Auxiliary deployment during trajan’s parthian war: Some neglected evidence from asia minor

(1)

vl\

Some Neglected Evidence from Asia Minor (1)

A recently published diploma for auxiliary veterans released from the garrison of Moesia Superior between 1 May and 31 August 115 reveals that at least nine of the auxiliary units assigned to the province on the day the document was certi-ﬁed were ‘translatis in expediti[0nem]’ — ‘absent on campaign’ (2). The units in question consisted of one cavalry regiment, the ala Praetoria Singularum, and

eight cohortes, the I Thracum Syriaca ', I Montanorum ', I Cilicum ', I Cisipaden-sium ; III Augusta Neruiana Pacensis ; IIII Raetorum ; V Hispanorum ; and VII

Breucorum c.R. Given the date of the document, the campaign is clearly Trajan’s Parthian War : preparations for this were well under way by the autumn of 113,

when Trajan left Rome for Antioch to make ready for his advance into Armenia

Major Via Satala in the spring of 114 (3). As such, then, the new diploma consti-tutes a valuable addition to our knowledge concerning the auxiliary units deployed for that campaign, a subject that has previously deﬁed exact analysis (4).

(1) This paper stems from work in 2006 while an Honorary Fellow at the School of Archaeology, Classics and Egyptology, Liverpool University 2 I am glad to formally rec-ognize the support given me there, especially by Dr.Alan Greaves. I am also most grate-ful to Dr. Jean Greenhalgh for her continued support in my epigraphic researches ; and to my Bilkent colleague Dr.Jacques Morin, for assistance with matters Greek. Most of all, however, I wish to acknowledge the great debt I owe Dr. Timothy Mitford (Christ Church, Oxford), for his exceptional generosity in allowing me to report the third text discussed here in advance of his own more detailed analysis of the same.

(2) W. ECK and A. PANGERL, Traians Heer im Partherkrieg. Z14 einem neuen Diploma aus dem Ja r 115 in Chiron 35, 2005, p. 49-67 Unfortunately, as is so often the case today, thi diploma is an unprovenanced item from the ever-expanding market for such document note, for example, how all ten diplomata published in two more recent papers by the same authors are unprovenanced but thought to be from the Balkans : W. ECK and A. PANGERL, Neue Diplomefu'r die Hilfstruppen von Brittania in ZPE 162, 2007, p. 223— 34 ; and IDEM, Weitere Militc’irdiplomeﬁir die mauretainischen Provinzen, loc cit., p. 235-47). Diplomata attain a greater economic value in the antiquities market from scholarly discussion of their contents : perhaps it is time to actively discourage the trade in those that lack a secure provenance by refusing academic comment on them : surely the short-term effect on scholarly analysis and reputation would be more than compensated for in the longer term by a wealth of securely provenanced items.

(3) J. BENNETT, Trajan Optimus Princeps, London, 2000, p. 183-204, esp. 189-91. (4) Cf. F. A. LEPPER, Trajan ’s Parthian War, Oxford, 1948, p. 173-74 : ‘The evidence for auxiliary units [involved in the Parthian War] is too scattered and not sufﬁciently com-plete’.

(2)

The editors of the new diploma were aware of and duly noted the documen-tary evidence available already indicating that two of the units listed therein,

namely the cohortes IIII Raetorum and VII Breucorum, were present in Asia

Minor at some time in their history, and most probably in connection with Trajan’s Parthian War (5). On the other hand they neglected to address other evi-dence showing how at least another six and possibly seven auxiliary units regu-larly recorded in the Danubian region before and after that campaign were also in Asia Minor at some point in the early 2““l century, and thus presumably there in connection with the same operation. They are : the cohors I Thracum Syriaca and (possibly) cohors I Montanorum, both listed as ‘translatis in expediti[one]’ on the new diploma ; the ala I Claudia Noua, also from Moesia Superior ; the

ala I Flauia Augusta Britannica >< c.R., and the cohors I Campanorum c.R.,

both from Pannonia Inferior and the ﬁrst of which is similarly listed as ‘missa in expeditionem’ on a diploma issued for its ‘home’ province on l-ix-114 ; and the cohors Lepidiana equitata bis torquata c.R., from Moesia Inferior. Moreover, there are good reasons for supposing the cohors IIII Gallorum equitata recorded in Thrace on 19-vii—1 14 is the homonymous cohort recorded as the garrison of Cilicia in 121, and that it too was deployed to the east in connection with Traj an’s Parthian War (6).

The failure to note the evidence indicating the possible/probable participation of these six or seven ‘European’ regiments in Trajan’s Parthian War might seem to be a somewhat surprising and remarkable omission in an article that ostensi-bly discusses auxiliary d ployments during that campaign, no matter how mis-leading the title. Yet, the f the matter is that the relevant evidence is not so easily available or necessarily familiar to scholars whose main ﬁeld of interest generally excludes the Roman Army in Anatolia. This paper, therefore, seeks to rectify this omission by reporting this evidence in a much more accessible form as a prelude to a future re-analysis of Trajan’s Parthian campaign.

1) Cohors IIII Raetorum equitata

Evidence for the cohors IIII Raetorum equitata having been in Asia Minor at the time of Traj an’s Parthian War comes in both epigraphic and literary form. A sarcophagus from Side, to begin with, bears the following text (7) :

M(<’xguov) Oiﬁkmov / ’Aootowov / Zenbv y’ / M(ét9uog) Obxmog / AoWog

(émrovrd)ox(ng) /oneioag 6’ / ‘Perdw /f)tov éomtof) léveétoe /évé081:o

(5) ECK and PANGERL, 2005 [n. 2], p. 61.

(6) Another text that may be relevant here is CIL 3, 12257 = AE 1890, 11, a funerary inscription recording the presence in Asia Minor of a cohors Lusitanorum and using epi— graphic conventions suitable to the period we are concerned with. However, the lack of an identifying numeral for the unit precludes fruitful analysis.

(7) AE 1915, 49 = J. NOLLE, Side im Altertum : Geschichte und Zeugnisse II (= [K 44),

(3)

‘Marcus Ulpius Longus, éxatowaupng in the [III cohors Raetorum, placed his own son, Marcus Ulpius Arrianus, three years of age, in here.’

Theﬁnd spot of an epitaph erected by a serving soldier normally indicates that he and his unit were present at that place at the time the text was inscribed, and the phrasing used in the text seems to conﬁrm that Marcus Ulpius Longus was indeed serving with the rank of centurion in the cohors IIII Raetorum at the time he placed his deceased three-years old son in a locally-made sarcophagus. In which case it follows that Longus’ regiment was at Side in Lycia-Pamphylia at some point in its history, a supposition that gains support in the knowledge that Side was one of two places regularly used as a home base by the garrison of that province (8). That aside the inscription is of note in that both father and son have

the trio nomina of full Roman citizens, and as they share the same nomen then

the son was born in legal wedlock : therefore Longus was a Roman citizen who chose to serve in the nominally peregrine auxilia (9). Moreover, given the age of

Longus’ son at the time of his death we might reasonably assume that his wife

was with him at Side, centurions being exempt from the rule that forbade ordi—

nary Roman soldiers from legal marriage during their military service (1°). As for the date of this text, the way that it registers the full tria nomina of both son and the father points to it having been inscribed no later than the opening decades of the 2'“l century AD, as later on it became more usual to give nomen and cognomen alone (“). On the other hand, our centurion (and his son) bears the same praenomen and nomen as the emperor Trajan, and acquired imperial nomenclature of this type were usually taken by auxiliaryman discharged by the emperor of the same name (‘2). In which case it could be objected that not only

(8) J. BENNETT, The Auxilia ofLycia-Pamphylia .' Identity, Deployment and Function in C. DEROUX (ed), Studies in Latin Literature and Roman History XIV, Brussels, 2008,

p. 283-305 : 297-299.

(9) Cf. P. A. HOLDER, The Auxilia from Augustus to Trajan, Oxford, 1980, p. 86-90, where it is noted that most Roman citizens who entered the auxilia often ranked as cen— turions (as here) ; see also M. P. SPEIDEL, The Captor of Decebalus .' a New Inscription from Philippi in JRS 60, 1970, p. 142-153 (= IDEM, Roman Army Studies I, Amsterdam, 1984, p. 173-87) ; and B. PFERDEHIRT, Die Rolle des Milita'rsﬁir den sozialen Aufstieg in der Romischen Kaiserzeit, Mainz, 2002, p. 16. The potential advantages to a citizen of service in the auxilia as opposed to the legions or another citizen unit are discussed in B. DOBSON, Legionary Centurion or Equestrian Oﬁicer ? A Comparison of Pay and Prospects in Ancient Society 3, 1972, p. 193-208.

