The Occasion of Arrian's "Tactica" Wheeler, Everett L Greek, Roman and Byzantine Studies; Jan 1, 1978; 19, 4; Periodicals Archive Online pg. 351

The Occasion of Arrian's Tactica 

Everett L. Wheeler 

ARRIAN STATES in his Cynegencus 1.4 that his lifelong interests 
were philosophy, generalship and hunting.1 With the topic 
of generalship the Anabasis Alexandri immediately springs to 

mind, but Arrian's interests in the art of command went far beyond 
his historical pursuits and assumed a personal and contemporary 
significance.2 As legatus Augusti pro praetore of Cappa do cia (ca 131-137), 
Arrian held a military post on the Euphrates frontier second in 
importance only to the governorship of Syria. To this point in his life 
Arrian had led one of the most brilliant public careers of the Hadrianic 
period, especially for a novus homo from Nicomedia. He had been the 
first Greek (known so far) to govern a western province, when he 
served as proconsul of Baetica in the late 120' s; he became the first 
Bithynian to hold the consulship, probably in 129; and he was only 
the second Greek to command in Cappadocia. (c. Julius Quadratus 
Bassus was the first Greek governor, 107/8-110/11.) Nor was Arrian a 
novice at military command: he probably saw action in Trajan's 
Parthian war as a tribune and possibly served as legatus legion is on the 
Danubian frontier in the early 120'S.3 

In 135 deteriorating relations between Rome and Caucasian Iberia 
prompted pharasmanes II to summon the Sarmatian Alani through 
the Caucasus in a display of force. Cappadocia was threatened. Arrian 
assembled his army, deterred the Alani, and Roman territory was 

1 The following abbreviations are used in this paper: HARTMANN I = Karl Hartmann, 
"Flavius Arrianus und Kaiser Hadrian," Programm Augsburg (Augsburg 1907); HARTMANN 
II = "Uber die Taktik des Arrian," Programm Bamberg (Bamberg 1895); KIECHLE = Franz 
Kiechle, "Die 'Taktik' des Flavius Arrianus," RGKomm 45 (1965) 87-129; KLUMBACH= Hans 
Klumbach, "Zum Straubinger Romerfund," Gymnasium 59 (1952) 71-76; SANDER = Erich 
Sander, "Die Kleidung des romischen Soldaten," Historia 12 (1963) 144-66; STADTER = 
Philip A. Stadter, "The Ars Tactica of Arrian: Tradition and Originality," CP 73 (1978) 

117-28; WHEELER = Everett L. Wheeler, "Flavius Arrianus: A Political and Military Biog­
raphy" (Diss. Duke University 1977). 

2 For Arrian's military inclinations and their effect on his literary activity see Gerhard 
Wirth, "Zur Tigrisfahrt des Kaisers Trajan," Philologus 107 (1963) 291. 

8 I discuss Arrian's early career in detail in Wheeler 24-53. 

351 



352 THE OCCASION OF ARRIAN'S TACTICA 

saved from destruction (Cass. Dio69.15.1, cf Them. Or. 34.8,32).4 At a 
later time he wrote an account of his campaign, the Alanica, of which 
only a fragment describing his contingency plans remains-the Acies 
contra Alanos. I) 

So far as we know, Arrian received no official recognition for his 
efforts against the Alani. He could have received a vote of thanks from 
the Senate, public praise from Hadrian, or even the ornamenta 
triumphalia; but perhaps Arrian did not deserve commendation. The 
battle described in the Acies is merely a plan, not an historical 
account; a major engagement between Arrian and the Alani cannot 
be proved. Arrian, however, served an emperor jealous of his generals' 
accomplishments, especially in the East, where success would im­
mediately provoke comparison with Hadrian's abandonment of 
Trajan's conquests.6 Conceivably we find a rather obscure reference 
to the Alan cam paign on Hadrianic coins bearing the legend 
EXER(CITUS) CAPPADOCIAE S.C. Weber conjectured that these 
coins commemorated Hadrian's review of the Cappadocian forces in 
129, but the issue dates to the period 134/5-138, probably too late to 
recall an event of 129.7 This type appeared in a series commemorating 
a number of provincial armies although not all of them: Africa, 
Pannonia, Egypt and Arabia were not included.s If the Cappadocian 

'See most recently A. B. Bosworth, "Arrian and the Alani," HSCP 81 (1977) 217-55; 
Professor Bosworth was extremely kind in sending me a draft of this paper before its 
publication. I cannot agree with his views concerning either Roman-Iberian relations or 
the Acies (see n.5 infra). See Wheeler 54-135, 149-58, 169-75, 213-59, 283-305. 

6 Flavii Arriani Quae extant omnia II: Scripta Minora et Fragmenta, ed. A. G. Roos with 
additions by G. Wirth (Leipzig 1967) 177-85. I have argued in detail that the Acies (despite 
its peculiar style) must derive from a literary work which can only be the Alanica: Wheeler 
272-76. Contra, A. B. Bosworth, "Arnan's Literary Development," CQ N.S. 22 (1972) 185 n.2, 
and op.cit. (supra nA) 247-48; Stadter 119. I wish to thank Professor Stadter for allowing me 
to read his paper before its publication. 

8 Festus, Brev. 20; Jordanes, Romana 270; Bpit. de Caes. 14.10; Eutrop. 8.6.2, cf 8.7.1.; and 
Hist.Aug., Hadr. 21.8: Bxpeditiones sub eo graves nul/ae fuerunt; bella etiam silentio paene 
transacta. 

