south to the Bulgarian Haemus Mountains
(modern Stara Planina), and as far west as modern Slovakia.3 More
recently (282. Stefans work dwarfs I. Olteans (Dacia: Landscape,
Colonisation and Romanisation [London: Routledge, 2007]) minimalist
view of Sarmizegethusa Regia: 87, 89. 3. On the Ceaucescu regimes
control of even dissertation topics in ancient history, see V.
Lica, The Coming of Rome into the Dacian World, trans. C. Patac and
M. Neagu, rev. A. R. Birley (Konstanz: UVK Universsittsverlag,
2000), 35 n.50; Romanian originswhether Geto-DacianMILITARY
HISTORY
I
1187
EVERETT L. WHEELER
September1 October 2006), an international conference at Cluj
commemorated Trajans creation of provincia Dacia (106 A.D.).4 The
current mania for commemorative academic conferences, however, is
not exclusively Romanian. The 2,000th anniversary of the reign of
Trajan (98117), the first Spanish emperor, prompted a conference in
Spain and, as Trajan was in Germany when Nerva (his predecessor)
died, a German conference of 1998 has been followed by a
semi-popular book of useful essays and nice pictures.5 Nor should a
recent biography of Trajan in English (now corrected and reprinted)
be ignored, although not a replacement for Paribenis substantial
two-volume study.6 Trajan and the Dacian wars are currently hot.
Apart from any nationalistic considerations (whether Romanian,
German, or Spanish), the Dacian wars present an interesting
methodological and historioor Romanhave been a political football
in Romania and a source of regional antagonism with Hungary and
Bulgaria, both of which still wince at Romanian possession of parts
of Transylvania and the Dobrudja (the area between the Danubes
northward bend and the Black Sea), respectively; for post-Ceaucescu
evaluations of these issues, see M. Babe, Devictis Dacis. La
conqute trajane vue par larchologie, in Civilisation grecque et
cultures antiques pripheriques. Hommage P. Alexandrescu, ed. A.
Avram and M. Babe, (Bucharest: Editura enciclopedica, 2000), 324
with n. 6, 325 n. 10; I. Haynes and W. Hanson, An Introduction to
Roman Dacia, in Roman Dacia: The Making of a Provincial Society,
ed. W. Hanson and I. Haynes, Journal of Roman Archaeology, Suppl.
56 (Portsmouth, R.I., 2004): 2729; K. Locklear, The Late Iron Age
Background to Roman Dacia, in Hanson and Haynes, eds., Roman Dacia,
3335; a convenient summary (by no means definitive) on Burebista
may be found in I. Cristan, Burebista and His Time, trans. S.
Mihailescu (Bucharest: Editura Academiei Republicii Socialiste
Romania, 1978), a work unknown in C. Bruuns bizarre paper, The
Legend of Decebalus, in Roman Rule and Civic Life: Local and
Regional Perspectives, ed. L. De Ligt et al. (Amsterdam: J. C.
Gieben, 2004), 15375. This reviewer, a participant in the 1980
congress, visited Sarmizegethusa Regia, at that time undergoing
conversion into a tourist attraction with dubious reconstructions
(in Stalinist concrete) of the monuments in the sacred area
(Terraces XXIII, featuring seven temples/ sanctuaries; cf. Stefan,
2269 with n. 235) and damage to the scientific understanding of the
site. Ascent to the site required four-wheel drive vehicles. 4. See
I. Piso, ed., Die rmischen Provinzen: Begriff und Grndung
(Cluj-Napoca: Editura Mega, 2008). 5. J. Gonzles, ed., Trajano
Emperador de Roma (Rome: LErma di Bretschneider, 2000); E.
Schallmeyer, ed., Traian in Germanien, Traian im Reich (Bad
Homburg: Saalburgmuseum, 1999); A. Nnnerich-Asmus, ed., Traian: Ein
Kaiser der Superlative am Beginn einer Umbruchzeit? (Mainz: Verlag
Philipp von Zabern, 2002); on Trajans somewhat peculiar position in
Germany in 98 (named Caesar, i.e., Nervas successor, in October 97
and thus subsequently possessing an imperium proconsulare, but on
present evidence not the provincial governor of either Germania
Superior, which he had been in 97, or Germania Inferior), see B.
Pferdehirt, Militrdiplome und Entlassungskunden in der Sammlung des
Rmisch-Germanischen Zentralmuseums (Mainz: Verlag des
Rmisch-Germanischen Kommission, 2004), 1:2627; cf. M. A. Speidel,
Bellicosissimus Princeps, in Nnnerich-Asmus, ed., Traian, 24. 6. J.
Bennett, Trajan Optimus Princeps, 2d ed. (Bloomington: Indiana
University Press, 2001); cf. my brief (and generous) review of the
first edition: Journal of Military History 62
1188
THE JOURNAL OF
The Dacian Wars of Domitian and Trajan
graphical problem for military historians concerned with
operations and strategy. Detailed information is at a premium.
Domitian, the victim of hostile sources and Trajans propaganda,
falls in the chasm of imperial biographies between Suetonius, whose
De vita Caesarum ends with Domitian, the last of the Flavians, and
that most curious assemblage of biographies and historical novellas
written in the late-fourth or early-fifth century, the Historia
Augusta, which begins with Trajans successor, Hadrian (r. 117138)a
remarkable phenomenon for a ruler hailed as the best emperor
(optimus princeps). Narrative surveys of Trajans reign survive
exclusively in the summaries of epitomators of the fourth century
and later. Tacituss Histories, covering the Flavian dynasty and
thus including Domitians campaigns (8489), survive only for events
up to 70 A.Dmost regrettably, as Orosius (7.10.4) reported that
Tacitus (a contemporary of Domitian and Trajan) recounted Domitians
Dacian war in great detail. Appians Dacica, written within a
generation or two of Trajans wars and known from brief references
by Photius in the ninth century and Zonaras in the twelfth century,
has no surviving fragments. Apparently few read it. Literary
accounts of events and motives are reduced to two sources: the
Roman History of the senator Cassius Dio, completed in the 220s and
for the period of Domitian and Trajan preserved in excerpts from
John Xiphilinuss eleventh-century epitome, supplemented by
scattered fragments in other Byzantine sources, and the Getica of
Jordanes (fl. 550), probably a Sarmatian Alan in Constantinople,
whose family had earlier assimilated with the Goths. Jordanes
claimed to be epitomizing the twelve-volume De origine actibusque
Getarum of the Ostrogoth bureaucrat and scholar Cassiodorus (c.
490c. 585), but he also cited a lost Gothic history by Ablabius of
unknown date. Jordanes Getica contains material from the
Stoic-Cynic orator and sophist Dio Chrysostom, whose time in Dacia
in the 90s (after being exiled from Rome by Domitian) inspired his
Getica, but whether Jordanes knew Chrysostoms Getica directly or
through another source is unknown. The archaizing tendency of Late
Roman authors like Jordanes in combining Dacians and Goths and
calling a work on Goths a Getica is clear.7 From the third century
on, the(1998): 38283; R. Paribeni, Optimus Princeps, 2 vols.
(Messina: G. Principato, 192627), with some echoes of the il Duce
of Paribenis time; note also: R. Hanslik, Marcus Ulpius Traianus,
Real-Encyclopdie der classischen Altertumswissenschaft, Suppl. 10
(1964): 10321113; M. Fell, Optimus Princeps? Anspruch und
Wirklichkeit der imperialen Programmatik Kaisers Traians (Munich:
Tuduv, 1992); and (more briefly) Speidel, Bellicosissimus Princeps,
2340. 7. Cassius Dios Books 6768 on Domitian and Trajan are
available in the Loeb Classical Library: Dios Roman History, trans.
E. Cary, vol. 8 (Cambridge, Mass.: Harvard University Press, 1925),
but the Greek text is best read in Cassii Dionis Cocceiani
Historiarum Romanarum Quae Supersunt, ed. U. P. Boissevain, 4 vols.
(Berlin: Weidmann, 18981931); Jordanes: T. Mommsen, ed., Iordanis
Romana et Getica, Monumenta Germanicae Historica, Auctores
Antiquissimi, 5 (Berlin: Weidmann, 1882); The Gothic History of
Jordanes, trans. C. C. Mierow (Princeton, N.J.: Princeton
University Press, 1915); Jordanes Alan ancestry: Alemany, Sources
on the Alans, 13637; fragments of Dio Chrysostoms Getica: F.
Jacoby, Die Fragmente der griechischen Historiker (Leiden:MILITARY
HISTORY
1189
EVERETT L. WHEELER
German Goths occupied some territory earlier belonging to the
Thracian GetoDacians (especially after Aurelians abandonment of
Dacia, c. 270). Some Dacian descendants, both native survivors of
the Roman conquest and the so-called Free Dacians, inhabiting
territory not annexed as part of Trajans province, may have
assimilated with Gothic intruders, although a Geto-Dacian culture,
distinct from the Sntana de Muresh-Cernjachov culture associated
with the fourth century Goths, continued into the Middle Ages.8
Detailed contemporary accounts did exist. In the tradition of
Caesars Gallic War, Trajan published his own commentaries on his
campaigns, a Dacica, fromE. J. Brill, 1923 ), nr. 707; Cassiodorus:
B. Croke, Cassiodorus and the Getica of Jordanes, Classical
Philology 82 (1987): 11734; for doubts about Ablabius as a source,
see A. Gillett, Jordanes and Ablabius, in Studies in Latin
Literature and Roman History, ed. C. Deroux, vol. 10, Collection
Latomus 254 (Brussels, 2000), 479500; Locklears skepticism (Late
Iron Age Background, 34), typical of new archaeologists, on the
value of literary sources like Jordanes, lacks authority, as his
view seems derived from English translations and not the original
Latin, nor does he understand the archaizing tendencies of Late
Roman authors. 8. On the Free Dacians, see G. Bichir, Die freien
Daker im Norden Dakien, and I. Ionita, Die freien Daker an der
nordstlichen Grenze der rmischen Provinz Dakiens, in Rmer und
Barbaren an den Grenzen des rmischen Dakiens, ed. N. Gudea, Acta
Musei Porolissensis 21 (1997): 785800 and 879888, respectively;
Batty, Rome and the Nomads, 36568; Lica, Coming of Rome, 256, 264,
although his citation of ILS nr. 854 seems more relevant to the
Costoboci than the Free Dacians; cf. Olteans misunderstanding of
Lica: (Dacia, 56). Batty (Rome and the Nomads, 485) erroneously
believes that Trajans wars depopulated Dacia; for correctives, see
Babe , Devictis Dacis, and literature at note 38 below. Ethnic
confusion can befuddle even modern authors: Stefan (359 n.2)
corrects a gaffe (the more egregious for a book on an ancient
geographer) that the Getae were Germans: D. Dueck, Strabo of
Amaseia: A Greek Man of Letters in Augustan Rome (London:
Routledge, 2000), 97; Oltean (Dacia, 47) erroneously equates the
German Bastarnae of southern Moldavia with Iranian Sarmatians;
evidence on the Bastarnae collected in Batty, Rome and the Nomads,
22124, 23656; see also M. B. Shchukin, Forgotten Bastarnae, in
International Connections of the Barbarians in the Carpathian Basin
in the 1st5th Centuries A.D., ed. E. Istvnovits and V. Kulscr
(Aszd/Nyregyhza: Jsa Andrs Museum; Osvth Gedeon Museum Foundation,
2001), 5764. On the Sntana de Muresh-Cernjachov culture, see the
convenient but largely inconclusive summary (as of 1991) in P.