(10) J. B. CAMPBELL, The Marriage of Soldiers under the Empire in JRS 68, 1978, p. 153—166 : for other cases of serving soldiers in legal wedlock see items 3, 7 and 8 below.

(11) Such at least is the commonly held opinion, although there are naturally

excep-tions to prove the rule.

(12) Cf. K. KRAFT, Zur Rekruitierung der Alen und Kohorten an Rhein und Donau,

(4)

is this text quite possibly somewhat later in date, but we should also allow for the possibility that Longus was actually a veteran who chose to settle at Side for one or other reason, making his presence there of no relevance at all with regard to the whereabouts of the cohors IIII Raetorum at any stage in its history. An objec-tion of this kind on these grounds is perfectly valid : but ignores the fact that there are at least two and perhaps four other auxiliarymen on record with Trajan’s praenomen and nomen who adopted these names before their discharge from military service, having been given an advance grant of Roman citizenship for their meritorious service in the Dacian Wars (‘3). It could be that Longus received citizenship in a similar way, especially as the cohors IIII Raetorum is likely to have taken part in the same campaign (14). Then again, as it was not unknown for Roman citizens to join the auxilia instead of (or after) service in a legion (15), it could be that Longus was the son of an auxiliary veteran who took Trajan’s nomenclature after his discharge at a earlier date in Trajan’s reign, and that for some reason or other the son, Longus, chose a military service in the auxilia rather than the legions (‘6).

Whatever the explanation for Longus’ possession of Trajan’s praenomen and nomen, the history of his unit, the cohors IIII Raetorum, is reasonably clear and allows the possibility that it was in Asia Minor at the time of Trajan’s Parthian War. Initially raised from the Raeti, an ethnos based in the region around mod-ern Lake Constance, and absorbed into the Roman Empire in AD 15 (17), the tribe eventually supplied the Roman army with at least eight cohortes equitatae, only one of which (fortunately for us) had the numeral “[111” (13). The earliest known diploma recording the existence of this particular unit shows it was assigned to Moesia Superior sometime before 16—ix—94, and it stayed ‘on the books’ of that province up to May-August 115, when the new diploma of 115 reports it as ‘absent on campaign’ ('9). On the basis of the Side text we can propose that it was

(13) J. C. MANN, Name Forms ofRecipients ofDiplomas in ZPE 139, 2002, p. 227-234 : 230, quoting CIL 16, 160 and 163 (both issued in 110) ; and RMD 86 (issued 113) and 20 (issued 118-120). To these might be added RMD 231 (issued 118) and 371 (issued 129). Cf. also RMD 250 (issued 139), recording M. Ulpius Cabelus : he was recruited in 109 and may have received citizenship prior to discharge for service in Trajan’s Parthian War.

(14) K. STROBEL, Untersuchungen zu den Dakerkriegen Trajans, Bonn, 1984, p. 149-150.

(15) Cf. J. C. MANN, op. cit. [n. 13], p. 229, quoting CIL 16, 164, to which we can now add RMD 443 and 447. See also the references cited above in [n. 9].

(16) Most Roman citizens who entered the auxilia often did so after service in a regu-lar citizen unit, and often ranked as centurions (as here) : cf. P. A. HOLDER, op. cit. [n. 9], p. 86-90.

(17) D. B. SADDINGTON, The Development of the Roman Auxiliary Forces from Caesar to Vespasian (49 BC - AD 79), Harare, 1982, p. 153.

(18) J. SPAUL, Cohors 2, Oxford, 2000, p. 276—87.

(5)

sent to Lycia-Pamphylia to replace the single auxiliary unit garrisoned there after this moved further east for active service in Trajan’s Parthian campaign (2°). If so, then the III] Raetorum had certainly left Lycia—Pamphylia by c. 125, by which

year, at the latest, the cohors I Musulamiorum constituted the garrison of

Lycia-Pamphylia (2‘). In fact as the 1111 Raetorum is registered among the Cappadocian garrison for Arrian’s sortie against the Alans in c. 135 (22), then it had quite prob— ably left Lycia-Pamphylia for Cappadocia sometime before 117. After all, the late summer of that year saw Hadrian regularising the provincial garrisons of Asia Minor before leaving in the autumn for the Danubian provinces, taking with him a number of those units Trajan had transferred from there for the Parthian War (see below) : thus as the cohors IIII Raetorum was ‘left behind’, so to speak, then it had presumably already long been established as one of the standing gar-risons of Cappadocia. Either way it seems that the cohors IIII Raetorum remained in Cappadocia from that time onward, as the next (and also the last)

documented reference to the unit is c. 395, when it is listed among the regiments commanded by the Dux Armeniae (23).

2) Cohors VII Breucorum ciuium Romanorum equitata

A funerary text found at Gordion in 1996 records the burial there of a soldier serving in the cohors VII Bruecorum at the time of his death (2“).

[D(is) M(anibus)] / TRITO BATONI(s) (ﬁlio) / MIL(iti) COH(ortis) VII / BREVC(0-rum) C(ivium) R(omanoBREVC(0-rum) EQ(uitata) / DOM(0) PANN(onia) / ANN(is) XXXII ST(ipendiorum) XII / MERSVA DASI (ﬁlius) / VEXIL(larius) COH(ortis) EIVSD(em) / HER(es) POSVIT

‘To the Shades ! To Tritus, son of Bato, soldier of the cohors VII Breucorum ciuium Romanorum equitata, whose home was Pannonia, who lived for 32 years, served for 12 years. Mersua, son of Dasius, uexillarius of the same cohort, his heir, placed this (monument) here.’

(20) The unit the cohors IIII Raetorum replaced in Lycia-Pamphylia was probably the cohors I Apula : J. BENNE'IT, op. cit. [n. 8], p. 289—291.

(21) J. BENNETT, op. cit. [n. 8], p. 293—295.

(22) ARK, Ext. 1.

(23) ND 0R. 38.28, with its base listed as Analiba (?Diviri). It is not certain when this

Armenian command was ﬁrst created. It could have been c. 293-294, when Sophene and

Ingilene became Roman protectorates and were attached to the newly established province of Armenia Minor : R. C. BLOCKLEY, East Roman Foreign Policy : Formation and Conduct from Diocletian to Anastasius, Leeds, 1992, p. 5-6. However, it could be later, when another provincial adjustment saw Armenia Minor become Armenia Prima and the eastern part of Cappadocia detached to form Armenia Secunda, apparently before 371 : T. A. KOPECEK, Curial Displacements and Flight in Later Fourth Century Cappa-docia in Historia 23, 1974, p. 319-42 : 320-21. On balance, then, it seems likely the com-mand was created by Valens in c. 364.

(24) A. L. GOLDMAN, A New Military Inscription from Gordian in Anadolu Mede-niyetleri Mﬁzesi .' 1996 yllllgl, Ankara, 1997, p. 45-57.

(6)

Although the name of the deceased is given in the dative, a practice more usual from the early to mid—2"" century onward, the new diploma for Moesia Superior lists his unit as ‘absent on campaign’ from that province in May-August 115, allowing us to assign the Gordion text to the period of Trajan’s Parthian War. In which case it is of some interest to see how evidently indigenous names predominate in the text : although Mersua is at present unparalleled, the others — Tritus, Bato and Dasius — are all of Balkan origin (25). Similarly, it is noteworthy how the text reveals that even as late as the Trajanic period there were long-serv-ing auxiliary soldiers who chose to maintain their indigenous name rather than adopt a Latin or at least Romanised ‘nom de guerre’ to suit their profession and status, a practice already common by the mid-15‘ century AD (26).

The cohors VII Breucorum itself was one of eight consecutively numbered units recruited from among the Pannonian Breuci (27). It is ﬁrst recorded in the garrison of Germania on 17—Vi—65, and appears next in Pannonia on 5-ix-85 (2‘). By 8-v-100 it was in Moesia Superior (29), and the new diploma of 115 for that province shows it remained ‘on the books’ after being transferred-out for service in association with Trajan’s Parthian campaign. The Gordion inscription sug-gests the unit spent at least a part of its time at that place while ‘on campaign’, perhaps serving the role of a local requisitions unit (3°), but by 7-Viii— 143, a

diplo-ma shows it had been re-assigned to Pannonia Inferior, where it rediplo-mained at least

until 31-viii-203 (3‘). That said, a dedicatory inscription found at Knodara, Cyprus, testiﬁes to the unit’s presence there at some point in its history (32) : this

(25) GOLDMAN, op cit, p. 47-48.