7 Wilhelm Weber, Untersuchungen V'r Geschichte des Kaisers Hadrianus (Leipzig 1907) 234; 
BMC, Imp. ill, cxviii, 498 no.1673; RIC II, 458 no.914; Anne S. Robertson, Roman Imperial 
Coins in the Hunter Coin Cabinet II (Oxford 1971) 159 no.588. 

8 BMC, Imp. ill, 497-504; RIC II, 457-62. An hitherto unknown type in this series has 
recently been found: EXERCITUS ALEXANDRI(NUS) without S.C. See Angelo Nicolini de 
Fazi, "Moneda inedita del emperador Hadrianus de la serie 'Adlocutiones' a sas tropas," 
Gaceta Numismatica (Barcelona) no.44 (Mar. 1977) 74-76 (non vidi), summarized in NL 99 
(Mar. 1978) 64 no.333. 



EVERETT L. ~EELER 353 

army earned recognition on the coins from the Alan campaign of 135, 
it was indeed an obscure reference to Arrian's activity. 

The following year, 136, Arrian composed a short treatise entitled 
TEXV'Y} 7'ctKTtK1} , his contribution to military theory. It is a peculiar work 
consisting of two parts. In the first section Arrian described the 
organization, drill and formations of an l-Iellenistic phalanx, while 

paraphrasing and often reproducing parts of an earlier treatise by 
Aelianus Tacticus, published about thirty-five years before and dedi­
cated to Trajan.9 The second section, an original composition of 
Arrian, recounted contemporary Roman cavalry exercises; its 
technical nature defied translation until 1964.10 

The seemingly disjointed structure of the treatise and Arrian's 
purpose in writing it have puzzled scholars for years. Schwartz 
believed it was an official report to which Arrian added the section 
on the Hellenistic phalanx to satisfy Hadrian's military interests. 
Davies asserts that Hadrian commanded Arrian to write the section 
on the cavalry exercises, while according to Bosworth Arrian simply 
reedited a standard textbook for military antiquarians.l1 Others 
claim a practical purpose for the work.12 Certainly the Tactica 
contains many contemporary references and possesses demonstrable 
relevance to the warfare of the period, but none of the explanations 
offered hitherto will account for the treatise as a whole. 

9 Aelian's TaK7"lK~ 8£wpta was quite similar to the T£xvTJ 'TaK7"lK~ of Asclepiodotus, a 
student of Posidonius, who earlier had composed a work of the same title. See K. K. 
Miiller, "Aelianus 10," RE 1 (1894) 482-86, and "Asclepiodotus 10," RE 2 (1896) 1637-41; 
PIR2 A 120; H. Kochly/W. Riistow, Griechische KTiegsschriftsteller 11.1 (Leipzig 1855, repro 
Osnabriick 1969) 77-79, 87-88; W. A. and C. H. Oldfather, in The Ulinois Greek Club, 
Aeneas Tacticus, Asclepiodotus, Onasander (LCL, London 1923) 230-43; Alphonse Dain, 
Histoire du texte d'Elien Ie tacticien (Paris 1946) 15-36; and Stadter 117-18. For a discussion 
of the whole problem see Wheeler 338-50, where I refute the theory of Dain, accepted 
by Stadter, that Aelian and Arrian independently used a lost Tactica derived from 
Posidonius' treatise. This lost Tactica, if it existed at all, cannot be conclusively dated 
to a period before Aelian. 

10 Charles Guischardt declined translating this part of the Tactica in volume II of his 
Mimoires militaires sur les Grecs et les Romains (Lyon 1760), and a century later Kochly and 
Riistow, op.cit. (supra n.9), did not include it in their Griech. Kriegsschriftsteller. Franz 
Kiechle, op.cit. (supra n.1), offered the first translation and commentary. 

11 E. Schwartz, "Arrianus 9," RE 2 (1896) 1233; R. W. Davies, "Cohortes Equitatae," 
Historia 20 (1971) 754; A. B. Bosworth, op.cit. (supra n.5) 183 and op.cit. (supra n.4) 242-44. 

12 Guischardt, op.cit. (supra n.lO) II 194 n.h; Adolf Bauer, Die griechischen Altertumer: Die 
Kriegsaltertumer (Munich 1893) 297; Max Jahns, Geschichte der Kriegswissenschaften vornemlich 
in Deutschland I (Munich 1889, repr. Hildesheim 1966) 99; Johannes Kromayer/Georg Veith, 
Heerwesen und Kriegfuhrung der Griechen und Romer (Munich 1928) 14; Kiechle 108-09; 
Stadter 119. 



354 THE OCCASION OF ARRIAN'S TACTICA 

I would argue rather that a unifying theme is to be found in 
Hadrian's vicennalia, although Arrian's ulterior motive in writing the 
piece, his desire for recognition, precludes assigning to the work a 
simplistic purpose. We must begin with the conclusion of the treatise, 
for the opening lines of the Tactica, which may have contained a 
dedication, are now lost. Arrian writes in 44.3: 

" , , ~ \ ~ R \' C\ 'A ~ \ , \ ~ , wc-rE EC rTJVOE rTJV 7Tapovccw ,..aCVIEtCW, TJV noptCWOC EtlCOCTOV TOVT 

ETOC {3aCtAEVEt, 7ToAV JLii'\'\ov gVJL{3atvEtV !LOt SOICE' T!X E7T7] Taiha .;prEp 

£C rTJv 7TaAat AalCESatJLova 

EvB' alXp,<X TE V€WV BillEt lCa;' JLwca Myna, 
lCa;' StlCa Evpvayvta lCaAwV £7TtTappoBoc EPYWV. 