Heather and J. Matthews, The Goths in the Fourth Century
(Liverpool: Liverpool University Press, 1991), 51101; B. V.
Magomedev, Die Cernjachov-Marosszentanna/Sntana de Mures-Kultur in
der Karpatenregion, in Istvnovits and Kulscr, eds., International
Connections, 22733; L. Ellis, Dacians, Sarmatians, and Goths on the
Roman-Carpathian Frontier: Second-Fourth Centuries, in Shifting
Frontiers in Late Antiquity, ed. R. Mathisen and H. Sivan
(Brookfield, Vt.: Variorum, 1996), 10525; note also the recent
archaeological survey of F. Curta, Southeastern Europe in the
Middle Ages, 5001200 (Cambridge: Cambridge University Press, 2006).
Batty (Rome and the Nomads, 250), who strangely omits discussion of
the Sntana de Muresh-Cernjachov culture, is skeptical of Romanian
scholars identification of various ethnicities (Costoboci, Carpi,
Bastarnae) with specific material cultures, although his own views
lack appreciation of archaic ethnic terms in late authors for
various tribes of their own day, and he uncritically accepts
material in (e.g.) Plinys Natural History, where earlier sources
are indiscriminately mixed with contemporary ethnographical
descriptions.
1190
THE JOURNAL OF
The Dacian Wars of Domitian and Trajan
which a single sentence survives in the grammatical work of
Priscian (fl. 500), although interpreters of Trajans Column
(including Stefan) see that monument as a massive illustration of
the Dacicas contents. A few Byzantine fragments of the Getica of
Trajans physician, Titus Statilius Crito, a participant in the
campaigns, provide valuable but limited details. Even Balbus, a
civilian surveyor called into service with Trajan, offers
intriguing hints of his duties in building roads, bridges, and
siege-works but regrettably without specific geographical
locations.9 Indeed Domitian, known as a good poet, commemorated his
Dacian war with an epic poem, of which a few lines, inscribed in
monumental letters on a block found in the Lateran area of Rome,
were first recorded by the humanist Petrarch.10 Archaeology,
epigraphy, numismatics, and papyrology strongly supplement the
sketchy literary sources. Many Dacian forts and Roman camps are
known, for which dates of construction or destruction can be
discerned or approximated. Dates for the beginning and end of
hostilities besides terms of peace are clear, as are which units of
the Roman army and their commanders participated in these
campaigns.119. Trajans Dacica: Priscian, Institutiones grammaticae
6.13=E. M. Smallwood, Documents Illustrating the Principates of
Nerva, Trajan and Hadrian (Cambridge: Cambridge University Press,
1966), no. 32: Traianus in I Dacicorum: Inde Berzobim, inde Aizi
processimus [From Berzobis, then from Aizis we advanced.]; Crito:
Jacoby, Die Fragmente, nr. 200; cf. J. Scarborough, Criton,
Physician to Trajan: Historian and Pharmacist, in The Craft of the
Ancient Historian: Essays in Honor of Chester G. Starr, ed. J.
Eadie and J. Ober (Lanham, Md.: University Press of America, 1985),
387405; for an attempt to find more fragments, see I. I. Russu,
Getica lui Statilius Crito, Studii Clasice 14 (1972): 11128 (cited
in Stefans footnotes, but missing from his bibliography); Balbus:
B. Campbell, The Writings of the Roman Land Surveyors, Journal of
Roman Studies Monograph 9 (London, 2000), xxxixxl, 2056 (Latin text
with English translation), esp. lines 1730. Despite Campbells
waffling between whether the emperor is Domitian or Trajan, Balbuss
references to mountain warfare in Dacia speak for Trajan, as the
Dacian campaigns under Domitian (none led by Domitian himself )
reached, but did not penetrate the Dacians Carpathian
fortifications; cf. Stefan 419 with n. 118. Like Crito and Balbus
in Trajans entourage, the famous architect Apollodorus of Damascus
and possibly Dio Chrysostom (cf. Bennett, Trajan Optimus Princeps,
67) were not soldiers but civilian comites (companions). For
attempts to connect Apollodorus of Damascuss Poliorcetica
[Siegecraft] to Trajans Dacian wars, see P. Blyth, Apollodorus of
Damascus and the Poliorcetica, Greek, Roman and Byzantine Studies
33 (1992): 12758; D. Whitehead, Apollodorus Poliorketika: Author,
Date, Dedicatee, in A Roman Miscellany: Essays in Honour of Anthony
R. Birley on his Seventieth Birthday, ed. H. Schellenberg et al.
(Gdansk: Foundation for the Development of Gdansk University,
2008), 20411. 10. Corpus Inscriptionum Latinarum, VI 1207; cf.
Quintilian, Institutio oratoria 10.1.91; Silius Italicus, Punica
3.61621; Suetonius, Domitianus 2.2; further commentary in Stefan
472. 11. See K. Strobel, Untersuchungen zu den Dakerkriegen Trajans
(Bonn: Rudolf Habelt, 1984), now updated at K. Strobel, Die
Eroberung DakiensEin Resmee zum Forschungsstand der Dakerkriege
Domitians und Traians, Dacia 50 (2006): 10514, and K. Strobel, Die
Donaukriege Domitians (Bonn: Rudolf Habelt, 1989), although many of
Strobels assertions require qualification: see F. Leppers review of
Strobels method: Classical Review 35 (1985): 33335; see also N.
Gostar, Larme romaine dans les guerres daces de Trajan (101102,
105106), Dacia 23 (1979): 15522; G. Cupcea and F. Marcu, The Size
and Organization of the Roman Army and the Case of Dacia under
Trajan, Dacia 50 (2006): 17594; basic remains K. Patsch, Der
KampfMILITARY HISTORY
1191
EVERETT L. WHEELER
Trajan's Column, scene LXV, courtesy Deutsches Archologisches
Institut, Rome, Italy
But the visible yet impenetrably silent glue holding together
the framework of Trajans Dacian wars is that towering shaft (28.9
meters on a pedestal of 6.2 meters) in the middle of Rome, with its
cartoon of over 2,500 figures twisting around it for 200 meters.
Trajans Column, dedicated in 113, tells the story of his Dacian
wars, which his victory arch at Beneventum (dated 113114) completes
by depicting his Dacian triumph. On the Column, Trajan, Decebalus
(the Dacian king), and various Roman units or hostile ethnic forces
(for example, Lusius Quietuss Moorish cavalry, Sarmatian
cataphracts of the Rhoxolani) are readily identifiable; there is a
splendid display of Roman military practices.12 Some scenes
correspond to the fragments of Cassius Dios account. Yes, the
Column tells a story, but theum den Donauraum unter Domitian und
Trajan, 5/2, Beitrge zur Vlkerkunde von Sdosteuropa, 5.2 (Vienna:
Verlag der sterreichischen Akademie der Wissenschaften, 1937); a
brief survey of Danubian legions and their bases is at J. Wilkes,
Roman Legions and their Fortresses in the Danube Lands (First to
Third Centuries), in Roman Fortresses and their Legions, ed. R.
Brewer (London/Cardiff: Society of Antiquaries of London; National
Museums & Galleries of Wales, 2000), 10119. 12. See I. A.
Richmond, Trajans Army on Trajans Column, Papers of the British
School at Rome 13 (1935): 140, reprinted in Trajans Army on Trajans
Column, ed. M. Hassall (London: British School at Rome, 1982),
which also includes Richmonds study of the Adamklissi monument
(Tropaeum Traiani): Adamklissi, Papers of the British School at
Rome 35 (1967): 2939.
1192
THE JOURNAL OF
The Dacian Wars of Domitian and Trajan
crux for operational analysis comes with topographical
identifications, pinpointing a scene with a specific site on the
ground, particularly as the theater of the Dacian wars included not
only modern Romania, but also parts of Hungary, Serbia, Bulgaria,
Moldavia, and possibly extreme southwestern Ukraine, and several
Roman armies operated simultaneously.13 Hence frustration and
scholarly debate flourish. Comparison with the problems presented
by the Bayeux Tapestry for William the Conquerors campaign of 1066
is apropos. Nor is study of the Columns reliefs uncomplicated.
Modern environmental hazards, damaging this monument like many
others, have obliterated or blurred details and necessitated a
cleaning and attempts at restoration (198188). All painted details
and likewise metal supplements (for example, spears in the figures
hands) have long vanishedthus the value of records of the Column in
earlier, better states of preservation. Happily, Napoleon IIIs
well-known interest in Julius Caesar led him to Rome. In 186162 he
had molds made of the Columns entire historical scroll, from which
three complete sets of plaster casts were later produced. These now
reside in Paris (Muse des Antiquits Nationales
Saint-Germain-en-Laye), Rome (Museo della Civilt Romana), and
London (Victoria and Albert Museum).14 Historical interpretation of
the Column began at the dawn of the twentieth century with Conrad
Cichorius, a student of the revered Theodor Mommsen, who, working
from the 414 casts at Rome made from Napoleon IIIs molds, produced
a multi-volume photographic archive of the casts with commentary.