(26) Cf. RIB 121 and 201 ; also CIL 13, 8312. Note also those 1“ century and later funerary texts that record men with both indigenous and alternate Latin names : e.g., ILS

2901, ‘C. RAVONIUS CELER QUI ET BATO SCENONARBI NATION DAL[m]’.

(27) SPAUL, op. cit. [n. 18], p. 317—327.

(28) Germania : RMD 79 ; Pannonia : CIL 16, 31.

(29) CIL 16, 46.

(30) J. BENNETT and A. L. GOLDMAN, Roman Military Occupation at Yassiho'yu‘k (Gordian), Ankara Province, Turkey’ in Antiquity Project Gallery, December 2007 : hgp://antiguity.ac.uk/ProjGall/bennett/ index.html. What that paper does not make entirely clear is just how many of the early ‘Roman’ structures at Gordion conform to Roman mil-itary norms : cf. IDEM, A Preliminary Report on the Roman Milmil-itary Presence at Gordian, Galatia in Gladius 13 (2), 2009 (= Articulos del 20 Congreso Internacional de la Frontera Romana), p. 35-45. More to the point, the failure to continue excavations on what is the only securely identiﬁed Roman fort in Anatolia is extremely lamentable.

(31) RMD 266 (for 143), 397 (for 144), and 401 (for 146) ; CIL 16, 179 and 180 (for 148) ; (?)RMD 415 (for 154/156), 102 and 103 (for 157) ; CIL 16, 112 and 113 (for 159), and 123 (for 167) ; RMD 448 (for 157/ 192), 446 and 447 (for 192), and 187 (for 203).

(32) T. B. MITFORD, New Inscriptions from Roman Cyprus in Opuscula Archaeologica 6, 1950, p. 1—95 : 55-56, revising CIL 3, 215 after personal inspection to read : ‘GENIO /

(7)

was most probably c. 116, when the Jewish community of Cyprus led by one Artemion rose up against Rome and massacred some (allegedly) 240,000

non-Jews on the island, an event that would have likely necessitated the dispatch of

additional Roman auxilia to assist the island’s standing garrison in restoring law and order (33).

3) Cohors I Thracum Syriaca equitata

The presence of the cohors I Thracum Syriaca equitata in Asia Minor at about the time of Trajan’s Parthian War is indicated by the text on a decorated funerary

monument found at Horopol, near Cengerli, some 15 km. southeast of

Refa-hiye (3“).

D(is) M(anibus) / IVL(ius) SEXTILIVS / LONGINVS / VIX(it) AN(nis) IIII / H(ic) S(itus) E(st) / M(arcus) SEXTILIVS / VALENS / DEC(urio) COH(ortis) I / THRAC(um) SYR(iacae) FIL(io) / SVO POSVIT

‘To the Shades ! Julius Sextilius Longinus, who lived 4 years : he lies here. Marcus Sextilius Valens, decurion of the cohors I Thracum Syriaca, placed this (monument) here for his son.’

The combination of the formulae ‘D M’ in the ﬁrst line and ‘H S E’ in the fourth point to this text dating to AD 75—125, and three other epigraphic features in the text are entirely consistent with such a time-span : the use of the nomina-tive case for the deceased ; the registering of the whole tria nomina of son and

father ; and the fact that their nomen is spelt out in full. Furthermore, as with the ﬁrst inscription discussed in this article, that from Side, as father and son share the same nomen, then the son was born in legal wedlock, and as both have tria nomina, both were Roman citizens : Valens, therefore, was evidently yet

anoth-er Roman citizen who chose to sanoth-erve in the nominally panoth-eregrine auxilia (35). Likewise, from his son’s age at the time of death we might assume that he was also yet another junior ofﬁcer in the auxilia who was both legally married and had his wife with him while in military service (3‘).

O_G_[V]L_N[I]VS TIR[O] E/AEF C [A P] AEmmT. See E.BIRLEY, The Religion of the Roman Army : 1895-1977 in ANRW 16.2, 1978, p. 1506-1541 2 1510, for how

simi-larly phrased dedications to Jupiter Optimus Maximus indicate the presence of the named unit at full strength at the place where the text was found.

(33) D10 68, 32, 3.

(34) The discovery of this epitaph and the name of the unit thereon were announced in T. B. MITFORD, The Roman Frontier on the Upper Euphrates in R. MATTHEWS (ed), Ancient Anatolia, Oxford, 0. 1999, p. 255-272 : 267 ; it will be fully discussed by the same author

in a future publication.

(35) Cf. [n. 9].

(8)

(

The I Thracum Syriaca equitata was one of four auxiliary cohorts originally recruited in Thrace and sharing the agnomen Syriaca (37). It is generally believed that when the adjectival form of a provincial name is found as part of a unit’s ofﬁcial title then the unit concerned spent a lengthy period of service in the same province and assumed its locative sufﬁx when on campaign elsewhere in order to distinguish it from other units with the same number and ethnic label (38) : in other words in the sense here of ‘the Thracian cohorts m Syria’. Conversely, it may be that the agnomen was initially bestowed on all four of these particular Thracian units when they were initially posted to Syria, that is, meaning ‘the

Thracian regiments i_n. Syria’. This seems possible as the other three units with

the same ethnic and agnomen continued to use the identifying locative even

while in Syria itself, their ‘home’ province (3").

Be that as it may, the unit is recorded on an inscription from Timacum Minus, Moesia (Ravna, Serbia) that can be plausibly dated to around 69/70, while four epitaphs naming regimental veterans in the nominative are known from the same place (4°). Evidently the cohors I Thracum Syriaca left Syria in July-August 69

with the expeditionary force taken to Europe by L.Mucianus in support of

Vespasian’s claim to the imperial purple (4'). In fact the unit is listed among the ’garrison of Moesia on a diploma of 7-ii-79 (42), and it was likely still at Timacum Minus in 86, for after the division of Moesia in that year, when the place become a part of Moesia Superior, the unit is next documented on a diploma on 8-v—100 for that same province (‘3). It quite probably took part in Trajan’s Dacian Wars (4"), but if so the new Moesia Superior diploma for May-August 115 reveals

(37) Although unit sufﬁxes of the locative or honoriﬁc type are often termed cognom-ina, they are surely agnomina in the sense of being descriptive additions to a unit’s for-mal nomenclature.

(38) Cf. L. CHEESMAN, The Auxilia of the Imperial Roman Army, Oxford, 1914, p. 47, and 62, n. 6 ', also D. L. KENNEDY, The Ala I and the Cohors I Britannica in Britannia 8, 1977, p. 249-55 : 249-50 and 255.

(39) The [I Thracum Syriaca : RMD 4 (for 91) and CIL 16, 106 (for 156/157) ; the III Thracum Syriaca 2 CIL 16, 35 (for 88), RMD 214 (for 91) and 372 (for 129) ', and the [III Thracum Syriaca : B. PFERDERHIRT, Ro'mische Militiirdiplome und Entlassungsurkunden in der Sammlung des Ro'misch-Germanischen Zentralmuseums, Mainz, 2004, p. 18—17, no. 6, esp. 17, n. 4 (for 91, but noting another unpublished example for 93).

(40) CIL 3, 8261 = ILS 2733 = PME 5, 59 (of c. 69/70) ; CIL 3, 8262, 14575 and 14579 (= AE 1901, 17 : also a veteran, given the subject’s age of 61 and 28 years service) ; and

AE 1910, 95.

(41) TAC., Hist. 2, 83.

(42) CIL 16, 22 and RMD 208. Note, however, a diploma for Moesia in 75, which may have named the unit : PFERDEHIRT, op. cit. [n. 39], p. 3-8, no. 1.

(43) CIL 16, 46.

(44) STROBEL, op. cit. [n. 14], p. 144 : but note that it is not named on either of the two Dacian diplomata issued in 110 : CIL 16, 57 and 163.