The fragment of Terpander (fr.6 Bergk = fr.4 Diehl), which ends the 
treatise, and the reference to Hadrian's twentieth year are intentionally 
juxtaposed. As Arrian states, the verses apply very well to Hadrian 
and reflect his interests in military training, Hellenic culture and law. 
Even the selection of Terpander fits Hadrian's preference for archaic 
authors.13 Hadrian's twentieth year fell in 136, and these concluding 
lines of the treatise suggest that the ostensible motive of the Tactica 
was to provide an occasional piece celebrating the emperor's 
vicennalia.14 

Apart from Arrian's Tactica Hadrian's vicennalia is not attested in 
literary sources, and the exact date of the renewal of Hadrian's 
tribunida potestas is not known, although it coincided very closely with 
the calendar year. Hadrian, however, became the first emperor since 
Tiberius to wear the purple for twenty years. Such an anniversary 
demanded celebration, as calendars of the Imperial cult and coins 
testify.1S Anniversary issues of coins did not enjoy great vogue in the 

18 Hartmann I 37. Military interests: Casso Dio 69.3.2, 5.2, 9.1-6; Hist.Aug., Hadr. 10.2-11.2, 
21.8-9; ILS 2487, 9133-35; Hellenic culture: Hist.Aug., Hadr. 1.5, 13.1-2,6,14.8-11,16.10-11; 
Casso Dio 69.3.1, 16.1-2; law: Hist.Aug., Hadr. S-9, IS, 21.1, 22.11; archaic preference: 
Hist.Aug., Hadr. 16.2, 5-6. 

14 Tact. 44.3 gives a precise date for the treatise. This is the only work of Arrian for which 
an undisputed date of composition can be determined. Although Arrian mentions the 
death of Cotys II of the Bosporan kingdom in Peripl. 17.3, an event which can be assigned 
to 131/2, this reference does not offer more than a terminus post quem for the treatise. See 
Wheeler 200-01,205 n.219, 396. 

16 The vicemutlia was cdebrated 13 December 136 on a calendar of the Imperial cult, 
P.Oslo ill 77, lines 15-16= E. Mary Smallwood, Documents ... ofNerva, Trajan and Hadrian 
(Cambridge 1966) no.147, which is close to 10 December, the traditional date for renewal 
of the tribunicia potestas. 



EVERETT L. WHEELER 355 

first three decades of the second century, but vota soluta and vota 
suscepta, common legends for announcing an anniversary, began to 
appear about 137. Strack believes there was an unusual abundance 
of aurei, which suggests a congiarium.I6 Thus Hadrian by no means 
passed over the anniversary in silence, and I suggest that neither did 
t:he governor of Cappadocia. 

It in no way speaks against the occasional nature of the piece that 
the treatise was apparently not directly addressed to Hadrian. The 
opening lines of the Tactica are lost, and Arrian always refers to 
Hadrian in the third personP He is called ,BaetA€Vc in Tact. 32.3, 42.2, 
4, 44.1, but he is mentioned by name only in Tact. 44.3, the very place 
where Arrian makes reference to the vicennalia. Just as Arrian is 
usually taciturn about himself in his works, he is equally reluctant 
to make direct reference to contemporary figures. Nor did the genre 
of military theory written under the Empire require that a treatise be 
addressed to the reigning emperor.IS Moreover, Arrian wrote the 
Tactica in Cappadocia and he would have had no opportunity to 

recite his piece for Hadrian. Although we have no basis by which to 
judge the extent of Arrian's literary fame in the 130' s, 19 it seems 
probable that a treatise by a man already as prominent in public 
service as Arrian, and holding a major military command, would 

16 Michael Grant, Roman Anniversary Issues (Cambridge 1950) 99 with n.2; H. Mattingly, 
"The Imperial 'Vota'," ProcBritAc 36 (1950) 158-59, 184 nn.15-16; and especially Paul L. 
Strack, Untersuchungen zur romischen Reichspritgung des zweitenJahrhunderts, II: Die Reichs­
pritgung zur Zeit des Hadrian (Stuttgart 1933) 184-87. BMC, Imp. m nos.775-77, 782; Strack 
nos.286-89, 699-700, 469, 248, 675, 249, 269-71. Mattingly offers some alternative explana­
tions for the legends vota soluta and vota suscepta, but his arguments are not compelling, 
and he often concedes Strack's views for the vicennalia. If Strack's date of 137 for the coins is 
correct, then Hadrian waited for the completion of his twentieth year before issuing most 
of the anniversary coins, and he did not anticipate his vicennalia on the coins, a common 
practice in the Late Empire. For our purposes, the date of the Taaica is 136, Hadrian's 
twentieth year. Whether most of the anniversary coins appeared in 136 or 137 has no 
bearing on our argument. Arrian's treatise need not have a direct correlation to the 
anniversary issues. 

17 Noted by Stadter 119. 
18 Karl Hartmann, II 20, claims all new military works had to be addressed to the 

emperor. Excluding Arrian's Tactica, four of the nine military tracts written during the 
first four centuries and preserved in toto or in extensive fragments were not addressed to 
an emperor: Frontinus' Strategemata, Ps.-Hyginus Gromatkus' De Munitionibus Castrorum, 
Onasander's (TpaTTl'Y'KOC and Heron's B€A07TOUK&. For the redating of Heron to the second 
half of the first century see E. W. Marsden, Greek and Roman Artillery: Technical Treatises 
(Oxford 1971) 208--{)9. 

19 I do not accept Bosworth's date for the Anabasis Alexandri between 115 and the early 
120's, and only presumption asserts a date for the Dissertationes Epicteti before Arrian's 



356 THE OCCASION OF ARRIAN'S TACTICA 

arouse public as well as imperial attention by its very novelty: duces­
scriptores were not a common phenomenon in the second century. 
Thus the Tactica can easily have been composed in honor of Hadrian's 
vicennalia without a direct address to the emperor. 