Cichoriuss division of the casts into 155 scenes (traditionally
given in Roman numerals) remains the standard method of citing the
Column, and his photographs are considered the best ever produced.
Nevertheless, Cichoriuss belief in the Column as a valid historical
account of Trajans wars and his identifications of Romanian sites
in the Columns scenes (at a time when excavations of Dacian sites
were still in their infancy) soon elicited harsh reviews and
alternative topographical views, such as those of G. A. T. Davies
(1920), who persisted, however, in seeing the Column as an
illustration of Trajans Dacica. Six years later the axe fell: the
art historian Kurt Lehmann-Hartlebens assessment of the Column as a
work of art demolished Cichoriuss case for the Columns precise
historical narrative and his topographical identifications.1513.
Stefans map (674 fig. 276) does not include all Dacian sites
relevant to the 106 campaign; for a supplement, see A. Diaconescu,
Dacia and the Dacian Wars, Journal of Roman Archaeology 21 (2008):
590. 14. A fourth set, the work of the now defunct cole Roumaine de
Rome in 193943, has been since 1967 on display at the National
Museum of History in Bucharest, where the bands of the spiral
(roughly 11.5 meters high) can be profitably studied at eye-level,
as this writer can attest from visits in 1980 and 1996. 15. C.
Cichorius, Die Reliefs der Traianssule, vols. 23 (Berlin: G.
Reimer, 18961900; vol. 1 not published); G. A. T. Davies,
Topography and the Dacian Wars, Journal of Roman Studies 10 (1920):
128; cf. his Trajans First Dacian War, Journal of Roman Studies 7
(1917): 7497; K. Lehmann-Hartleben, Die Trajanssule. Ein rmisches
Kunstwerk zu Beginn der Sptantike, 2 vols. (Berlin/Leipzig: W. de
Gruyter & Co., 1926); a more detailed account of the Columns
history is in R. Lepper and S. Frere, Trajans Column (Gloucester,
U.K.; Wolfboro, N.H.: Alan Sutton Publishing, 1988), 14; cf. Stefan
34.MILITARY HISTORY
1193
EVERETT L. WHEELER
Henceforth advocates of the Column as an historical source have
been on the defensive. Even the Roman army seen on the Column has
become a victim of the artists (or artists) supposed inaccuracies
and generalizations, although with little appreciation of what was
possible in limited space. Accordingly, individual legions cannot
be identified; the equipment of both legionaries and auxiliaries is
misrepresented; even the legionary and Praetorian signa (standards)
are wrong.16 Perhaps the apogee of the anti-Column movement came in
1988, when Frank Lepper and Sheppard Frere republished Cichoriuss
plates (long out of print) and subjected all aspects of Trajans
Dacian wars to keen critical analysis. If their reproduction of
Cichoriuss folio plates in an octavo volume was a major
disappointment (miniscule and often unclear)an Italian volumes
reproduction of the plates the same year is far superiortheir tome
represented a useful status quaestionis.17 Inter alia, a frequent
target of their criticism became the idea that Apollodorus of
Damascus, the architect behind construction of Trajans famous stone
bridge over the Danube at Drobeta (modern Turnu Severin) and the
supposed designer of Trajans Forum, was the genius (the Maestro)
behind the Columns reliefs, although they concede that he might
have been the architect of the Column. Indeed, as generally agreed
whoever the Maestro wasthe Columns scenes (in whole or part) derive
from paintings of the wars events displayed in Trajans triumph.18
For the history of the16. J. C. N. Coulston has led the charge
against the Column: The Value of Trajans Column as a Source for
Military Equipment, in Roman Military Equipment: The Sources of
Evidence, ed. C. van Driel-Murray (Oxford: British Archaeological
Reports, 1989), 3144; The Architecture and Construction Scenes on
Trajans Column, in Architecture and Architectural Sculpture in the
Roman Empire, ed. M. Henig (Oxford: Oxford University Committee for
Archaeology, 1990), 3950; Three New Books on Trajans Column,
Journal of Roman Archaeology 3 (1990): 290309; cf. M. C. Bishop and
J. C. N. Coulston, Roman Military Equipment (London: B. T.
Batsford, 1993), 2123; M. Charles, The Flavio-Trajanic Miles: The
Appearance of Citizen Infantry on Trajans Column, Latomus 62
(2002): 66695; C. G. Alexandrescu, A Contribution on the Standards
of the Roman Army, in Limes XIX: Proceedings of the XIXth
International Congress of Roman Frontier Studies Held in Pcs,
Hungary, September 2003, ed. Z. Visy (Pcs: University of Pcs,
2005), 14756; for a more sympathetic view of the accuracy of the
Column and what was really possible, see D. Richter, Das rmische
Heer auf der Trajanssale. Propaganda und Realitt: Waffen und
Ausrstung, Marsch, Arbeit und Kampf (Mannheim: Bibliopolis, 2004),
and M. Galinier, La representation iconographicque du lgionnaire
romain, in Les lgions de Rome sous le Haut-Empire, ed. Y. Le Bohec
and C. Wolff, 3 vols. (Paris: De Boccard, 20002003), 2:41739; a
credulists position on the Columns representation of the army is in
L. Rossi, Trajans Column and the Dacian Wars, trans. J. M. C.
Toynbee (Ithaca, N.Y.: Cornell University Press, 1971). 17. Lepper
and Frere, Trajans Column; S. Settis et al., La Colonna Traiana
(Turin: G. Einaudi, 1988); note also the reproduction of the plates
in F. Coarelli, The Column of Traian, trans. C. Rockwell (Rome:
Colombo, 2000). 18. See J. N. C. Coulston, Overcoming the
Barbarian. Depiction of Romes Enemies in Trajanic Monumental Art,
in The Representations and Perception of Roman Imperial Power, ed.
L. De Blois et al. (Amsterdam: J. C. Gieben, 2003), 42021,
elaborating on an idea of LehmannHartleben, Die Trajanssule);
Apollodorus: Lepper and Frere, Trajans Column, e.g., 1819, 149 50,
271; the bridge, depicted on the Columns scenes XCVIIIXCIX (=Stefan
fig. 267), was
1194
THE JOURNAL OF
The Dacian Wars of Domitian and Trajan
wars military operations, however, Lepper and Frere marched on
well-trodden paths. Despite nearly a century of faultfinding, it
cannot be denied that the Columns spiral tells a story. But what
story? The book discussed here seeks in part to revalidate some
aspects of the Columns historical worth. Stefans massive monograph,
the fruit of over thirty years of archaeological and topographical
studies of ancient Romania, appears in the distinguished series of
the cole Franaise de Rome, often reserved for the ultimate in
French dissertations, the doctorat dtat. The work bears the
characteristics of such: thickness (704 pages of text in double
columns!) and exhaustive discussions (in the sense of both
depleting what can be said and trying a readers patience). No one
except a reviewer (or perhaps a graduate student) would ever read
this book cover to cover. Nevertheless, its bibliography alone
renders the volume an indispensable reference tool, and no serious
future work on the Dacian wars of Domitian or Trajan will be able
to ignore Stefans gold mine of information. The text is excellently
complemented by 286 illustrations (photographs, maps, sketches)all
in the superb quality traditional in this series. Indeed the volume
contains some of the best photographs of scenes from Trajans Column
available. Regrettably, in the French tradition of academic
publications this bulky monument of scholarship (over 800 pages) is
in paperback!19 This is not a typical monograph, but actually at
least three books in one. Stefans first book offers a detailed
archaeological analysis of all known Dacian fortifications in the
Carpathians, the Dacian forts (discovered so far) in southern
Moldavia (guarding passes into the Dacian heartland from the east),
and Dacian forts on the north bank of the Danube (a Dacian limes of
sorts) in the area west of the Iron Gates Gorge, where a branch of
the Carpathians extends into northern Serbia and divides the Middle
from the Lower Danube. In addition, he examines most Roman camps in
Romania associated with the wars of Domitian and Trajan, with
particular attention to the controversial Roman camp at
Sarmizegethusa Regia in 102 and its successor in 106. If a complete
archaeological record of the Dacian wars lies far in the future
even for the area of Sarmizegethusa Regia, Stefan presents the most
extensive record of Romanian archaeology on these wars to date. The
highly techcompleted in 105 for the start of Trajans second war;
for the bridges remains and bibliography, see Stefan 64142, who
corrects (641 n. 34) the archaeological misconceptions of S. P.