(9)

it had returned ‘home’ prior to being transferred for extra-provincial service dur-ing Trajan’s Parthian War. At that time it seems to have been commanded by M.Sextius Proculus: his cursus honomm shows he was praefectus of the I Thracum Syriaca when it was brigaded with vexillations from the I Cilicum and VII Breucorum, two of the other units named on the new diploma as having been absent from Moesia Superior in 115 (“5).

Following its service ‘abroad’ under Trajan, the I Thracum Syriaca returned to Europe sometime before 125 when it was listed among the garrison of Moesia Inferior (45). It most likely arrived in this province in the early winter of 117 along with the ‘sacred armies’ that Hadrian led from Parthia via Asia Minor to Moesia in response to the attacks then being made on that region by the Roxolani (47) : this force would naturally have included those Danubian units previously on active campaign in the east and which were now needed to rein-force the garrison in those provinces threatened by the Roxolani. Either way, the I Thracum Syriaca then remained in Moesia Inferior until at least 157 (48), an altar set up by its praefectus at Transmarisca (Tutrakan, Bulgaria) suggest-ing it was based tﬁno later than the Antonine period (49).

Although the Horopol inscription places the cohors I Thracum in Armenia Minor at some point in its history, the chronological features in the text and the unit’s history indicating that this was probably at the time of Trajan’s Parthian War, there is no other evidence for any Roman-period activity in the immediate vicinity of its ﬁnd spot (5"). Thus it may be that this stone was transported there in post—Roman times from somewhere else, and if so, then perhaps from Melik Sérif, 10 km. to the north, and the approximate location of the Chorsabia of Ptolemy’s Geography and the Carsagis of the Antonine Itinerary, a place located on one of the main east-west routes across Pontic Anatolia (51). More to the point,

(45) AE 1926, 150, for M. Sextius Proculus (PME $.25). The inscription was previ-ously considered Hadrianic in date : cf. STROBEL, op. cit. [n. 14], p. 144, with n. 315.

(46) RMD 235 (cf. RMD 364).

(47) SHA, Had. 6, 6. Note IGRR 3, 208, from Ancyra, recording ‘the transit of the emperor Hadrian with his sacred armies’ : this must date to the late autumn or early

winter of 117, as IGRR 4, 349 shows that by ll-xi-117, Hadrian and his entourage had

reached Juliopolis/Gordioukome (Eskisehir), 140 km to the west on the road to Byzan-tium via Nicaea and Nicomedia.

(48) RMD 241 (for 127) ; 165+399 (for 145) ; 270 (for 146) ; 414 (for c.155) ; and 50

(for 157).

(49) AE1939, 101.

(50) Pers. comm., T. B. Mitford.

(51) Chorsabia : PTOL., Geog. 5, 6, 20 ; Carsagis : Ant. Itin. 208 and 215. That suitable stones could travel some distance for re-use in another setting is vividly demonstrated in England by those inscriptions transported in Saxon and medieval times from Corbn'dge to Hexham, 7 km away : cf. RIB 1120, 1122, etc.

(10)

Melik Serif has produced a milestone of 92/94 (reused in Hadrianictimes), sug— gesting that this route (if not necessarily this explicit location) was already of some importance when Domitian re-organised the Euphrates frontier (52). Fur— thermore, the same place has also produced a building inscription of c. 199 recording work by the Cohors I Lepidiana equitata bis torquata CR. (53). It is highly likely, therefore, that Chorsabia/Carsagis or its general vicinity had a recognised tactical if not strategic value from an early date, and as such could be the location for a garrison in the Trajanic period.

4) Cohors I Montanorum (equitata ?)

The presence of a cohors I Montanorum in Phrygia at the time of Trajan’s Parthian War is suggested by a funerary text seen in 1924 at Gorgoromeis (Ak Kilisse) in Isauria (5").

[---] / KovéLah/og Ma]/§Lurp 11[L'q3] / éoroaravodumv év] / <I>gv(yta) e’Lg xtbg‘tmv) a M[--- rt]/e§(‘>g um innehﬁg] / 0LVY<lL>dQLOg aorta] / Bav(v) e’Lg tn[v udr(m)]

/ MUOLa(v) dig duhmto‘I] / owy<k>dQLo<g> um é [v e’L’Ml] / ’Attemépwv s’L[g owy Mid/own; (!) ma Mﬁgdo(tog) [BevcEaéLQLOQl /dXTdQ(LO§) xogvmmvLog --- ] / [---l

(?This monument was erected by NN?) Kondianos to his son Maximus : “I was in mil-itary service in Phrygia in the Cohors I M[. . .] as an infantryman ; then (I became) an eques singularis ; (then) going down to Lower Mysia (= Moesia Inferior) as (7eques) singularis; then (I was serving) in the ala Atectorigiana in the singulari; then a librarius, (then) (?)beneﬁciarius, (then) actarius, (then) comicularius ...”

Aside from this text lacking any epigraphic features of clear chronological value (55), it will be seen that several of the restorations and expansions are exem-pli gratia, and thus of uncertain value. Even so, there is general agreement on six

(52) T. B. MITFORD, Cappadocia and Armenia Minor: Historical Setting of the Limes in ANRW 2, 7, 2. 1980, p. 1169-1228 : 1185. For the Flavian frontier arrangements in this region, cf. J. BENNETT, The Origins and Early History of the Pontic-Cappadocian Frontier

in Anatolian Studies 56, 2006, p. 77-93 : 89-90.

(53) AE’ 1908, 22.

(54) W. H. BUCKLER, W. M. CALDER, and C. W. M Cox, Asia Minor, 1924.1. Monu-ments from Iconium, Lycaonia and Isauria in JRS 14, 1924, p. 24-84. 74-75, no 1090, with Plate XVII, 109C (= AE 1926, 74); but see now G. LAMINGER-PASCHER, Zu zwei kleinasialischen Militiirinschriﬂen 1n Wiener Stud. 7, 1973, p. 249—63: 257-63; also M. P. SPEIDEL, Guards of the Roman Armies, Bonn, 1978, 101—102, no. 57.

(55) Unless, that is, the naming of the ala Atectorigiana without the qualifying ethnic Gallorum has a chronological signiﬁcance. This is also missing from the unit’s name on a diploma of 105 (PFERDEHIRT, 2004 [n. 39], p. 30-32, no. 10), but it is included on the next recording the unit, issued in 122 (PEERDEHIRT, 2004 [n. 36], p. 57-61, no. 20), and on all subsequent issues (RMD 241, for 127 ', 270, for 146 ; and 50, for 152/ 154—157). However, while this allows the possibility that the ethnic was only adopted between 105-122, it is also missing from an early 3rd century text naming the regiment (J. F112, Honoriﬁc Titles

(11)

aspects of this text. First, that Kondianos erected the monument in memory of his

son Maximus. Second, the main part of the text describes Maximus’ career in the

form of a autobiographical oration, perhaps based on a letter or his testament.

Third, Maximus enlisted as an infantryman in a cohors I M[—-] at a time when

that unit was in Phrygia. Fourth, Maximus was subsequently appointed as an eques singularis in the guard unit of the provincial governor, his rank suggesting that he had previously graduated as a cavalryman in his parent regiment and that this was therefore an equitate unit. Fifth, he travelled with his parent regiment to Moesia Inferior, for an eques singularis remained formally attached to his unit throughout his term as a governor’s guardsman (5”). Sixth, after arriving in Moesia Inferior, Maximus was re-appointed as (eques ?) singularis before trans-ferring to the ala Atectorigiana, one of the ‘permanent’ units stationed in that province (see below), and in which he served first as a singularis for his com— manding officer, before being made a beneficiarius (probably), then actarius and finally cornicularius.

The main problems here, of course (insofar as far as this paper is concerned), are those of establishing the date of the text and identifying the cohort Maximus joined while it was in Phrygia. To deal with the second of these issues ﬁrst, the unit could in theory be almost any one of at least nine cohortes now known to

exist with the numeral ‘1’ and the initial letter ‘M’ : I Mattiacorum ; I Mauro-rum ; I MenapioMauro-rum ; three (or more) I MontanoMauro-rum (57) ; I MorinoMauro-rum ; I

Musu-lamiorum ', and I (Flauia) MusuMusu-lamiorum. However, the I Mattiacorum disap-pears from the record after a single appearance on a Moesian diploma of 78 ; the I Menapiorum and the I Morinorum seem to have been permanently based in

Britannia; the I Flauia Musulamiorum never seems to have left Mauretania Caesariensis; and the I Musulamiorum remained in Syria until at least 113

before being transferred to Lycia-Pamphylia between c. 118-125 (58).

of Roman Military Units in the 3rd Century, Bonn, 1983, p. 66, no. 244). In which case its omission on the text discussed here might be of no chronological signiﬁcance whatso— ever

(56) SPEIDEL, 1978 [n. 54], p. 7.