In view of the lost proemium of the Tactica and the reference to 
Hadrian by name in Tact. 44.3, we should examine Tact. 32.3, where 
Arrian concludes the section on the phalanx and announces the 
discussion of cavalry exercises to follow. In the sole manuscript 
(Laurentianus gr. 55.4) the sentence reads: 

, , ~ \ \ t' \ I "tp "'" f ~ I Y ,_ E"yw OE" Ta t7T7TtKa yvp.vaeta, oca wp.awt t11'7T7JC yvp.va."OV'Tat, €V Tep 
, 't: \() , .. \ , \ " ,/.(} '" \ ~ , ~ l7apOV'Tt E"17E"5E"1\ wv, OTt Ta l7a')"1] . ..•.•• Kat E"'f' T/V 0T/I\wcat €V rn 

.L:: fI • \ , ~ ~ Q \ , '.1. '''' " cvyypa'f'[/ T/VTtVa V1TE"P aV7'OV TOV fJactl\E"WC CVVE"ypa'f'a, TOOE fLOt ECTat 

T€AOC TOO AOyOV TOO TaK'TtKOO. 

«Now after relating in detail the cavalry exercises which the Roman 
horsemen practice, seeing that I have already explained the [infantry 
drill 1] in the treatise which I wrote in behalf of the emperor himself, 
this will be the conclusion of my Tactica." The corruption 
1TCXyrJ • ••••• • Kat was emended by F. Haase to l7€'tKeX, by C. Muller 
to l7ap' "EAA'T/CtV. Either emendation yields essentially the same 
meaning of the sentence: l7€'tKO: to contrast with the {l71TtKO: YVf'vancr., 

or l7ap' "EAA'T/ctv to counterbalance the Roman material to follow. 
Adoption of Haase's l7€'tKO: in the Teubner edition has resulted in 
confusion. Kiechle takes the phrase 'TO: l7€'tKeX .. . CVV€ypcr..pcr. as a 
separate report by Arrian on ludi castrenses sent to Hadrian earlier, 
while several others conjecture a lost Tactica of Arrian on Roman 
infantry drill dedicated to Hadrian.20 Since a lost Tactica is not other­
wise attested, Hartmann offered a more plausible solution. Accepting 
Muller's emendation, he took the phrase to refer to the first section 
of the Tactica, which concerns the Hellenistic phalanx.21 

Acceptance of Hartmann's view affords another explicit reference 
to Hadrian and the purpose of the Tactica. The treatise is written 
{J1TEP cr.thoO 'ToO f3cr.nA€Wc. Since Arrian ended the work with a reference 
to Hadrian and now we find another precisely in the middle of the 

governorship of Cappa do cia. See Bosworth, "Arrian in Baetica," GRBS 17 (1976) 62-64, with 
the same views repeated in his "Errors in Arrian," CQ N.S. 26 (1976) 117-19. I discuss the 
dates of Arrian's works in Wheeler 395-419. 

10 Kiechle 88; Roos, op.cit. (supra n.5) xxxiv; Stadter 119. 
t1 Hartmann II 17 and I 34 n.l. 



EVERETT L. WHEELER 357 

piece, the probability of lost introductory remarks concerning 
Hadrian is substantially increased.22 

A vicennalia was a time for glorifying an emperor and his rule. In 
this context we can see how Arrian interlaces his work with references 
to Hadrianic policy and the themes of the reign. In Tact. 44.1-2 Arrian 
extols Hadrian's cavalry reforms, carefully noting their consistency 
with the mos maiorum. Hadrian introduced horse-archers on the 
Parthian-Armenian model, the wheelings and counter-wheelings 
associated with the charge of Sarmatian and Celtic KOY7"OcPOPOt, and 
the various complex forms of long-range fire in attack (a,Kpo{3oAtCiwl). 
He also urged that each national group of auxilia continue to use its 
native war cry and instituted exercises for leaping ditches and walls. 
Other minor improvements were prescribed to increase the cavalry's 
elegance in public performances and quickness in battle.23 

Furthermore, Arrian's cavalry exercises of Tact. 34-43 do not have a 
didactic purpose. Arrian writes as a spectator from the reviewing 
stand watching a show (Tact. 38.2); his aim is to describe a festive 
performance, not to instruct his reader on the mechanics of cavalry 
drills. In fact his descriptions sometimes lack lucidity (e.g., 38.1, 40.2) 
and are often mere summaries of the action, especially toward the 
end of the treatise (e.g., 43.1-4). The exercises, although inspired by 
practice for battle, are performed for display and for the amusement 
of the onlookers; Arrian repeatedly refers to the beauty, grace, 
elegance and the astounding qualities of the exercises (Tact. 34.2, 4, 
5; 35; 36.4; 38.2-4; 40.7, 12; 44.2). 

The maneuvers described by Arrian were not serious training in 
techniques and weapons or a general review, such as Hadrian con­
ducted at Lambaesis in 128 (ILS 2487,9233-35), but sportive and festive 
exercises called ludi castrenses or armatura.24 All units of the Roman 

ii Hartmann II 18. He also notes, I 34 n.1, that if Arrian wrote a lost Tactica on the 
Roman infantry, his discussion of the testudo in Tact. 11.4-5 would be a needless repetition 
from the earlier work. Stadter, 121 n.15, calculates that Arrian's preface was probably 
about twenty-six Teubner lines. 