Mattern (Rome and the Enemy: Imperial Strategy in the Principate
[Berkeley/Los Angeles: University of California Press, 1999], 149
n. 113), although he does not respond to Lepper and Freres claim
(Trajans Column, 14950) that the Column misrepresents the
superstructure of the bridge; see, most recently, M. Serban,
Trajans Bridge over the Danube, International Journal of Nautical
Archaeology 38 (2009): 33142. 19. This non-native reader of French
detected only occasional misprints in the texttoo infrequent to
catalogue or distractand two over-inked pages (485, 488) are the
only production errors. Embarrassing must be Stefans confusion (754
in the Bibliography and passim in footnotes, e.g., 532 n. 190) of
Michael P. Speidel with his nephew, Michael A. Speidel, the real
author of Bellicosissimus Princeps, in Nnnerich-Asmus, ed., Traian,
2340.MILITARY HISTORY
II
1195
EVERETT L. WHEELER
Map 1: Dacia KEY Passes: A) Iron Gates; B) Vulcan; C) Red Tower
Geto-Dacian Sites: 1) Sarmizegethusa Regia; 2) Buridava; 3) Deva;
4) Apulum; 5) Piatra Craivii; 6) Porolissum; 7) Piroboridava; 8)
Troesmis; 9) Noviodunum; 10) Aegyssus; 11) Popeshti; 12)
Sboryanovo; 13) Borovo; 14) Divici; 15) Berzobis; 16) Aizis Greek
Cities: 17) Tyras; 18) Histria; 19) Tomis; 20) Callatis; 21)
Dionysopolis; 22) Odessus; 23) Mesembria; 24) Apollonia; 25)
Axiopolis Roman Sites: 26) Aquincum (Budapest); 27) Lugo; 28)
Szeged; 29) Singidunum (Belgrade); 30) Viminacium; 31) Lederata;
32) Oescus; 33) Novae; 34) Nicopolis ad Istrum; 35) Serdica
(Sofia); 36) Durostorum; 37) Adamklissi; 38) Colonia
Sarmizegethusa; 39) Drobeta
nical discussion of construction techniques for walls, towers,
and buildings may be impenetrable for the uninitiated in this type
of archaeological argument, although one can admire Stefans
diligence in re-examining the entire record of excavation reports
and earlier sketches of the sites (often ignored by later
excavators) to determine the pristine states of sites and the
excavators original findings. Stefan has also pioneered use of
aerial and satellite photography in Romanian archaeology, not only
to find sites but also to discern the lines of fortification walls
not visible at ground-level. This highly technical archaeological
discussion (Parts III equal 339 pages), essentially half the book,
is not well integrated, however, with Parts IVV (pp. 397704),
detailed discussion of the wars of Domitian and Trajan, where
the1196THE JOURNAL OF
The Dacian Wars of Domitian and Trajan
reader is expected to recall too much from what he/she read,
say, 400 pages earlier. But then again, this is not a typical
monograph. Stefans compendium of archaeological data provides the
building blocks for various theses. In the case of Sarmizegethusa
Regia, for example, it has been much debated whether Gradishtea
Muchelelui was in fact the Dacian capital in Trajans time. Some
once preferred the site of Varhly (43 kilometers west-southwest of
Gradishtea Muchelelui, as the crow flies), where a Colonia Ulpia
Traiana Sarmizegethusa was founded c. 107 on the site of a Roman
camp 101105.20 The Colonia lies in a plain just east of the famous
Iron Gate Pass, but the site lacks evidence of pre-Roman
occupation. Gradishtea Mucheleluis role as a major Dacian religious
centermost probably on the holy mountain Cogaeonum, mentioned in
one of Strabos reports (7.3.5) on Burebistaseems clear: the sacred
area of Terraces XXIII features seven major sanctuaries and
comprises 15 to 20 percent of the three-acre site, the only portion
so far subjected to extensive excavation. Gradishtea Mucheleluis
sanctuaries exceed in size and number the four at Costesti, a major
fortress guarding the northwestern approach to Gradishtea
Muchelelui, and the three at Fetele Alba, on a ridge of equal
elevation opposite Gradishtea Muchelelui. Such a concentration of
sanctuaries surely indicates the sacred character of the area.21
Further, Stefan now argues that Gradishtea Muchelelui was not an
open site or mere citadel of refuge, but the Dacian capital and a
major fortified city with curtain walls and, as he conjectures,
towers at every 25 to 30 meters. More complete study of the
archaeological evidence also permits a better understanding of the
citys water system attested on the Column as well as other
architectural20. On the problem of a precise date for foundation of
the Colonia, see I. Piso, Les dbuts de la province de Dace, in
Piso, ed., Die rmischen Provinzen, 31922; on the camp see note 27
below; the confusion of Colonia with Regia is perpetuated in Batty,
Rome and the Nomads, 529, whose fig. 41.8 (290) also incorrectly
transposes the locations of the forts Blidaru and Costesti. 21.
Strobel (Die Eroberung Dakiens, 111 n. 29) rejects without argument
the identification of Cogaeonum with Gradishtea Muchelelui, but
ignores that Strabos Cogaeonum is a mountain, not a specific site;
for lists and discussions of Dacian sanctuaries, see Babe, Devictis
Dacis, 33031; Locklear, Late Iron Age Background,5763.MILITARY
HISTORY
1197
EVERETT L. WHEELER
features.22 Here Stefan initiates his motif of confirming the
accuracy of individual details on the Column concerning siegecraft
and Dacian architecture with archaeological evidence. In fact, he
asserts (against current opinions) that all major Dacian forts,
generally atop hills or mountains with difficult lines of approach,
were totally enclosed fortifications, that is, even on sides
bordering cliffs or inaccessible slopes. Moreover, he demonstrates
that the sophistication of Dacian building techniques conformed to
Hellenistic Greek practices. Dacian walls, often 3 meters thick in
the murus Dacicus technique (front and back stone abutments with
earth/rubble filling and both transverse and lateral timbers as
stabilizers and linkage), employed large, wellcut, nicely faced
stone blocks (some marked with Greek letters). Dacian military
architecturefar from being barbarianreflected (with some local
modifications) the recommendations of Philo Mechanicus (fl. 225
B.C.) for construction of walls immune to battering rams, and
likewise Philos view on the architectural integrity of towers (that
is, structures independent of the walls, so that a collapsing wall
did not bring a tower down with it).23 Dacians built huge towers
(surface area up to 221.12 square meters) to cover curtain walls
and isolated, independent towers (up to c. 225 square meters)
outside the enceinte, often within a few hundred yards of a forts
principal gate. Stefan convincingly argues that these monstrous
towers, even the so-called tower-palaces generally on the highest
point(s) of the interior of a Dacian fort, were artillery
platforms.24 Dacian artillery is known from both Cassius Dio and
Trajans Column.2522. E.g., Dacian use of monumental roofed streets
(scene CXIV=Stefan fig. 33) is confirmed by finds at Sarmizegethusa
Regia: Stefan 7476; on the water system: 7681, 9899; some earlier
views posited no springs or water system inside the city; for
Stefan (6012) the Columns scenes LXXVLXXVI (fig. 252: Decebaluss
surrender in 102) unquestionably show Sarmizegethusa Regia, but
Lepper and Frere (Trajans Column, 271) cite (captiously?) depiction
of polygonal rather than the ashlar walls known from the site as
proof that the Maestro had never seen the Dacian capital; but cf.
C. D. Stoiculescu, Trajans Column: Documentary Value from a
Forestry Viewpoint, Dacia 29 (1985): 8198: the Columns accuracy in
representing flora found in modern Romania. 23. For a critical
edition of the Greek text of Philo on siegecraft (=his Syntaxis
mechanike, Book 5) with French translation and commentary, see Y.
Garlan, Recherches sur de poliorctique grecque (Paris: cole
franaise dAthnes, 1974), 279404; English translation: A. W.
Lawrence, Greek Aims in Fortification (Oxford: Clarendon Press,
1979), 69107. 24. Three tables (27476) feature the measurements of
all known Dacian towers on the walls, within the walls, and outside
the walls. Stefan (271, caption to fig. 134) assumes a minimal
artillery range of 250 meters; cf. E. W. Marsden, Greek and Roman
Artillery: Historical Development (Oxford: Clarendon Press, 1969),
91, 13132: maximum effective range for both arrowshooters and
stone-throwers was 400 yards (c. 366 meters); Oltean (Dacia, 7880,
116) posits these tower-houses as status symbols of the Dacian
elite. 25. Of course Stefans view depends on the construction of
such towers not antedating Dacian possession of artillery, which on
present evidence would not be before Decebalus received military
engineers (mechanopoioi) from Domitian in the peace terms of 89
(Dio 67.7.4; cf. 68.9.5); ordnance captured earlier in Domitians
war is another possibility; cf. Stefan 436,
1198
THE JOURNAL OF
The Dacian Wars of Domitian and Trajan
MAP 2: Sarmizegethusa Theater (not to scale)
SR = Sarmizegethusa Regia SC = Sarmizegethusa Colonia
The limited topographical access to the Dacian forts (often only
one way up the hill) exposed attackers to a withering crossfire
from the independent towers in conjunction with fire from the
curtain-wall towers and the elevated interior towers (for example,
271 fig. 134: illustration of artillery coverage at Costesti).
Stefans emphasis on the significance of artillery in Dacian
defenses is new. His exploitation of the archaeological data drives
home what a grueling and onerous conflict of siegecraft and
mountain warfare the conquest of Dacia was. The northwestern
approach to Sarmizegethusa Regia from the Muresh River valley
(north of the central Carpathians) demonstrates the extent of the
Dacian defensive system. An army marching south toward the Dacian
capital must first take Costesti, a major religious center and
perhaps Burebistas capital before the construction of
Sarmizegethusa Regia. Located on a promontory (elevation c. 600
meters) overlooking a narrow defile and approachable only from its
southern side, Costestis double ring of walls encloses about nine
hectares. As the crow flies, one is only about 14 kilometers from
the Dacian capital. But 300 meters to the southeast across a valley
begins the tip of the mountain ascending to Sarmizegethusa. An
earthen rampart traceable for over 300 meters guards the initial
ascent; behind it are at least eighteen isolated towers, which
impede the approach to another fort, Blidaru, only 1,500 meters
from Costesti (as the crow flies) but 150 meters higher in
elevation. A third major fort, Piatra Rosie (Red Rock), c. 7
kilometers south of Blidaru, guarded the western and southwestern
approaches to Sarmizegethusa (c. 13 kilometers away) from the Strei
River valley. On51617; given the sophistication of Dacian
architecture, however, earlier possession of artillery could be
conjectured.MILITARY HISTORY
1199
EVERETT L. WHEELER
present evidence, this northwestern approach to Sarmizegethusa
Regia was the most heavily fortified. Besides this Costesti
corridor, Stefans astute birds-eye perspective has discerned two
circles of Dacian forts protecting access to the Dacian heartland
of the Carpathian interior and extending throughout this entire
mountain range. This new and more comprehensive understanding of
the Dacian defense system in turn leads to new topographical
arguments for the routes of Roman armies in 8889, 101102, and
105106, as passes dictated the avenues of access (discussed below).
Nevertheless, many of these hilltop forts have Hallstatt (Iron Age)
origins. Stefan generally assumes rather than proves that the
Dacian fortification system was initially Burebistas work with some
later improvements.26 Worthy of report is also Stefans remarkable
analysis (32355) of the Roman camps at Sarmizegethusa Regia.