(57) The existence of two such units is reported as ‘I at I Montanorum’ on a di lama

for Pannonia of 85 (CIL 16, 31), one probably equitate, the other peditate (cf. : J. ASEL,

Cohors I Montanorum in C. UNZ, Studien zu den Militiirgrenzen Roms XIII, Stuttgart, 1986, p. 782-786 : 765). Analysis of the available diplomata suggests one remained in Pannonia and then later in Pannonia Inferior, while the other was deployed in Moesia Superior before 96. However, a third I Montanorum is recorded in Syria-Palestina in 136/137 (RMD 160), and again in 142 (PFERDEHIRT, op. cit. [n. 39], p. 83-85, no. 29), and in 160 (RMD 173).

(58) For the I Mattiacorum : RMD 208 ; for the I Menapiorum and I Morinorum : SPAUL, op. cit. [n. 18], 185 and 186; for the I Flavia Musulamiorum: SPAUL, op. cit. [n. 18], p. 47 ; and for the I Musulamiorum : SPAUL, op. cit. [n. 18], p. 472, with BENNETT, op. cit. [n. 8], p. 293-95.

(12)

Thus as far as the evidence goes, it is probable that the cohors I M[---] our Maximus joined was one or other of the cohortes I Montanorum (59) 2 and this, of course, is where the new Moesian Superior diploma becomes relevant, in con— ﬁrming that the province’s cohors I Montanorum was absent ‘on campaign’ in 115 (5°). The unit itself is ﬁrst recorded there in 96 and again in 103/106, and geo— graphical proximity suggests it was probably the same unit with this name recorded in Dacia in 109, 110, and again on 3/4—v-1 14 (6'). If so, the new diplo-ma for Moesia Superior shows that it was re-assigned to its ‘home’ province before August 115 at the very latest as by then it was absent, ‘translatis in expe-diti[one]’. Other diplomata demonstrate that it returned ‘home’ sometime before 9-ix-132, and is subsequently attested in Moesia Superior in 157, 160 and 161 (62). True, this particular regiment is nowhere registered as being equitata, the type of unit that Maximus seems to have joined in Phrygia, while in any case it might be thought unlikely that units raised from among the Montani would combine cavalry with mountain infantry. As it is, however, aside from the simple fact that such units could have been elevated to equitate status subsequent to their original formation, the cohors I Montanorum stationed in the province of Pannonia Inferior in 192 or 193 was certainly an equitate unit, a fact known only from the diploma issued to one of its former members in that year and in which the unit is described as such (63).

The available data thus allows for the following scenario. In 114/115, the Upper Moesian cohors I Montanorum was assigned to Phrygia for support duties in connection with Trajan’s Parthian War (6“). It was there when our Maximus

(59) LAMINGER-PASCHER, 1973 [n. 54], p. 261, came to the same conclusion, albeit by a different route.

(60) It needs to be noted that the editors of the new Moesian Superior diploma for 115 also accepted the possibility that the I Montanorum was named on this text, but rejected its potential value with regard to the document they were concerned with on the grounds that being an epitaph erected in the soldier’s own home, then it had no bearing on whether or not the unit had served in the east (cf. ECK and PANGERL, 2005 [n. 2], p. 59, n. 40) : they were apparently unaware of the full contents of the text and its reference to the unit Maximus joined being in Phrygia at the time he was recruited.

(61) Moesia Superior : RMD 6 (for 96), and PFERDEHIRT, op. cit. [n. 39], p. 38, no. 13 (for 103/106) ;Dacia : RMD 148 (for 109) ; CIL 16, 63 (for 110) ; and RMD 226, for 114. Note that the homonymous unit in Pannonia Inferior is recorded as being in its ‘home’ province on 1—ix-1 14 (RMD 344/ 152+228).

(62) RMD 247 (for 132) and 419 (for 157) ; PFERDEHIRT, op. cit. [n. 39], p. 117-118, no. 40 (for 160) ; and RMD 55 (for 161).

(63) PFERDEHIRT, op. cit. [n. 39], p. 126-131, no. 44 : remarkably enough the designa— tion equitata for this unit appears on the extrinsecus only.

(64) That this hypothesised transfer could only have taken place after the Parthian War began does not preclude it happening in connection with that campaign : the course of events in the east between 114—117 (cf. BENNETT, op. cit. [n. 3], p. 196-202) reveals that

(13)

enlisted in the cohort, his later appointment as an eques singularis suggesting it may well have been equitate, but in 117 the unit left Phrygia for Moesia Inferior — presumably with Hadrian’s ‘sacred armies’ in response to the increasing threat

from the Roxolani (65). When that matter was resolved, the I Montanorum

returned to its ‘home’ province of Moesia Superior, but by then Maximus had

been promoted to the ala Atectorigiana, a unit that had been based in Moesia

Inferior from at least 105 until 154/157, apart from a period of service in Dacia Inferior (“5). Unfortunately, despite the allure of this possible scenario it is con— ceded that we cannot know if the cohort Maximus joined was indeed the I Montanorum, aside from which there is nothing in the text to certify that it belongs to the relevant period. As such, then, all that can be proposed is that the inscription allows for the possibility that the I Montanorum was in Phrygia at the time of Trajan’s Parthian War.

5) Claudia Nova (miscellanea)

Another unit from Moesia Superior that appears to have been in Asia Minor at the time of Trajan’s Parthian War was the ala Claudia Nova, whose presence in the region is shown by a funerary text from Amaseia, modern Amasya (67).

L SEMPRONIVS L(ucii) [F(ilivs)] / SCA[P(tia)] ALTI[NI] M[A]/CEDO DEC(uri0) ALAE / CLAVDIA(e) NOVA(e)

‘Lucius Sempronius Macedo, son of Lucius, of the Scaptia voting tribe ; of Altinum ; decurion in the ala Claudia Nova’

The use of the nominative to describe the deceased in this text and the way it spells out our soldier’s nomen in full suggest that it belongs to the period before ' AD 125. A similar date is likewise indicated from our soldier’s entire ofﬁcial nomenclature as a Roman citizen being presented as praenomen, nomen, ﬁlia— tion, tribus, origo, and cognomen. Our soldier, therefore, provides us with yet another example of how a Roman citizen might prefer military service in the

Trajan seriously underestimated the forces he needed for his Parthian campaign (as he also did with his First Dacian War), and as will be shown below, at least one other unit seems to have been sent from Europe to the east in the course of 114/115, namely the cohors IIII Gallorum equitata.

(65) Two and possibly three of the Moesian Superior units recorded as ‘translatis in expediti[0ne]’ on the new diploma were certainly subsequently assigned to Moesia Inferior, albeit on a permanent basis, having presumably been sent there by Hadrian : cf. ECK and PANGERL, 2005 [n. 2], p. 62-63.

(66) Moesia Inferior : PFERDEHIRT, op. cit. [n. 39], p. 30-32, no. 10 (for 105) ; and RMD

(for 113), 341 (for 127) ; 365 (for 138/142), 165/399 (for 145), 270 (for 146), and 50 (for

152/ 154/157). Dacia : PFERDEHIRT, op. cit. [n. 39], p. 57—61, no. 20 (for 122).

(67) CIL 3, 13635 = J. G. C. ANDERSON, F. CUMONT, and H. GREGOIRE (edd), Studio Pontica 3, Paris, 1910, p. 131, no. 105.

(14)

auxilia instead of (or perhaps after) volunteering for the legions or another citi-zen unit (63).

As its name indicates, the ala Claudia Noua was inducted by a Julio-Claudian

emperor, most probably Claudius himself, as epitaphs for serving members and

veterans of the unit show it was assigned to the garrison of Dalmatia from about

AD 45 (69). However, a diploma of 74 establishes it was then in Germania (7°),

perhaps having arrived there via Italy when Otho and Vitellius were warring

against each other (7‘). It remained in Germania until shortly before 20-ix-82,

when another diploma for the same military region records that it had recently been transferred to Moesia (72). Then it appears on a diploma for Moesia Superior of 93, and again for that same province in 94 and 100 (73). It evidently took part

in Trajan’s Dacian Wars, however, for a commander at that time was decorated for service ‘bello Daccico’ (7“), and a diploma for 110 shows that it was then

assigned to the Dacian garrison (75). However, it is perhaps referenced on a diplo-ma for Moesia Superior of 132, and it certainly appears on a diplodiplo-ma for the same province on 20-i-151, where it appears for theﬁrst time with the agnomen miscellanea. The unit is also named on diplomata for the same province issued

in 157, 159/160, 160, and ﬁnally in 161, in each case with the like agnomen (7").