23 Only the list of Tact. 44 constitutes Hadrianic reforms. The exercises described in 
Tact. 34-43 are pre-Hadrianic. See Friedrich Lammert, Die romische Taktik zu Beginn der 
Kaiserzeit und die Geschichtschreibung, Philologus Suppl. 23.2 (Leipzig 1931) 49. Kiechle, 126, 
claims the Hadrianic reforms of Tact. 44.1-2 are unique because for the first time the 
Romans adopted the practices of nations not in Roman service; but in fact 'stimulus diffu­
sion', the dissemination of new techniques, weapons, etc. through contact with other 
peoples, was already common. 

24 Cf Vegetius 2.23: A rmaturam, quae festis die bus exhibetur in circo .... 



358 THE OCCASION OF ARRIAN'S TACTICA 

army practiced the armatura and all soldiers learned a form of 
armatura, although the festive performances were reserved for a 
special unit, also called armatura, under a decurio or duplicarius if 
cavalry, and probably under an exercitator armaturarum, often an 
evocatus, if infantry.25 This special unit consisted of the most efficient 
performers and received special training as well as higher pay. An 
armatura was intended to create an image of Roman military capability 
which could never be achieved throughout the whole army. The 
differences between Arrian's armatura in Cappadocia and Hadrian's 
inspection at Lambaesis are first, magnitude, and second, competitive 
show as opposed to demonstration of training. Arrian's exercises 
involved units specially trained for shows, whereas Hadrian reviewed 
the maneuvers of a whole legion, whole auxiliary cohorts and alae.26 

Arrian's description of the participants' armament and attire (Tact. 
34.2-35) confirms that the exercises are sport rather than training. The 
riders make a spectacular entry onto the parade-ground wearing 
gilded iron or bronze helmets with yellow plumes made from horses' 
manes and embroidered Cimmerian tunics of scarlet or dark blue. 
Their helmets cover the whole head and face with openings only for 
vision and breathing. More than thirty examples of these helmets with 
face-masks have been found, most notably the eight at Straubing in 
Bavaria. The Straubing helmets divide into four Hellenistic types 
and four oriental. Apparently the difference in style of helmets 
distinguished the two contesting parties in the armatura and would 
also account for the difference in the color of the tunics.27 Trousers 
are also worn, bound tightly around the legs rather than loose in the 
Parthian or Armenian fashion. The riders carry a form of scutum but 

25 Evidence collected by Sander 153 n.89a, 15!H>1. Also see CIL X 3344; Not.Dig.Or. 11.9 
and Oc. 9.6; and R. W. Davies, "The Training Grounds of the Roman Cavalry," Arch] 125 
(1968) 73-100. 

28 I discuss the similarities and differences between the Cappadocian armatura and the 
maneuvers at Lambaesis in more detail in Wheeler 3~2. If Arrian were discussing a 
general review of all cavalry in the Cappadocian army, the individual competition in 
aKOV'Tlc,...OC (Tact. 42), in which each participant is called forward by name to perform, 
would surely produce among the spectators the tedium loathed by Hadrian in military 
displays. Cf Hadrian's remarks at Lambaesis (ILS 3194): Difficule est cohortales equites etiam 
per se placere, dijJidlius post alarem exerdtationem non displicere; alia spatia campi, alius iaculan­
tium numerus . .. Verum vos f<lStidium calore vitastis . ... Also see Sander 159 and Klumbach 
71. 

27 For the helmets see Klumbach 73-74; Kiechle 117; Sander 1~1; Graham Webster, 
The Roman Imperial Army (London 1964) pI. xxna; Paul Couissin, Les armes romaines (Paris 
1926) 410-22; and H. Russell Robinson, The Armor of Imperial Rome (New York 1975) 107-35. 



EVERETT L. WHEELER 359 

lighter in weight and painted.28 Their weapons are small javelins 
without iron heads (&KOJlTta),29 which in later exercises are replaced 
by gvcTa oopaTa (Tact. 40.4), heavier javelins, and by lanceae (Tact. 
41.2).30 For protection from errant missiles the horses also wear 
armor, frontlets for the eyes (7TPOP.€Tw7Tt8ta) and flank covers 
(7Tapa7TA£tJpifHa) in addition to saddle cloths.31 Such colorful and 
expensive equipment would hardly serve as <fatigues' in training. 

We should note also that one of the most colorful and surprising 
aspects of Arrian's armatura is the use of dracones, the hissing cloth 
snakes used as standards, which Arrian calls CKv8tKCt. c'Y}/L€La and 
describes fully in Tact. 35.3-4. The origin of dracones is debated but 
was without doubt barbarian.32 Arrian states that the horsemen of 
the armatura are distinguished by Roman standards and the dracones 
(Tact. 35.2), which could correspond to the different colors of tunics 
and possibly different styles of helmets. The dracones are not simply 

I have not seen the recent republication of the Straubing finds: J. Keirn/H. Klumbach, 
Ver rlhnisc/re Schatvund von Straubing (Miinchener Beitrage zur Vor- und Friihgeschichte 
3 [1951], 2nd ed. 1976). 

B8 The famous painted shield from Dura-Europus may be an example of the type Arrian 
mentions. See Franz Cumont, "Fragment de bouclier portant une list d'etapes," Syria 6 
(1925) 1. 

28 Headless practice javelins called pila praepilata or hastilia are attested and were also 
used in gladiatorial schools: Polyb. 10.20.3-4; Liv. 26.51.4; BAfr. 72; Pliny, HN8.17;Veg.1.14; 
Quint. Inst. 5.12.17. Arrian mentions cavalry greaves in Tact. 4.8 but does not include them 
in the equipment of the armatura. Greaves with movable knee protectors made from 
bronze sheet metal with gold and silver plating were found with other equipment of an 
armatura at Straubing. See Klumbach 74. 

80 For the final phases of the performance in which lanceae are used, the participants 
change into regular battle dress with iron helmets, breastplates and heavy scutum (Tact. 
41.1), but the exercises remain competitive. 