According to Cassius Dio (68.9.7), Trajan left a stratopedon
(ambiguous whether: army camp or legion) at Sarmizegethusa after
Decebuluss capitulation in 102 to ensure Dacian compliance with the
terms of peace, which included dismantling Dacian defenses. Roman
garrisons also were scattered at other sites key in the war of
101102 south and west of the capital, including one at the later
Colonia Sarmizegethusa, besides several south and east of the
Carpathians.27 Despite Dios explicit testimony, a Roman camp at
Sarmizegethusa Regia, its garrison, and even the continued presence
of the Dacian king Decebalus at his capital after 102 have been
much debatedin part from disbelief that Decebalus could function as
a king and perpetrate violations of the 102 agreement with a Roman
camp at his very doorstep, and not least from the lack of
archaeological evidence for a Roman camp there in 102; further, Dio
refers to the Dacian capital, but never calls it (in extant
excerpts) Sarmizegethusa. Nevertheless, the Colonia, the only
alternative, was a Roman creation without evidence of pre-Roman
Dacian occupation.28 Some26. Stefan 11356 (Costesti), 156200
(Blidaru), 21829 (Piatra Rosie). New major Dacian fortified sites
continue to be discovered: one in 2004 near Poiana Brasnov and
Rsnov, comprising twenty-two to thirty hectares with twenty-two
towers, dated to at least 1416 A.D., and identified as the Cumidava
of Ptolemy (Geography 3.8.8); another at Cetatea Znelor near
Covasna with fortifications dating first century B.C.first century
A.D.; details in Strobel, Die Eroberung Dakiens, 11112. 27. A
burned layer beneath the site of the later forum of the Colonia may
attest a Roman camp destroyed by the Dacians at the start of the
second war in 105: see I. Piso, Les lgions dans la province de
Dacie, in Le Bohec and Wolff, eds., Les lgions de Rome, 1:209, and
Les dbuts de la province de Dace, 32021; Stefan 28081 with n. 41;
Strobel, Die Eroberung Dakiens, 108 with n. 12. The partial
deconstruction of many Dacian forts in 102 (e.g., Costesti,
Blidaru) and hasty reconstructions in 105 are clear in the
archaeological record: Stefan 65457; cf. Column scene CXXXII=Stefan
fig. 272 with caption. E. Sauers critique of Stefan on this point
(review of Stefan, American Journal of Archaeology 112 [2008]: 196)
seems captious, as Sauer cannot cite a specific historical context
(other than 105) for use of such fortifications near Blidaru for
Dacian resistance; Roman forts outside the Carpathians: Stefan
64243. 28. Lepper and Frere (Trajans Column, 3049) discuss the
problem of the name Sarmizegethusa for both sites; Ptolemy
(Geography 3.8.4), writing within a generation of two of Trajans
Dacian wars, lists Zarmizegethousa to Basileion, that is, the royal
residence, but omits the Colonia; cf. Piso, Les dbuts de la
province de Dace, 322.
1200
THE JOURNAL OF
The Dacian Wars of Domitian and Trajan
would even have Decebaluss kingdom after 102 reduced to the area
north of the Muresh River with a new Dacian capital established at
Piatra Craivii (the Apoulon of Ptolemy, Geography 3.8.8 [?], c. 20
kilometers north of modern Alba Iulia) and the whole theater of the
second war (105106) transferred to extreme northern Dacia. Piatra
Craivii, a rocky protrusion (elevation 1,083 meters), fortified in
the second half of the first century B.C. (era of Burebista) and
later topped by a small (67 by 36 meters) medieval citadel
(thirteenth-fifteenth centuries), suffered destruction by fire. As
often for pre-Roman fortified Dacian sites, that destruction is
assumed to be Trajanic, but no evidence ties Decebalus to Piatra
Craivii or attests a transfer of his residence north of the
Muresh.29 Certain is that the southern end of the curtain wall at
Sarmizegethusa Regia, enclosing Terraces IVVI, was dismantled.
Decebaluss palace, thought to have occupied the highest elevation
within the city (Terraces IIII) was left untouched. Stefan asserts
that the rampart (vallum) of a Roman camp of 102105, incorporating
Dacian architectural fragments, can be traced on the ground and
enclosed an area of six to seven hectares. It included part of the
former southern interior of the city, but also extended 100 to 200
meters south of the former defensive wall. Stefans aerial
photograph (to this reviewers eye) does not prove the case.
Besides, he would have the Romans copying Dacian construction of
walls 3 meters thicknot normal Roman practice.30 Decisive in
support, however, is an argument from stratigraphy. Excavation of
part of the wall of the Roman camp of 106 along Terrace V revealed
Roman renovations: the camp wall of 106 rested on two layers of
filling (nearly 1 meter thick): beneath one layer the remains of a
Roman forge came to light, which had been installed over the site
of a Dacian mint at a lower level. That Roman forge buried under
the level of the 106 camp must surely belong to the camp of 102105.
Whatever its precise perimeter, the camp of 102105 no longer seems
in doubt. The camp of 106 was built on higher ground after Romans
raised the surface of the entire southern half of the former city
and leveled off the area, including obliteration of the palace
complex on Terrace I.3129. Decebalus and Piatra Craivii: C.
Opreanu, Bellum Dacicum Traiani, Dacia 50 (2006): 11820,
reasserting his earlier views: e.g., The Consequences of the First
Dacian-Rumanian War (101102). A New Point of View, in Gonzlez, ed.,
Trajano Emperador de Roma, 39697; on the site, see Stefan 24755;
contra Opreanu, Piso, Les dbuts de la province de Dace, 297 n. 2;
Strobel, Die Eroberung Dakiens, 112 with n. 35. 30. An objection
raised by I. Bogdan Cataniciu, Dacias Borders under Trajans
RuleRemarks, in Visy, ed., Limes XIX, 726 n. 24, 727 n. 31, and
Cataniciu, Daci i Romani. Aculturatie n Dacia (Cluj-Napoca:
Academia Romana, 2007), 127; Piso (Les dbuts de la province de
Dace, 304 with n. 5) and Strobel (Die Eroberung Dakiens, 111) also
reject a 102 Roman camp; Oltean (Dacia, 56) favors a 102 camp.
Certainly 3-meter thick walls are un-Roman, but use of Dacian
architectural fragments in building the camp wallsa practice even
more extensive in construction of the Roman camp walls of 106 after
the citys destructionas well as other evidence (see below) speak
against this objection. The Romans did incorporate part of the
original city wall in the northwest side of the camp of 106. 31.
Stefans view of the 102105 camp develops an earlier thesis espoused
by A. Diaconescu, who (to no surprise) endorses Stefans position in
his review: Dacia and the Dacian Wars,MILITARY HISTORY
1201
EVERETT L. WHEELER
No such problems beset the camp of 106, for its outline in the
current forest is clearly discernible from aerial photographs (20
fig. 4, 332 fig. 168). This much smaller camp (2.8 hectares) on
Terraces IV overlapped only partially the camp of 102105 and
incorporated the area of the former palace. Two gates, six towers,
a trace of barracks, a cistern, and (of course) a bath (outside the
camp to the south and exploiting earlier Dacian water works) are
known, although the camp has not been studied or dug in detail. The
respective garrisons of these camps remain controversial.32
Evidence from the fill used for the camp walls of 106 attest the
presence of vexillationes (detachments) of the legions II Adiutrix
and VI Ferrata for the garrison of 102105, which in Stefans
reconstruction of events would have been wiped out or taken
prisoner by the Dacians in 105.33 Fragments used for fill between
wall faces in 106 must surely represent older discarded material.
Less clear is a detachment of IV Flavia Felix at the 102105 camp,
although a vexillatio of this legion manned the site in 106 and
later, when its chief duty was destruction of the Dacian capital.
The size of both camps indicates a Roman presence not even close to
the strength of a full legion (c. 5,000 men), but rather
approximates the size of camps allotted to auxiliary units of 1,000
or 500 men.34 How long after 106 or 107 troops remained at a city
now
592; cf. his Dacia under Trajan. Some Observations on Roman
Tactics and Strategy, in Beitrge zur Kenntnis des rmischen Heeres
in den dakischen Provinzen, ed. N. Gudea, Acta Musei Napocensis
34.1 (1997): 1822; Opreanu, The Consequences, 397401; on the mint,
producing Dacian imitations of Roman denarii dated 126 B.C., 68
B.C., and 1437 A.D., see Stefan 71. 32. Stefan 34854; for a
different view, see Piso, Les lgions dans la province de Dacie,
21113. 33. Stefan (34951) persists in including I Adiutrix in the
102 garrison despite C. Opreanu, Legio I Adiutrix in Dacia.
Military Action and its Place of Garrison during Trajans Reign, in
Roman Frontier Studies. Proceedings of the XVIIth International
Congress of Roman Frontier Studies, ed. N. Gudea (Zalau: County
Musuem of History and Art, 1999), 57172, followed by Piso, Les
dbuts de la province de Dace, 304 with n. 50. Oltean (Dacia, 56)
would include the legion IV Flavia Felix in the 102 camp. The
fragmentary text attesting a vexillatio of the VI Ferrata (AE 1983
nr. 825) suggests restoring VI Ferrata (a legion based in Syria) in
a lacuna detailing the transfer of detachments of eastern legions
to the 101102 Dacian war, noted in the career inscription of C.
Iulius Quadratus Bassus (consul 105), who later died fighting the
Sarmatians (c. 118) as Hadrians first Dacian governor (Smallwood,
Documents Illustrating the Principates of Nerva, Trajan and
Hadrian, no. 214): E. Dabrova, The Governors of Roman Syria from
Augustus to Septimius Severus (Bonn: Rudolf Habelt, 1998), 88 with
n. 889. This new text of VI Ferrata from the 102105 camp disproves
previous reconstructions of this legions history, which put it the
105106 war: e.g., K Strobel, Zu Fragen der frhen Geschichte der
rmischen Provinz Arabia und zu einigen Problemen der
Legionsdislokation im Osten des Imperium Romnum zu Beginn des 2.
Jh. n. Chr., Zeitschrift fr Papyrologie und Epigraphik 71 (1988):
25253, followed by Speidel (Bellicosissimus Princeps, 35), who
erroneously thinks the whole legion went to Dacia. 34. On the size
of legionary and auxiliary camps, see Y. Le Bohec, The Roman
Imperial Army (New York: Hippocrene Books, 1994), 16162.