Amaseia is best known for being the main polis of Pontus and the home-town of Strabo. Its location on the northern highway directly linking Byzantion with Satala, trejﬁmnkt for Traj an’s advance into Armenia and the capture of Artaxata in the ﬁrst year of his Parthian War, would account for the discovery there of a

(68) Cf. [n. 9].

(69) G. ALFOLDY, Ro'mische Heeresgeschichte, Amsterdam, 1987, p. 242-243.

(70) CIL 16, 20.

(71) ALFOLDY, op. cit. [n. 69], p. 242.

(72) CIL 16, 28.

(73) CIL 16, 39 (for 93) ; RMD 335 (for 94) ; and CIL 16, 46 (for 100).

(74) AE 1972, 573 for M. Gavius Bassus (PME G8); also V. A. MAXFIELD, The Military Decorations of the Roman Army, London, 1981, p. 170-171 ; D MERKELBACH, H. ENGELMANN and D. KNIBBE (edd), Die Inschriﬁen von Ephesos 111 (= [K 13), Bonn,

1980, p. 76-77, no. 680 : and PLINY, Ep. 10, 21, and 86a.

(75) CIL 16, 163 (stamped bricks with the abbreviated title A CL(audia) recorded from Slaveni, Romania, suggest it was based there for at least some time: N. GUDEA, Der Dakische Limes .' Materialien zu seiner geschichte in RGZM 44, 1997, p. 497-610

(*1-*113) : 580-582 (*83-*85), Abb. 69, Z.1).

(76) RMD 247 (for 132) ; PFERDEHIRT, op. cit. [n. 39], p. 89-91, no. 31 (for 151) ;RMD

418 and (?)419 (for 157) ; CIL 16.11 (for 159/160) ; PFERDEHIRT, op. cit. [n. 39], p.

117-118, no. 40 (for 160) ; and RMD 55 (for 161). The origin of the agnomen is uncertain, alt— hough some have suggested it derives from its original founding complement of Spanish and Gaulish recruits (cf. ALFOLDY, op. cit. [n. 69], p. 242, with his note 28) : however, this seems unlikely as the title does not appear before 151. Alternatively, it could indicate a major inﬂux of recruits from a variety of sources to make up for losses while on campaign shortly before that date.

(15)

funerary monument to a serving member of a unit directly involved in that cam-paign : he could have died while either on his way to the front in c. 113 or dur—

ing the unit’s return west with Hadrian in 117. That being so, then the

chrono-logical features in the text together with the record of the diplomata all point to the probability that the ala Claudia Nova played a part in Trajan’s Parthian War. 6) Ala I Flavia Augusta Britannica milliaria c.R.

Evidence for the ala I Flavia Augusta Britannica milliaria c.R. having been in Asia Minor at some point in its history is given by a religious dedication found at Amaseia (77).

[I(oui)] 0(ptimo) M(aximo) / [AV]G(usto) (7“) SAC(rum)] / [AL]A I FLAVIA / [AV]G(usta) BRITAN[N(ica) / [ ><] C(iuium) R(omanorum) BIS TO[R]/[Q]VATA OB [V]I[RTV]/TEM CVI P[RAEEST] /[--]VS BON[--]

‘Sacred to Jupiter Optimus Maximus Augustus. The ala I Flavia Augusta Britannica milliaria, ciuium Romanorum, honoured with two torques because of its courage; commanded by ---us Bon-—-’

The text lacks any epigraphic features to indicate a date although the unit’s battle honours were probably acquired during Trajan’s Dacian Wars (see below). Otherwise this altar conforms to a familiar pattern in which the unit’s title is given in the nominative, followed by the phrase cui praeest to introduce the name of its commander at the time the altar consecrated. As such it falls into a class of dedications that auxiliary units are thought to have made to Jupiter Optimus Maximus either when renewing their annual oath of allegiance or on the birth— day of the reigning emperor (79). In other words, this altar provides clear proof that the ala I Flavia Augusta Britannica was at Amaseia in full strength at the time this inscription was set up.

The ﬁrst reference on a diploma to the ala Flavia Augusta Britannica mil-liaria c.R. is on that for Pannonia issued on 19-xi-102, its adjectival agnomen indicating earlier service in Britannia (3°). It evidently played a part in Trajan’s Dacian Wars as an inscription records that its commander at the time was pre-sented with military honours by that emperor for service ‘bello Dacico’ : given that the unit’s own battle honOurs are ﬁrst recorded on the same text, they too

(77) CIL 3, 6748 = ANDERSON, CUMONT, and GREGOIRE, op. cit. [n. 67], p. 130—131, no. 104

(78) Dedications to Jupiter Optimus Maximus Augustus would appear to be of some rarity: cf. CIL 3, 5167 (Celeia, Noricum), and 6423 (Lissa, Dalmatia); CIL 5, 4014 (Arilica, Regio X) ; CIL 8, 1327 = 14875 (Chidibbia, Africa Proconsularis) ; and CIL 12, 2410 (Vicus Augusti, Gallia Narbonensis).

(79) BIRLEY, op. cit. [n. 32], p. 510. (80) Cf. KENNEDY, op. cit. [n. 38].

(16)

were most likely won during that same campaign (8‘). It appears next on a diplo-ma for Pannonia Inferior issued on 2—viii-1 10 (82), but diplodiplo-mata issued for the same province on 1-ix-114 report it as being ‘missa in expeditionem’ (83). Another diploma of 135 shows it was back ‘home’ by then, and it is thereafter attested in Pannonia Inferior on a regular basis into the late 2nd century (3“).

As the ala I Flavia Augusta Britannica is listed as absent ‘on campaign’ from its ‘home’ province of Pannonia Inferior in 114, it seems logical to assume that the Amaseia text indicates its presence in Cappadocian Pontus at the time of Trajan’s Parthian War: as we have already seen, Amaseia is on a route likely used by Trajan’s army at the start of his Parthian campaign and also by Hadrian at its end. However, this same Byzantion-Satala route was quite probably also

used at the beginning and the end of Verus’ Parthian War, for Verus, like Trajan,

began his campaign in 163 with an advance along the Upper Euphrates into

Armenia Major (85). Moreover, Verus used at least one legion from Pannonia

Inferior in his campaign (the II Adiutrix), and thus (presumably) also auxiliaries from the same province. In which case the lack of any clear chronological indi-cators in the Amaseia text allows the possibility that it could well relate to this

later enterprise (86).

7) Cohors Campanorum (Voluntariorum c.R.)

A funerary text of unknown provenance but of locally available stone in the

museum at Amastra, ancient Amastris, and presumably found there or in the

vicinity, conﬁrms the presence in Asia Minor of the Cohors Campanorum, a unit otherwise reported only in the Balkans (“7).

L(ucio) SENPRONIO MILITI ET TV/BICINI COHORTIS CAMP(anorum) > L(uci) ALLI/DI MAGNA TITI ATTI FILIA VXOR / EIVS

(81) AE 1980, 496 for P. Cassius Secundus (PME C97 ter); cf. MAXFIELD, op. cit.

[n. 74], p. 171-173, and 223.

(82) CIL 16, 164 (for 110 : the ala I Britannica c.R. on a diploma issued for Dacia the same day (CIL 16, 163) is another unit).

(83) CIL 16, 61, and RMD 87 (name restored) and 152/228/345 (all for 114).

(84) RMD 251 (for 135) ; CIL 16, 175 (for 139) ; RMD 266 (for 143), 401 (for 146) ; CIL 16, 179 and 180 (for 148), 99 (for 150); RMD 102 and 103 (for 157), 415 (for 154/156) ; CIL 16, 112 and 113 (for 151/160) ;RMD 110 (for 154/161) ; CIL 16, 123 (for 167) ;RMD 446 and 447 (for 192).

(85) A. R. BIRLEY, Marcus Aurelius, a Biography, London, 1987, p. 128-129, with fur-ther references.

(86) The text is unlikely to date to the time of Severus’ eastern campaigns as inscrip-tions and the literary record show he crossed Anatolia by way of Ancyra and Cilicia.