81 Eight gold and silver plated 7TPOP.€TW7Tl8,a were discovered at Straubing. Sander 160; 
Klumbach 73-74. Photographs of a 7TPOP.€TW7Tl810" from Straubing and horse armor from 
Dura-Europus in Webster, op.dt. (supra n.27) pIs. xxnb, xx. 

82 Scythian origin: Webster, op.cit. (supra n.27) 136; Sarmatian origin: Ortwin Gamber, 
"Dakische und sarmatische Waffen auf den Reliefs der Traianssaule," ]KS 60 (1964) 62; 
Thracian origin: Kiechle 12~22; Parthian origin: Emilio Gabba, "SuIle influenze reciproche 
degli ordinamenti militari dei Parti e dei Romani," in La Persia e il mondo greco-romano 
(AcLincei, Problemi 76, Rome 1966) 62. We can dismiss Gabba's view since his only evidence is 
Lucian, Hist.conser. 29, which cannot be trusted for accurate military details. Kiechle 
argues that a Scythian origin is not archaeologically attested, while a Thracian dragon can 
be shown to antedate the arrival of the Scythians. Kiechle's case is strong but not conclusive. 
We know the Scythians strongly influenced the Thracians, who borrowed their shield, the 
pelta, from the Scythians. See J. K. Anderson, Military Theory and Practice in the Age of 
Xenophon (Berkeley 1970) 113, 30~1 nn.9-10. 



360 THE OCCASION OF ARRIAN'S TACTICA 

a showpiece. The draconarius is the tactical leader of his unit, and the 
men are trained to follow his maneuvers (Tact. 35.6). Possibly by the 
late second and certainly in the third and fourth centuries the dracones 
became regular standards in the Roman army.33 They appear on 
Trajan's Column in the Dacian army of Decebalus, and Arrian is the 
first source to attest Roman use of dracones. Quite possibly Hadrian 
introduced dracones in the Roman army along with the other barbar­
ian practices mentioned in Tact. 44.1-2. Thus we may have still 
another allusion to Hadrianic reforms in the Tactica. 

Moreover, Arrian's description of individual competition in 
aKOVTtC}"Oc (Tact. 42) merits our consideration. He writes that good 
army commanders (oco, &ya8oc. c-rpaTtac TJYEP.OVEC) call out the names 
of the participants according to rank, beginning with the decurio of 
the unit. As each man is summoned, he rides out with three lanceae to 
perform a series of throws at two targets, hurling two at one target 
while charging straight ahead and the third at a second target while 
simultaneously wheeling his horse to the right. He says that the 
addition of the second target and the requirement of hurling during a 
right wheel were specified by Hadrian's order (Ka'T<1 7Tpoc-rag,v 
f3anAlwc), probably a reference to an official drill manual. The 
passage is notable because Arrian refers to Hadrian (f3anAEvc) three 
times in this section, more than in any other part of the treatise, and 
because he concludes Tact. 42 with a direct statement about his own 
activity. He writes that he commended the unit with the greatest 
number of outstanding hurlers because this unit was well trained for 
real warfare. Arrian's purpose here is quite subtle. By injecting at 
the end of this section his own action, about which he is usually silent 
throughout the work, and by the repeated mention of Hadrian's 
drill instructions, Arrian seeks to demonstrate that he conscientiously 
follows the emperor's orders to the letter in the public performance 
of the armatura, and to identify himself as &yal/dc cTpa-rdic .fJYE/u.fJJl. 

If Hadrian's vicennalia furnished the occasion for composition of 
Arrian's Tactica and the cavalry exercises of Tact. 34-43 constitute the 
festive military show of an armatura, it is reasonable to conclude that 
the two are related. The Cappadocian army celebrated the emperor's 
vicennalia with an armatura, which Arrian quite appropriately in­
corporated in his treatise. Ludi militares were often held for occasions 
such as an imperial birthday (Hist.Aug., Maxim. 2.3-3.5). A direct I. Evidence in Kiechle 121. 



EVERETT L. WHEELER 361 

reference to the purpose of the games is not given because this would 
be obvious to the reader from the nature of the exercises and pre­
sumably from the treatise's lost introduction. 

In addition to the armatura still other Hadrianic passages await our 
attention. Arrian discusses in Tact. 11.1-2 thinning and increasing the 
depth of the phalanx. Offensively a deep formation is needed to drive 

back the enemy with a compact mass, just as Epaminondas used at 
Leuctra and Mantinea, while defensively a thick formation should be 
used for repelling an enemy charge, such as against the Sarmatians 
and Scythians. No doubt Arrian's mention of Sarmatians is a direct 
reference to his own Alan campaign of the previous year; the discus­
sion of this defensive formation resembles Arrian's plans given in 
Ades 15-16, 26. The example of Epaminondas, however, provokes 
curiosity. One might expect Arrian to mention Alexander. Pausanias 
8.11.8 provides an explanation: Hadrian composed an epitaph for the 
grave of the famous Theban general, which implies Hadrian's 
admiration. Thus Arrian cited Epaminondas because he was a known 
favorite of Hadrian. Moreover, Arrian sought through the indirect 
reference to his Alan campaign to associate himself with the tactics of 
the Theban commander and to promote his own accomplishments.34 

Another peculiar passage is Tact. 33, the transition between the 
sections on the phalanx and the cavalry exercises. Arrian states that 
his task in the following section will be difficult because the Romans 
have adopted Celtic terminology as well as Celtic practices. Arrian 
exaggerates, for in the body of the second section he employs only 
three Celtic terms: petrinos (37.4), xynema (42.4), toloutegon (43.2), and 
on Spanish practice, the Cantabrican charge (40.1). His claim is an 
attempt to continue the procedure announced in Tact. 1.2, and 
borrowed from Aelian (Tact. 1.3-4), to define technical terms for 
novice readers.35 

Celtic cavalry techniques become a pretext for a panegyric on 
Roman borrowings from other peoples, a topos found in many 
authors.36 Arrian writes in glowing terms of the Roman genius (Tact. 