1202
THE JOURNAL OF
The Dacian Wars of Domitian and Trajan
destroyed and its population forced to migrate is unknown.35 As
usual, Romans destroyed and burned the camp of 106 when it was
abandoned. It should not be forgotten, however, that Sarmizegethusa
Regia was the chief cult site of Dacian religion and located on its
holy mountain. Directly across the Alb River valley from the
capital and at the same elevation (1,800 meters away, as the crow
flies, but c. 4,500 meters for someone descending one slope and
ascending the other), lay Fetela Alba, another cult site with three
sanctuaries. It shared the capitals fate of total destruction in
106 after damage in 102.36 The rise of the Dacian empire under
Burebista in the first century B.C. had involved religious fervor,
in which the holy man Decenaeus had played a role.37 Romans in 106
wiped out not only a Dacian kingdom but also native Dacian
religion. A Roman garrison at Sarmizegethusa may have lingered
longer than the time needed to destroy the city to ensure
enforcement of the new prohibition of the native Dacian
cults.38
35. A unit of Germaniciani exploratores, permanently stationed
at Orashtioara de Sus, not far from the entrance to the Costesti
corridor, perhaps had among its duties surveillance of the former
sacred Dacian district: Piso, Les dbuts de la province de Dace,
306; on the unit, see N. Gostar, Ein numerus Germanicianorum
exploratorum im oberen Dacia, Germania 50 (1972): 24147, although
his identification of this unit with German infantry seen on
Trajans Column is dubious: exploratores are generally mounted
units; on this Roman camp, see Stefan 283; N. Gudea, Der dakische
Limes: Materialen zu seiner Geschichte, Jahrbuch des
rmischgermanischen Zentralmuseums Mainz 44 (1997): 1045. 36. For a
classic example of a new archaeologists attempt to turn an event
like the destruction of Sarmizegethusa Regia and Fetele Alba into a
process, see Locklear, Late Iron Age Background, 5051, who also
(45) claims Fetele Alba was unfortified; contra, Stefan 21317. 37.
For one view of Decenaeus, see R. Vulpe, Dcne, conseiller intime de
Burbista, in his Studia Thracologica (Bucharest: Editura Republicii
Socialiste Romnia, 1976), 6268. Although the accounts of Decenaeus
include many Greek topoi associated with lawgivers and creators of
civilization (cf. A. Szegedy-Maszak, Legends of the Greek
Lawgivers, Greek, Roman and Byzantine Studies 19 [1978]: 199209),
it is a mistake, ignoring valid information in Strabo, to reject
Decenaeus as a myth: sic P. Heather, Goths and Romans 332489
(Oxford: Clarendon Press, 1991), 36. 38. Dr. Mihai Popescu (Paris)
advises me that absolutely no trace of the native Dacian religion
survived in Roman Dacia. See his La religion dans larme romaine de
Dacie (Bucharest: ditions de lAcadmie Roumaine, 2004); cf. Oltean,
Dacia, 111, 200; Babe, Devictis Dacis, 33136; Piso, Les dbuts de la
province de Dace, 31617; D. Ruscu, The Supposed Extermination of
the Dacians: The Literary Tradition, in Hanson and Haynes, eds.,
Roman Dacia, 7585. Lepper and Frere (Trajans Column, 3078, 318) on
a survival of Dacian religion are in error; see also G. Florea and
P. Pupeza, Les dieux tus. La destruction du chef-lieu du Royaume
dace, in Piso, ed., Die rmischen Provinzen, 28196. Romans did wage
religious warfare on occasion, and the religious and cultural
contexts of Roman battles are generally not appreciated: see my
Shock and Awe: Battles of the Gods in Roman Imperial Warfare, Part
I, in LArme romaine et la religion sous le Haut-Empire, ed. Y. Le
Bohec and C. Wolff (Paris: De Boccard, 2009): 22567; a Part II is
in preparation.MILITARY HISTORY
1203
EVERETT L. WHEELER
Stefans second book (43684, 66793) treats the art and
iconography celebrating the victories of Domitian and Trajan. A
single thematic chapter might have been preferable to chronological
placement, that is, after the campaigns of Domitian and Trajan
respectively. Much here will interest art historians more than
their military counterparts.39 Nevertheless, art plays an important
role in Stefans continuation of attempts to rehabilitate Domitians
reputation, a trend begun with Stphane Gsell (1894) and reflected
in recent biographies.40 Readers of Tacitus, Suetonius, Pliny the
Youngers Panegyric to Trajan in 100, and even Cassius Dio are
familiar with Domitians bad press from writers of the senatorial
class, although he was not disliked by the army, whose pay he had
raised. The Senate decreed a damnatio memoriae after his
assassination in 96. Hence Domitian vanished from the public
record: his name was erased from all public documents and
monuments. Moreover, what Domitian had proclaimed as a Roman
victory in Dacia in 89, Trajan, especially after his annexation of
Dacia in 106, could recast in the public eye as a humiliating
defeat now avenged. Trajan readily added Dacicus to his titulature,
an honor that Domitian had declined. Forgetting Domitian would not
have been easy for the Eternal Citys inhabitiants, as the extent of
Domitians building projects in Rome rivaled those of Augustus.
Trajans Forum and Market superseded Domitians Forum Magnum, still
under construction at the time of his assassination. Many large
fragments of sculptures from colossal victory monuments,
transferred to depots of marble, escaped destruction. The two
gigantic trophies, sitting today on the balustrade of39. Stefans
painstaking work in assembling his material from studies of
architectural and sculptural fragments and archaeologists endless
debates on the topography of the city of Rome cannot be fully
expounded here. Curiously, Stefan does not engage more fully with
J. E. Packers massive study, The Forum of Trajan in Rome: A Study
of the Monuments, 3 vols. (Berkeley/Los Angeles: University of
California Press, 1997), although Packer is occasionally cited.
Indeed yet another study of Trajans Forum and Column has recently
appeared: M. Galinier, La colonne trajane et les forum impriaux
(Rome: cole franaise de Rome, 2007), reviewed by Packer, The Column
of Trajan: The Topographical and Cultural Contexts, Journal of
Roman Archaeology 21 (2008): 47178. 40. S. Gsell, Essai sur le rgne
de lempereur Domitien (Paris: Bibliothque des coles franaises
dAthenes et de Rome, 1894); B. W. Jones, The Emperor Domitian
(London: Routledge, 1992); P. Southern, Domitian: Tragic Tyrant
(London: Routledge, 1997); see also K. H. Waters, Traianus
Domitiani Continuator, American Journal of Philology 90 (1969):
385405, and a conference at Toulouse: J.-M. Pailler and R.
Sablayrolles, eds., Les annes Domitien (Toulouse: Presses
universitaires du Mirail, 1994). Stefan (699 n. 5) castigates F. M.
Ahl, The Rider and the Horse: Politics and Power in Roman Poetry
from Horace to Statius, Aufstieg und Niedergang der rmischen Welt
II.32.1 (1984): 40110; R. Saller, Domitian and his Successors,
American Journal of Ancient History 15.1 (1990 [2000]): 418; and
contributors to A. J. Boyle and J. W. Dominik, eds., Flavian Rome
(Leiden: E. J. Brill, 2003), for perpetuating the literary
traditions negative views.
III
1204
THE JOURNAL OF
The Dacian Wars of Domitian and Trajan
the Capitol at Rome and representing Germania and Dacia, are (as
attested by an inscription) Domitianic.41 Similarly, as Stefan
argues (47277 with fig. 209), the Dacian frieze, re-used on the
Arch of Constantine and assumed to be Trajanic, is in reality
Domitianic. For historical reliefs Domitian had to surpass the
celebrations of the Jewish War seen on his older brother Tituss
Arch; likewise Trajan had to excel (hence the Column) Domitians
displays.42 Trajans Forum, in fact, may have included as many as
eighty-two colossal statues of Dacians, excelling in number, size,
and expense (use of the rarest high-quality marble) Augustuss forty
of Orientals commemorating his so-called Parthian accord of 20
B.C., which involved the return of lost legionary standards.43
Stefan (690) may well be correct that of all barbarians only
Parthians and Dacians (not Germans) occupied a special place in the
Roman public memory. Domitians artistic celebrations of his Dacian
war may also be credited with establishing in Roman iconography the
female type of a personified Dacia and the conventional portrayal
of Dacians, which Trajan exploited.44 As Stefan demonstrates
(contrary to some views), Domitian did commemorate his Danubian
wars on coins, which for the first time depicted Dacian arms.45
Further, Domitian revived the concepts both of the emperor on
horseback (his colossal equestrian statue known only from coins and
poetic references) and of the emperor as a leader at the front in
major warsa model followed by Trajan and Marcus Aurelius.4641.
Erroneously associated with Mariuss victories over the Cimbri and
Teutones two centuries earlier by A. Goldsworthy, The Complete
Roman Army (London: Thames & Hudson, 2003), 164; Stefan 456 n.