(87) AE" 1993, 1429 = 1995, 1425 = C. MAREK, Katalog der Inschriften im Museum von Amasra in Epig. Anat. 6, 1985, p. 133-156 : 140, no. 19 = IDEM, Stadt, Ara und Territo-rium in Pontus-Bithynia und Nard-Galatia, Tubingen, 1993, p. 99-100, with 171, no. 52.

(17)

‘To Lucius Senpronius (”3), miles and tubicen of the cohors Campanorum, of the cen-tury of Lucius Allidius. His wife, Magna, daughter of Titus Attius (set this up for him)’.

Although the name of the deceased is given in the dative, two features of this particular inscription point to a date around the turn of the 15‘ century AD. First, the nomina of both our subject and his centurion are spelt out in full, an

epi-graphic trait usually indicative of the early rather than mid or later principate.

Second, while our soldier and his centurion are clearly ciues Romani neither have cognomina, as was almost universally the case among free—born Roman cit-izens by the end of the 1“ century (89).

The cohors I Campanorum Voluntariorum c.R. was originally raised by Augustus using citizen volunteers from Campania, and funerary texts and other inscriptions show it in Dalmatia for at least a part of the 1“ century AD (90).

However, a cursus honorum from Rome reports the regiment in Pannonia

Inferior by 116 at the latest (91) : it was presumably transferred there to replace a

unit from that province which had moved east in connection with Trajan’s

Dacian Wars. It is next documented on a diploma for the same province on 7-viii-143, which parenthetically reveals it had been receiving peregrine recruits from

at least 118, and is then listed in Pannonia Inferior on a regular basis until

192 (92).

Although the exact provenance of the text we are concerned with here is unknown, the port-town of Amastris is as likely a place as any other along the Pontic coast to produce the tombstone for a serving soldier from a unit otherwise only epigraphically recorded in the European provinces. Moreover, the stone used for this epitaph appears to be of a type that was available locally in the

(88) The nomen Senpronius would seem to be of some rarity, but cf. the Senpro(nius) or Senpro(nianus) of the legio XI Claudia on a Severan inscription from Phrygia: cf. M. CHRISTOL and T. DREW-BEAR, Inscriptions militaires d’Aulutrene et d’Apamée de Phrygie in Y. LE BOl-IEC (ed), La hie’rarchie (Rangordnung) de l’armée romaine sous le Ham-Empire. Actes du Congrés de Lyon, 15-18 septembre I 994, Paris, 1995, p. 57—92 : 70. For what it is worth, of the other seven recorded cases, three are of Iberian origin : CIL

2.2972 (Arroniz) ;AE’ 1987, 526 (Baedro) ; and AE’ 1946, 3 (Tarragona).

(89) CIL 16, 39 and 42, issued in 94 and 98 respectively to Roman citizens who lack cognomina, would appear to be that latest dated cases of this practice.

(90) M. P. SPEIDEL, Citizen Cohorts in the Roman Imperial Army in TAPA 106, 1976, p. 339-48 : 341 and 345 (= IDEM, Roman Army Studies I, Amsterdam, 1984, p. 91-100 : 93 and 97).

(91) CIL 6, 3520 = [LS 2731.

(92) RMD 266 (for 143 : cf. CIL 16, 38 for 94, for cohors VIII Voluntariorum c.R., another citizen unit, evidently accepting peregrines from c. 69) ; RMD 397 (for 144) ;

(?)RMD 401 (for 146) ; CIL 16, 179 and 180 (for 148) ; CIL 16, 113 (for 160) ;RMD 415 (for 154/156) ; RMD 110 (for 154/161) ; RMD 102 and 103 (for 157) ; RMD 448 (for 157/192) ; and RMD 446 and 447 (for 192).

(18)

Roman period. To that we might add how the stylistic and epigraphic features of this monument suggest a date on either side of 100. Thus, given that at least one other unit from Pannonia Inferior was probably in Asia Minor during Trajan’s Parthian War, namely the ala Flauia Augusta Britannica milliaria c.R., it seems entirely possible that the cohors I Campanorum was also here at that time, either in Amastris or passing by or through its vicinity when our soldier died.

That aside it is of interest that our soldier’s wife, Magna, designated here as his ‘uxor’ (93), is named on the text, suggesting she accompanied him on cam-paign (9“). As previously observed, until the time of Severus serving soldiers below the rank of centurion were prohibited from contracting a legal marriage, while any lawful marriage contracted by a man before he entered the army was formally and legally voided on enlistment for the duration of his military ser— vice, (95). The logic behind these regulations was to allow the state to avoid any form of responsibility for the wife and children of serving soldiers, especially if (as would seem to be the case here) they followed their men folk when they were temporarily or permanently moved from their original base to another place of duty. Nonetheless, it is clear that at regimental level at least individual Roman commanders effectively turned a Nelsonian blind eye towards the private lives of the ordinary soldiers they ostensibly controlled (’6).

8) Cohors I Lepidiana equitata bis torquata c.R.

A funerary text found at Magnesia ad Syplum suggests that the cohors I Lepidiana, was also in Asia Minor at the time of Trajan’s Parthian War (97).

D(is) M(anibus) / L(ucius) CALPVRNIVS VALENS OPTIO / COH(ortis) I LEPID(ianae) EQ(uitatae) C(iuium) R(omanorum) > PONTICI / VIXIT AN(nis) XXXX MIL(itavit) AN(nis) / XVIII CALPVRNIA LEDA / CO<N>IV<N>X FECIT

‘To the Shades ! Lucius Calpumius Valens, optio in the cohors I Lepidiana equitata civium Romanorum, of the centuria of Ponticus. He lived 40 years and served 18 years. His wife, Calpumia Leda, (had this) made.’

(93) After ‘CONIVNX’, ‘VXOR’ is the second most common epigraphic term denot-ing the wife of a servdenot-ing soldier: M. M. ROXAN, Women on the Frontiers in V. A. MAXFIELD and M. J. DOBSON (edd), Roman Frontier Studies, Oxford, 1991, p. 462-467 : 462.

(94) It could be, of course, that she had remained ‘at home‘, and paid for the monu-ment to be erected after learning of his death, although this seems less likely to be the case.

(95) J. B. CAMPBELL, op. cit. [n. 10].

(96) Cf, also C. VAN DRIEL-MURRAY, Gender in Question in P. RUSH (ed), Theoretical Roman Archaeology .' 2nd Conference Proceedings, Oxford, 1995, p. 3-21.

(97) CIL 3, 12251 = ILS 2590 = AE 1890, 159 : cf. CHRISTOL and DREW-BEAR, op. cit. [n. 88], p. 62, for the correct provenance of this item, sometimes mistakenly attributed to Izmir.

(19)

The combination of the formulae ‘D M’ in the ﬁrst line and the inscription

naming the deceased in the nominative case implies a date either side of 100, as

does the reporting of the deceased’s full tria nomina and the way his nomen is

spelt out in full. Just as was the case in at least three of the texts discussed above,

the subject of this epitaph was clearly one of those Roman citizens who chose to enlist in a peregrine regiment. His rank as an optio, an administrative grade immediately below that of centurion or decurion, was slated among the princi-pales, showing that he was quite probably due for promotion at the time he died (98). Finally, as with the text just previously discussed, here again we have a case of a serving soldier whose wife was most probably with him at the time of his death.

The cohors I Lepidiana equitata ciuium Romanorum is unique in being the only cohort named with a gentilicium in adjectival form, presumably that of its ﬁrst or an early and notable commander. This practice was, however, relatively common amongst some of the alae raised in the early principate from the Celtic-speaking parts of the Roman Empire (99), which at the very least suggests a ter-ritorial origin for those recruits who formed the core of the initial cohors I Lepidiana equitata ciuium Romanorum even though who the ‘Lepidianus’ in question might be is unknown. Be that as it may, the available diplomata

indi-cate that the unit was in Pannonia on 13-vi-80, and that it had been posted to

Moesia Inferior by 97, where it is registered for the

ﬁrst time with the battle

hon-our ciuium Romanorum, presumably won for service in one of Domitian’s Danubian escapades (10°). Other diplomata show that the unit was assigned to

Moesia Inferior in the years 99, 105, 113, 125 and 127 (1“), although the epi—

graphic features of the Magnesia text point to a stay in Asia Minor at the turn of

the 1“ century. As it was, sometime before c. 199, the unit was evidently

(re-)posted to Asia Minor, for as noted above, a building stone records its presence at Melik Sérif (Chorsabia/Carsagis) by then (‘02). This transfer to what later became the province of Armenia Minor was presumably made in connection with the eastern campaigns of either Verus or Severus, after which the unit

evi-(98) D. J. BREZE, The Organisation of the Career Structure of the Immunes and

Principales of the Roman Army in B] 74, 1974, p. 245-292 : 270 ; also IDEM, A Note on

the Use of the Titles Optio and Magister below the Centurionate during the Principate in Britannia 7, 1976, p. 127-139, esp. 132.