1& Hartmann, II 15, claims that Arrian's reference to Epaminondas shows use of Xenophon. 
The connection is allegedly illustrated through the word ~p.f3oAov in Arr. Tact. 11.2 and 
Xen. Hell. 7.5.22, 24. The argument is hardly convincing. 

15 Lammert, op.at. (supra n.23) 48; Kiechle 115. 
Ilpolyb. 1.20.5-16,6.25.11; Posidonius, FGrHist87 F 59= Ath. 6.273E-F; Diod. 5.40.1-2, 

23.2; Ined.Vat. 3 (Hermes 27 [1892] 118); SaIl. Cat. 51.37-39; Cic. Resp. 2.16.30; Aur.Vic. 
Caes. 11.13. 



362 THE OCCASION OF ARRIAN'S TACTICA 

33.3-5): "So in fact you would find that the Romans took military 
equipment from others-and now it is called Roman, because they 
made the best use of it. They also took military exercises (yvp.vacta 

c-rpa'nw'nKa) from others, the chairs of their magistrates [sella curulis], 
and the toga praetexta."37 Arrian does not stop here. He continues 
with Roman adoption of Greek religious practices and the Phrygian 
rites of Attis and Rhea, and the Athenian origin of the earliest Roman 
law-the Twelve Tables. Kiechle has already noticed that Arrian's 
panegyric reveals his complete support for Hadrian's introduction of 
barbarian tactics, for which exempla maiorum are cited as proof that 
Hadrian's policy is not revolutionary.38 We should also note, however, 
that the borrowed yvp.vaCLa cTpa'nw'nKa of Tact. 33.3 could be a direct 
reference to Hadrian's reforms listed in Tact. 44.1-2 already men­
tioned.39 Most significantly, Tact. 33 taken as a whole is a direct 
reflection of the cosmopolitan atmosphere of Hadrian's reign-the 
diversity of every corner of the ancient world made Roman. 

Finally, Arrian made a conscious effort to Romanize the first section 
of the Tactica, on the Hellenistic phalanx. In Tact. 18.3 he equates the 
Hellenistic cavalry unit of 512 men, Z1T1Tapxla, with the Roman ala,40 
and in Tact. 10.3 he notes that the commander of a unit of 100 men is 
called an €KaTOvTCfpx'Y/c, the Greek term for a Roman centurion.41 

Brief notes on Roman equipment and contemporary references also 
occur. In contrast to the Greeks of old who wore two greaves, the 
Romans use only one, attached to the leg extended in battle (Tact. 
3.5).42 References to the charge of the Alan and Sarmatian KOVTOcp6pOt 

11 cf Onasander 30, and Ovid, Met. 4.428: Pas est et ab Iwste doceri. 
as Kiechle 126. 
ae Hartmann I 36. Kiechle, 88, prefers to interpret the exercises as the games called 

Troia. See Verg. Am. 5.548f and Suetonius in Servo ad Am. 5.602. 

'0 Whether 512 should be taken as the exact number in a Roman ala or only an approxi­
mation is debated. See G. L. Cheesman, The Auxilia of the Roman Imperial Army (Oxford 
1914, repro Chicago 1975) 26-27, 29; Davies, op.cit. (supra n.11) 753. 

U The equation of the centurion and the Hellenistic 'TaeU1.pxrlC or €Ka'Tov'Tapxr1C is found 
already in Asclepiodotus, Tact. 2.8, and repeated in Adian, Tact. 9.3. Thus insertion of 
£KaV'TOV'Tapx'I]C in the treatise is not original with Artian, but its equation with centurion is 
clear even in Asclepiodotus. 

U Sources differ on which leg the greave was worn. Left leg: Liv. 9.40.3 (Samnites); 
Juv. 6.256-58 (gladiators); Sil.Ital. 8.419 (Sabines). Right leg: Veg. 1.20; Verg. Aen. 7.689-90. 

Polyb. 6.23.8 says only that greaves were worn. Couissin, op.cit. (supra n.27) 350-67, argues 
that Arrian's statement is an anachronism from Republican times. In the Imperial period 
greaves are attested only on stdae for centurions, officers, urban cohorts and praetorians. 
Sander. 155-57, accepts Arrlan. 



EVERETT L. WHEELER 363 

as well as to Armenian and Parthian horse-archers appear in Tact. 
4.3-certainly another allusion to Arrian's personal experience on the 
Euphrates frontier. British chariots used against the Romans are 
mentioned (Tact. 19.2-3) ;43 and we find the phalangic formation 
cvvac7TLCf-L6c equated with the Roman testudo (Tact. 11.4--6, cf Acies 
26).44 At the conclusion of his discussion of different types of cavalry, 
Arrian digresses on the Roman cavalry of his own time (Tact. 4.7-9). 
He denotes two basic kinds: KOVTOcp6pOt and lancearii. The former 
charge in the manner of the Alani and the Sarmatians, which is 
probably a reference to the Hadrianic reform mentioned in Tact. 44.1. 
Arrian goes on to list the standard equipment of the Roman cavalry­
men: the spatha, scutum, iron helmet and small greaves. The lancearii 
use their weapon for hurling at long range or for hand-to-hand 
combat, for which the spatha is also used. Some carry a mace besides.41> 
In the text of an otherwise dry account of the Hellenistic phalanx 
these Roman references enliven the treatise and add contemporary 
relevance. 