249. 42. Stefans thesis of Trajans borrowing from Domitianic
projects finds support in Aurelius Victor (De Caesaribus 13.5) and
is not criticized in the two reviews (so far) of Stefan in major
Anglophone archaeological journals: Sauer (above, note 27), 19596;
Diaconescu, Dacia and the Dacian Wars, 58994; cf. Coulston,
Overcoming the Barbarian, 41619. 43. For Augustuss Parthian accord
as a contrivance of propaganda and one of the great non-events of
Roman history, see E. L. Wheeler, Roman Treaties with Parthia:
Vlkerrecht or Power Politics? in Limes XVIII. Proceedings of the
XVIIIth International Congress of Roman Frontiers Studies Held in
Amman, Jordan (September 2000), ed. Philip Freeman et al. (Oxford:
Archaeopress, 2002), 1:28792. Stefan (535 with n. 214), like
Mattern, Rome and the Enemy, 60, falls victim to Augustuss
propaganda: the standards lost in Mark Antonys Parthian war (36
B.C.) had already been returned in 33 B.C. 44. An earlier
personification of Dacia, labeled in Greek ethnos Dakon (nation of
Dacians), had appeared among the figures of conquered peoples
(gentes captae) in the late Neronian Sebasteion at Carian
Aphrodisias in Asia Minor: see Stefan 46162, 484. The Roman
populace had first seen Dacians in 29 B.C., when Octavian
(Augustus) staged a combat of Dacians vs. German Suebi as part of
his victory celebration for the defeat of Marc Antony: Cassius Dio
51.22.6. 45. Coulstons remarks (Overcoming the Barbarian, 392, 421)
on the Trajanic origin of depictions of Dacians and their weapons
should now be modified. 46. On the portrayal of Trajan on his
Column as the perfect general conforming to the principles of
Onasanders Strategikos, see G. Picard, Tactique hellnistique et
tactique romaine: le commandement, Comptes Rendus de lAcadmie des
inscriptions et belles-lettres, 1992, 17883.MILITARY HISTORY
1205
EVERETT L. WHEELER
Notable, too (for Stefan), is Domitians concern for Roman war
dead. Commemoration of the slain on state-sponsored monuments was
not a Roman custom. Germanicuss recovery and proper burial (15
A.D.) of the dead from Quinctilius Varuss Teutoburg Forest disaster
(9 A.D.) offered the only precedent.47 Atop the hill of Adamklissi
(extreme southeastern Romania), Trajan erected in 108109 a
cylindrical Tropaeum Traiani, 44.44 meters in diameter and at c. 37
meters tall, approximately the same height as his Column in Romea
monument visible north of the Danube (10 kilometers away); its
attraction and location on major north-south, east-west routes
spawned a town (Municipium Tropaeum Traiani) flourishing into the
Late Roman period. The Tropaeum Traiani would be merely an
interesting detail were it not for the survival of fifty-four
metopes (1.48 meters high), which once decorated the frieze of this
monument (destroyed by an earthquake c. 500). These remarkable
examples of early second-century provincial art illustrate, as
Stefan believes, the Dacian and Rhoxolan counterattack across the
Danube in the winter of 101102, not only supplementing Trajans
Columns scenes of the same operations, but also providing a
different view of Dacian and Roman armor and weapons.48 As the
metopes, however, were not found in situ, whether they tell a story
and in what sequence they should be read provoke discussion. Yet
two other enigmatic monuments share the site: on another hill 127.5
meters north-northwest of the Tropaeum lies a Roman-style mausoleum
with little datable material and 255 meters east of the mausoleum
stands a monumental altar inscribed with an extremely fragmentary
casualty list estimated at 3,000 to 4,000 dead, to which the
titulature of either Domitian or Trajan could be restored.49
Debate47. Tacitus, Annals 1.6162; Suetonius, Caligula 3.2; Cassius
Dio 57.18.1; on the Roman attitude, see G. Clementoni, Germanico e
i caduti di Teutoburgo, in Dulce et decorum est pro patria mori. La
morte in combattimento nelantichit, ed. M. Sordi (Milan: Vita e
Pensiero, 1990), 197206; J. Rpke, Wege zum Tten, Wege zum Ruhm:
Krieg in der rmischen Republik, in Tten im Krieg, ed. H. von
Stietencron and J. Rpke (Munich: Verlag Karl Alber, 1995), 23334;
cf. M. Reuter, Gefallen fr Rom. Beobachtungen an den
Grabinschriften im Kampf getteter rmischer Soldaten, in Visy, ed.,
Limes XIX, 25563. 48. The metopes, now housed in a museum at the
site, are reproduced in F. B. Florescu, Das Siegesdenkmal von
Adamklissi: Tropaeum Traiani (Bonn/Bucharest: Verlag der Akademie
der Rumnischen Volksrepublik, 1965); cf. Stefans republication of
nine of them at fig. 233, which he believes depict the Romans
surprise attack by night on the Dacians and Rhoxolani at scene
XXXVIII of the Column. A recent view that the Dacian-Rhoxolan
attack occurred somewhere north of the Danube (e.g., Bennett,
Trajan Optimus Princeps, 9293) has little to support it,
particularly as the invaders are shown on the Column (scene XXXI)
swimming a major river, surely the Danube, and the Dacians captured
a slave of Laberius Maxiimus, the governor of Moesia Inferior
(Pliny, Letters 10.74.1): see Stefan 56068, and on the Tropaeum
Traiani: 693 with figs. 27778. 49. ILS 9107; on Roman
identification of battle dead and construction of such casualty
lists, see D. Peretz, Military Burial and the Identification of the
Roman Fallen Soldiers, Klio 87 (2005): 12338; Charless denial (The
Flavio-Trajanic Miles, 669) of any Domitianic association with the
site, based in part on A. Poulter, The Lower Moesian Limes and the
Dacian Wars of Trajan, in Studien zu den Militrgrenzen Roms III:
Acten des 13. Internationalen Limeskongrees, Aalen 1983, ed. C. Unz
(Stuttgart: Landesdenkmalamt Baden-Wrttemberg, 1986),
1206
THE JOURNAL OF
The Dacian Wars of Domitian and Trajan
rages whether the mausoleum and the altar are originally
Domitianic or the whole site is Trajanic. In Stefans
reconstruction, Adamklissi marks the site of a major Roman defeat,
when the Dacians initiated Domitians war in late 84 or early 85 by
an attack across the Danube and killed the provincial governor of
Moesia, Oppius Sabinus. Domitian (pace Stefan) later erected a
victory monument (Tropaeum Domitiani = the mausoleum) west of the
altar (89 or 90?) to celebrate his Dacian war, although this
Tropaeum was destroyed probably in 96 as part of the damnatio
memoriae and Domitians name was erased from the altar.50 During
Trajans first war, as Stefan argues, the Dacian counteroffensive
(with their Rhoxolan allies) in the winter of 101102, after Roman
forces had gained access to the Costesti corridor to Sarmizegethusa
Regia, followed the same route as the attack of late 84,51 crossing
the lower Danube into the Dobrudja and sweeping south with the
Danube on their right past Adamklissi and west into Roman Moesia
Inferior. Trajan halted the Dacians and Rhoxolani 45 kilometers
south of Novae and commemorated the victory by founding the city of
Nicopolis ad Istrum (Victory City at the Istrus River), of which
the construction (pace Stefan) appears in scene XXXIX of Trajans
Column, where (besides civilian settlers) soldiers are shown
working with chisels, thus indicating building in stone (that is, a
city) rather than the earth-wood construction of an army camp.
After the war Trajan built at Adamklissi his own Tropaeum,
dedicated to Mars Ultor (the Avenger) to obscure the earlier
Domitianic monuments, just as he obliterated Domitians building
projects at Rome. Sic Stefan, generally following traditional views
on the 101102 campaign. Something happened at Adamklissi, a rather
isolated site hitherto of no significance, to occasion a rare
state-sponsored war memorial; the dedication of the51928, is too
sweeping. Poulters case for the legion I Minervas connection with
Adamklissi monuments is not convincing, as Poulter cannot establish
that Durostorum (Bulgarian Silistra) c. 60 kilometers from
Adamklissi, was ever a base of the I Minerva. This legions
participation in Trajans Dacian wars is not at issue: see Strobel,
Untersuchungen zu den Dakerkriegen Trajans, 8687; Y. Le Bohec,
Legio I Minervia (1er-IIe sicles), in Le Bohec and Wolff, eds., Les
lgions de Rome, 1:8385; cf. on the enigimatic text ILS 4795, on
which Poulters case is partially based, E. Wheeler, A New Book on
Ancient Georgia: A Critical Discussion, Annual of the Society for
the Study of Caucasia 67 (199496): 7071 with n. 76. 50. Stefan
43738, 44244 with fig. 191; the attacks date of late 84 or early 85
derives from Dacian arms, signaling military action, depicted
already on Roman coins dated to January March 85: see 400, 402; the
semi-popular work of I. M. Ferris, Enemies of Rome: Barbarians
through Roman Eyes (Stroud, U.K.: Sutton Publishing, 2000), 70-71,
concedes a Domitianic origin of the mausoleum and the altar. 51. J.
Wilkes, Les provinces danubiennes, in Rome et lintgration de
lEmpire 44 av. J.-C.260 ap. J.-C., II: Approches rgionales du
Haut-Empire romain, ed. C. Lepelley (Paris: Presses universitaires
de France, 1998), 263, would put the Dacian inroad of late 84 west
of the Iron Gates Gorge; Strobel (Die Donaukriege Domitians, 43)
locates the attack east of Novae (Bulgarian Svishtov), the base of
the legion I Italica under the Flavians; on this important site,
see J. Kolendo and V. Bozilov, eds., Inscriptions grecques et
latines de Novae (Msie Infrieure) (Paris: De Boccard, 1997); M.
Absil, Legio I Italica, in Le Bohec and Wolff, eds., Les lgions de
Rome, 1:22738.MILITARY HISTORY
1207
EVERETT L. WHEELER
Tropaeum Traiani to Mars Ultor suggests revenge for a defeat.
Geographical details and literary narratives are lacking. Oppius
Sabinus, consul with Domitian in early 84, subsequently (perhaps
already later in 84) became governor of Moesia (in 86 divided into
two provinces, Superior and Inferior, as a result of the Dacian
war). The Dacians plundered Moesia in 85 and were still in Moesia
in 86.52 A second Roman defeat followed (although not necessarily
in Moesia and possibly in 87 rather than 86) with the loss of the
Praetorian Prefect Cornelius Fuscus and a legionary or praetorian
standard. Stefan (405) equates the severity of the situation to the
Teutoburg Forest debacle. Positing Adamklissi as the site of
Oppiuss defeat goes back to Sir Ronald Syme (1928). Likewise
conjectural is Stefans route of the attack through the Dobrudja in
both 8485 and 101102, reasoning that the Dacians attacked the less
well-defended sector of Moesia, as no legions were yet stationed on
the Danube east of Novae, although scattered units of auxilia and
the Moesian fleet (established by Vespasian) surveyed the area
between Novae and the Danubes mouth.53 For the Dacian-Rhoxolan
offensive of winter 101102 (absent in Cassius Dios fragments and
Jordanes) interpretative possibilities are reduced to scenes on
Trajans Column and the metopes of the Tropaeum Traiani. Nicopolis
ad Istrum, however, as the site of Trajans victory finds
confirmation in two literary sources, and the capture of the
governor of Moesia Inferiors slave would seem to attest Dacian
operations south of the Danube. Diaconescu, highly critical of
Stefans views of the campaigns associated with the Adamklissi
monuments and an advocate of their exclusively Trajanic origin,
would derive from the scenes of the Column and the metopes a second
Roman victory of Trajan at Adamklissinot the first attempt to tie
Adamklissi directly to the events of 101102.54 The Adamklissi
monuments remain enigmatic, but on present evidence Stefans viewsby
no means definitive and largely in accord with traditional
interpretationsmake sense of what is available52. Jordanes, Getica
76 (the fullest account); Eutropius 7.23.4; cf. Suetonius,
Domitianus 6.1; Tacitus, Agricola 41.2; excavations at Novae and
Viminacium (Serbian Kostolac; base of the legion VII Claudia) show
some destruction in this period possibly associated with Dacian
attacks; Stefan (400, 406) attributes the same destruction layer at
Viminacium to Dacian attacks in both 85 and 86. Available evidence
permits various reconstructions of the chronology. 53. R. Syme,
Rhine and Danube Legions under Domitian, Journal of Roman Studies
18 (1928): 47; Stefan 400 with n. 6, where he rejects an
alternative view (above, note 51), placing the Dacian attacks in
the sector between Viminacium and the Danubes Iron Gates Gorge;
cf., most recently, L. Petculescu, The Roman Army as a Factor of
Romanisation in the North-Eastern Part of Moesia Inferior, in Rome
and the Black Sea Region. Domination, Romanisation, Resistance, ed.