Thus Arrian's Tactica is best interpreted as an encomium on the 
occasion of Hadrian's vicennalia. He has interlaced the work with 
repeated references to the emperor's reforms, personal preferences 
and the atmosphere of the reign. Sometimes the references are direct 
and clear, but often his veiled allusions appear quite subtle. In a 
manner equally clever Arrian frequently mentions the Sarmatians 
and the Alani, certainly an attempt to spread the fame of his 
own campaign in Transcaucasia; and in Tact. 42 he portrays 
himself as the' good commander', conscientiously following Hadrian's 
drill instructions. Arrian's desire for recognition should not surprise 
anyone, since he boasts openly about his fame and his literary talent 
in Anab. 1.12.4-5. 

48 See Stadter 125 nn.31-32. 
"I treat the significance of this equation in detail in "The Legion as Phalanx," forth­

coming in Chiron 9 (1979). 
45 The reference to the spatha and mace as weapons for close combat echoes Arrian's 

instructions in Aaes 31. For the mace see Couissin, op.at. (supra n.27) 387-88; DarSag I 
(1909) s.v. "clava," 1237-38 and figs. 1581-83; Franz Cumont, "Le sacrifice du tribun 
romain Terentius et les Palmyreniens a Doura," MonPiot 26 (1923) 34 fig.5, 35. Stadter, 123 
n.24, corrects Webster, op.at. (supra n.27) 151, who claims Arrian's mace is an ax. The 
Romans did, however, add axes to their armament for use against Sarmatian lassoes. See 
Jacques Aymard, Essai sur les chasses romaines (BEFAR 17l, Paris 1951) 495, and Wheeler 
281-82. 



364 THE OCCASION OF ARRIAN'S TACTICA 

Nevertheless, Arrian's motive is puzzling. If one accepts the con­
ventional view that Arrian and Hadrian were close personal friends,46 
there seems to be no problem: Arrian has written an anniversary 
piece for his friend Hadrian. A close personal relationship between 
this Greek and the Graeculus, however, cannot be proved; the pertin­
ent evidence is weak and circumstantia1.47 Throughout the treatise, 
Arrian writes to please Hadrian (to attract his attention?) and con­
tinually shows his support for the emperor's military reforms. Could 
not Arrian's motive be private and personal? 

Arrian's Cappadocian command probably ended in 137. He dis­
appears from the sources until the 140's, when he reappears as a 
private citizen in Athens and archon in 145/6.48 The shadowy end of 
Arrian's Roman career has prompted the conjecture that he fell into 
disfavor with Hadrian, a view which cannot be directly supported by 
evidence.49 But when we consider that Arrian probably left Roman 
service within a year or two after the Tactica's composition, the 
treatise can be seen in a different light. Arrian employs the occasion 
of the vicennalia as a pretext to laud the emperor's military reforms, 
to appeal to Hadrian's military interests, as well as to demonstrate 
his approval of the regime's military policies and to remind both 
Hadrian and the general public of his recent defense of Cappadocia. 
As a consular legate, Arrian had reached the upper limit of an imperial 
career, and few options for further public service remained open to 
him. If the emperor chose to retain his military services, he might 
govern Moesia Inferior or perhaps even Syria, the most prestigious 
eastern post. Relieved of command and reduced to the lot of a 

.. E.g., Schwartz, op.cit. (supra n.ll) 1231; Hartmann IT 31 and "Zu den Schriften und 
Fragmenten des Flavius Arrianus," BPW 30 (1910) 606; Henry Douicet, Quid Xenophanti 
debueTit Flavius ATTianus (Paris 1882) 7, 88; Franz Ferdinand Schwarz. "Arrian's IndOre on 
India: Intention and Reality," East and West N.S. 25 (1975) 198. 

417 See Wheeler 47-49. 
u ILS 8801 (IGRR III 133); Werner Eck. SenatoTen von Vespasian Ins Hadrian (Munich 1970) 

215 n.418; Arr. Cyn. 1.4; IG ITt 2055. 
" Schwartz. op.cit. (supra n.11) 1230. 1231. In a similar manner G. W. Bowersock connects 

the end of Arrian's career with Hadrian's displeasure with the four sophists C. Avitius 
Heliodorus. Valerius Eudaemon. Favorinus and Dionysius of Miletus: see Greek Sophists in 
the Roman Empire (Oxford 1969) 52. Arrian. however. was not a sophist; and it is probably a 
mistake to compare a consular legate in a major military command with sophists. Cf 
Gerhard Wirth, "Arrian und Trajan-Versuch einer Gegenswartsdeutung," StudClas 16 
(1974) 190 n.83. The conjecture that Arrian was governor of Syria in the late 130's or early 
14O's is unproven and awaits epigraphical confirmation. See G. A. Harrer. "Was Arrian 
Governor of Syria?" CP 11 (1916) 338-39. 



EVERETT L. WHEELER 365 

private citizen, he could participate in the Senate and possibly crown 
his career with the proconsulship of Asia or Africa. The staid life of a 
senator would be pleasing to many but surely not to a man whose 
passion was generalship (Cyn. 1.4). In short, after probably at least 
twenty years in Roman service Arrian had to face a third possibility, 
the most dreaded prospect of an active man-retirement. I suggest 

that the Tactica was a subtle plea to prolong his military career. 50 

DUKE UNIVERSITY 

September, 1978 

60 A version of this paper was read at the Atlanta meeting of the American Philological 
Association 30 December 1977. I wish to thank Professors K. J. Rigsby and P. A. Stadter for 
beneficial criticism and hdpful suggestions.