T. Bekker-Nielsen (Aarhus: Aarhus University Press, 2006), 3135;
for a survey of previous work on the problems of the Adamklissi
monuments, see Lepper and Frere, Trajans Column, 295304, although
they offer no real solutions. 54. Victory at Nicopolis: Ammianus
Marcellinus 31.5.16; Jordanes, Getica 101; Stefan 400 with
additional evidence, including coin hoards; Diaconescu, Dacia and
the Dacian Wars, 593; cf. Lepper and Frere, Trajans Column, 87; see
also above, note 48, for a view that there was no Dacian-Rhoxolan
invasion of Moesia Inferior in 101102. Scene XXXIX may indicate
Trajans foundation of Nicopolis ad Istrum, as Stefan believes, but
it seems unlikely that Trajan devoted
1208
THE JOURNAL OF
The Dacian Wars of Domitian and Trajan
.
Stefans third book (359441, 485673) covers the entire history of
the Dacians from the sixth century B.C. to the conclusion of
Trajans second war in A.D. 106. His extensive pre-history to the
wars of Domitian and Trajan seeks to validate his theses concerning
Hellenization of the Dacians (for example, Sarmizegethus Regia as a
Dacian version of Attalid Pergamum: 10911) and the rule of
Burebista and Decebalus over essentially Hellenistic kingdoms.
These thorny questions can only have their surface pricked here.
Greek influence on Dacian architecture and urbanism is less
problematic than the character of the Dacian state. Some impression
of these issues will precede discussion (V) of Stefans innovative
reconstructions of the topography of Domitians and Trajans
campaigns. Part II of this article will assess Dacians and Roman
strategy on the Lower Danube. Stefans emphasis on Dacians
marginalizes the conglomerate of other peoples (Greeks, Iranians,
Germans, Celts) sharing the Carpathian basinan ancient ethnic mix
antedating medieval complications (for example, Magyars, Slavs).
These other major players demand acknowledgement. In the Greek
Classical period Thracian tribes, of which the Getae were a branch,
dominated the area between the Strymon River (Thraces border with
Macedonia) and the Black Sea, and from the northern coast of the
Aegean Sea northward through the Carpathians and beyond the Siret
River, as well as westward through modern Bulgaria into Serbia. The
Odrysae, the most powerful Thracian tribe in the fifth and fourth
centuries B.C. south of the Haemus Mountains (modern Stara
Planina), figured prominently in the power politics of Athens and
Philip II of Macedon in the northern Aegean during the fourth
century B.C. An Odrysian dynasty retained varying degrees of
authority over much of Thrace into the forties B.C., when a dynasty
of the Sapaean tribe emerged. Rome tolerated Thrace as a
client-kingdom as long as order was preserved, but dynastic
disputes compelled annexation in 46 A.D.55 Historically, Macedonia
buffered Greece and the Mediterranean world generally from inroads
of Thracians as well as Illyrians northwest of Macedonia. After
Macedonias annexation in 146 B.C., it fell to the Roman governor of
Macedonia to deter threats anywhere in the Balkans, a task
subsequently incumbent on Roman commanders in Moesia (the area
between the Haemus Mountains and the Danube) from the time of
Augustus.56time and manpower to building and settling a city in the
middle of the 101102 campaign. If Stefan is correct, the scene may
be an anachronistic interpolation in the Columns narrative. 55. On
the Thracians, see now Z. H. Archibald, The Odrysian Kingdom of
Thrace: Orpheus Unmasked (Oxford: Clarendon Press, 1998); opinions
differ on whether the Dobrudja, including the Greek cities of the
coast, was annexed as part of Thrace in 46 and when the Dobrudja
became part of the Moesian province: cf. Suceveanu and Barnea, La
Dobroudja Romaine, 2528, and De nouveau autour de lannexion romaine
de la Dobroudja, in Piso, ed., Die rmischen Provinzen, 27180;
Petculescu, The Roman Army, 31, 41 n. 2; Lica, Coming of Rome,
14647. 56. Macedonia as buffer: Livy 33.12.1011; cf. 42.52.1;
Strabo 7.3.11; note Alexander the Greats campaign (335 B.C.) to the
Danube against the Triballi to secure Macedonias northernMILITARY
HISTORY
IV
1209
EVERETT L. WHEELER
Decline of Odrysian power in the fourth century B.C. favored a
Getic rise north of the Haemus. Two branches can be discerned: the
Getae proper, extending on both the northern and southern banks of
the Danube from the Iron Gates Gorge east to the Black Sea coast,
of which the Trizes/Terizes tribe was prominent in the Dobrudja,
and the Crozbyzes of the Carpathiansthe mountain Getae, which Latin
sources called Daci. Ptolemy, however, could list fifteen
Geto-Dacian tribes.57 Toponyms ending in -dava distinguish
Geto-Dacian sites. In the Classical Greek and Hellenistic eras, the
Getae often allied with the Thracian Triballi, with whom they
shared the cult of Zalmoxis, the chief deity of Dacian religion.
The Triballi stretched from Oescus (modern Gigen) on the Danube
southward and west into the Serbian mountains. By the time of Roman
occupation they were more a name than a people.58 Non-Thracians,
however, also inhabited or migrated into the Carpathian basin. From
the seventh century B.C. Greek colonies dotted the west coast of
the Black Sea: (for example) in Romania, Histria, Tomis (modern
Constantia), Callatis, and in Bulgaria, Dionysopolis (modern
Balcic), Odessus (modern Varna), and Mesembria (modern Nesebur). A
Greek market (emporium) even arose at some point far inland at
Axiopolis (modern Cernavoda, c. 60 kilometers due west of Tomis on
the northward bend of the Lower Danube), an oddity not matched by
Greek ventures into theborder before his Persian expedition:
Arrian, Anabasis Alexandri 1.46.11; cf. Strabo 7.3.8; Lica, Coming
of Rome, 38, 74. Battys denial of Augustuss annexation of Moesia
(Rome and the Nomads, 429; cf. 460) obscures the complex and poorly
reported history of Roman military activity on the Middle and Lower
Danube; the absence of a consular governor of a province of Moesia
before Claudius (r. 4154) does not preclude Moesia as a provincia
in the sense of a military command: on the problems, see R. Syme,
The Early History of Moesia, in Syme, The Provincial at Rome and
Rome and the Balkans 80 BCAD 14, ed. A. R. Birley (Exeter:
University of Exeter Press, 1999), 193220; cf. H. Wolff, Die
rmische Erschliessung der Rhein- und Donauprovinzen im Blickwinkel
ihrer Zielsetzung, in Rmische InschriftenNeufunde, Neulesungen und
Neuinterpretationen: Festschrift fr Hans Lieb, ed. R. F. Stolba and
M. A. Speidel (Basel: F. Reinhardt, 1995), 32527; and, most
recently, M. Mirkovic, Die Anfnge der Provinz Moesia, in Piso, ed.,
Die rmischen Provinzen, 24970. 57. Strabo 7.3.1213; Ptolemy,
Geography 3.8.14; Stefan 35960; cf. 37678: a superfluous commentary
on ancient authors preferences for Getae or Daci, synonymous terms
from Burebistas time. 58. Strabo 7.3.13; Syme, The Early History of
Moesia, 200201, 21718 with n. 110; Batty, Rome and the Nomads,
52022; Ptolemy, Geography 3.10.5: Oescus of the Triballians (Oiskos
Triballon). Oescus, the northern terminus of the shortest route
between Macedonia and the Danube, was the first legionary camp
downstream from the Iron Gates Gorge and became the base of the
legion V Macedonica perhaps as early as Tiberius (r. 1438), but
certainly during Claudiuss reign (3854): N. Gudea, Die Nordgrenze
der rmischen Provinz Obermoesian. Materialen zu ihrer Geschichte
(86275 n. Chr.), Jahrbuch des rmisch-germanischen Zentralmuseums
Mainz 48 (2001): 1112; cf. M. Zahariade and N. Gudea, The
Fortifications of Lower Moesia (Amsterdam: Adolf M. Hakkert, 1997),
72: a synopsis of the Roman forts history; Batty, Rome and the
Nomads (405; without sufficient evidence) would have Oescus as a
permanent legionary base from 2/3 A.D. or 11 A.D.; cf. Wilkes,
Roman Legions and their Fortresses, 102 for the V Macedonica at
Oescus as early as 9 B.C.
1210
THE JOURNAL OF
The Dacian Wars of Domitian and Trajan
interior elsewhere.59 Of equal significance, Iranian peoples
from the Ukrainian steppe were already migrating into Transylvania,
the plains of Wallachia, and the Dobrudja by the sixth century B.C.
The Agathyrsi, apparently a Thracian people on the Muresh River,
had a king with an Iranian name (Spargapeithes) in the early fifth
century B.C.; burials show a mixture of Geto-Dacian and Scythian
elements until the midfifth century B.C.60 Intermittent Scythian
migrations and raids from the Ukraine through the fourth century
B.C. later intensified, as the Scythians were pushed from the
steppe into the Crimea and the Dobrudja by the Sarmatians, another
Iranian people, whose successive waves (Iazyges, Rhoxolani, Aorsi,
Alani) dominated the steppe from the Don to the Dnieper, until the
Goths arrived in the third century and the Huns in the fourth. The
Dobrudja became Little Scythia (in Greek, Mikra Skythia), a term
revived in the Late Empire, when the Dobrudja (detached from Moesia
Inferior) became Scythia in the Diocletianic reorganizatio