(99) E. BIRLEY, Alae Named after their Commanders in Ancient Society 9, 1978, p. 257-273, esp. 264—271 (= IDEM, The Roman Army: Papers 1929-1986, Amsterdam,

1988, p. 375—382, esp. 375-382.

(100) Pannonia : CIL 16, 26 (for 80) ; Moesia Inferior : RMD 337 (for 97).

(101) CIL 16, 45 and PFERDEHIRT, op. cit. [n. 39], p. 20-21, no. 8 (for 99) ; PFERDEHIRT, op. cit. [n. 39], p. 33-34, no. 11 (for 105) ; CIL 16, 38 (for 113) ; RMD 235 and (?) 364 (for 125), and 241 (for 127).

(20)

dently remained in the east henceforth, being named in the Notitia Dignitatum of c. 395 as one of the regiments commanded by the Dux Armeniae (”3).

Given that the epigraphic features of the Magnesia ad Syplum text indicate a date around the turn of the 15‘ century, then we might reasonably assume that the cohors I Lepidiana was there in connection with Trajan’s Parthian War when our soldier died. Yet Magnesia ad Syplum is not the most obvious place to expect to ﬁnd a tombstone of a serving Roman soldier at any point in time during the prin-cipate, as by then it seems to have been a place of little signiﬁcant status or

importance. Even so, it does lie at one end of the fertile Hermus Plain and it was

a way point on a minor trans—Anatolian route connecting the western Aegean coast and the Hermus valley with the Anatolian Plateau via Apamea ('0‘). Perhaps, then, the cohors I Lepidiana was passing through Magnesia on its way to or from central Anatolia or parts further east when our soldier died : but it is much more likely that it was stationed there or in the vicinity in some form of overall supervisory capacity, controlling the collection of food supplies from the broad and fertile Hermus valley (”5). Whatever explanation for the unit’s pres-ence at Magnesia, though, this inscription provides us with yet another example of how at least the junior ofﬁcers of Roman auxiliary units might travel with their legal spouse while on active duty.

9) cohors IIII Gallorum (equitata)

Another European unit that was seemingly sent to the east in connection with Trajan’s Parthian War was that cohors IIII Gallorum equitata recorded in Cilicia on 19-viii-l2l when one of its infantrymen received his honourable

dis-charge (106).

‘.../ PEDIT(ibus) ET EQVIT(ibus) QVI M(ilitaverunt) IN COH(orte) IIII GALL(orum) QVAE EST / IN CILICIA. ..’ (intus tabella I)

‘.../ PEDITIB(us) ET EQVITIB(us) QVI MILITAVERVNT IN COHORT(e) / IIII

GALLOR(urum) QVAE EST IN CILICIA. ..’ (extrinsecus tabella I)

(103) Not.Dig. Or. 38, 35, its then base being at ‘Caene Parembole’ : for views on the location of this site, see A. BRYER and D. WINFIELD, The Byzantine Monuments and Topo-graphy of the Pontos, Washington DC, 1985, p. 327 and 330, with J. CROW and A. BRYER, Survey in Trabzon and Giimii hane Vilayets, Turkey, 1992-1994 in DOP 51, 1997, p. 283-289, 287, n. 14, both suggesting Eski Pazar; while E. L. WHEELER, The Army and the Limes in the East in P. ERDKAMP (ed), A Companion to the Roman Army, Oxford, 2007, p. 235-266, 256, favours Apsarus.

(104) Although no Roman period road has yet been identiﬁed along this route, it was one of some antiquity 2 cf. B. LEVICK, Roman Colonies in Southern Asia Minor, Oxford,

1967, endnote map ; also CHRISTOL and DREW-BEAR, op. cit. [n. 88], p. 57.

(105) Cf. the cohors VII Breucorum, registered at Gordion during Trajan’s Parthian War, and most likely there to supervise food supplies from the region for that campaign : BENNETT and GOLDMAN, op. cit. [n. 28].

(21)

e

‘...To the infantrymen and cavalrymen who have served in the cohors IIII Gallorum, which is in Cilicia. ..’

The exact identity of this cohors IIII Gallorum equitata presents something of a mystery. From the available diplomata we learn that a cohors IIII Gallorum equitata was in Moesia and subsequently Moesia Inferior between 75 and 105 (‘07), while a variety of epigraphic evidence in addition to diplomata register a unit with the same number and name in Britannia from Trajanic times (if not before) until at least 178 (‘08). There again, a cohors IIII Gallorum is also record-ed on diplomata for Raetia on a repeatrecord-ed and regular basis between 107 and 167/168 (‘09), although it is not known whether this was equitate or peditate, while a cohors IIII Gallorum equitata c.R. is also regularly listed among the gar-rison of Mauretania Tingitana between 88 and 161 ~ albeit it is not named with the title c.R until 109 (11°). Finally there was a cohors IIII Gallorum equitata in Thracia on 19 vii 114, when one of its equites came to the end of his 25 years of military service (1“). In other words, the evidence currently available shows that there were perhaps no less than ﬁve cohortes [III Gallorum equitata in addition to the one named on the diploma for Cilicia of 121.

As it is a consideration of geographical distance and a comparison of the chronological markers makes it highly unlikely that our Cilician cohors IIII Gallorum equitata was the same as any one of those like-named and numbered

units recorded in either Britannia, Raetia, or Mauretania Tingitania. Therefore,

unless we are dealing with an entirely new regiment, the Cilician cohors IIII Gallorum equitata should be the same as either that recorded in Moesia Inferior between 75 and 105 or its homonym in Thrace in 114. On the other hand, the simple matter of geographical proximity and historical context likewise makes it more than probable that these two regiments — and thus also our Cilician cohort — were one and the same unit, having been deployed domino-fashion from Moesia Inferior to Thrace sometime after 105, and then to Cilicia after 114. \Obviously, if we are indeed dealing with a single unit, then such a series of

(107) RMD 2 (for 75), and 338 (for 97) ; and CIL 16, 50 (for 105).

(108) RIB 619, 620, 2458, 9-10, and 2472.1-2 ; and CIL 16, 69 (for 122) ;RMD 24 (for 127) ; CIL 16, 93 (for 145/146) ; RMD 420 (for 158), and 184, 29 3 and 294 (all three for 178).

(109) CIL 16, 55 (for 107) ; RMD 155 and 229 (both for 116), and 243 (for 129), and 387 (for 140), and 94 (for 138/140) ; CIL 16, 94 (for 140) ; RMD 170 and 275 (both for 157) ; CIL 16, 117 (for 153/157) and 16, 183 (for 156/157) ; (?)RMD 278 (for 160), and 434 (for 157/161) ; CIL 16, 21 (for 166) ; and RMD 68 (for 167/168).

(110) CIL 16, 159 (for 88), and 16, 161 (for 109), and 16, 165 (for 116), and 16, 73 and 16, 169 (both for 122) ;RMD 157 (for 131), and 382 (for 135), and 409, 410 and 411 (all three for 153), and 53 (for 159), and 107 (for 161).

(111) RMD 227/14 ; it had left Thrace by 138, when RMD 385/260 was issued for that

(22)

icomedeia Trapezus /- Amaseia ‘0 O . 5-6 NIcaea /\/.Nicopolisr‘ Juliopoiis _{Sebastopolis’k} ./'Satala yCarsagis

Gordlum.2/Ancyra _Sebasteia

5 8 Docimium Caesars la A/rabissus zM/litene . Magnesia_ . ___________ /" " Sardis Smyrna Ephesus ArchelaIs W'Cocusus lconium Samosa = V v o GorgoromeIs ‘/ Zeugma ‘ ‘ Hierapolis. . Antiocheia 3&m

FIG. 1. —— Asia Minor showing the places and locations refelred to in the text along with other key geographical locations (illustration by Ben Claasz Coockson)

.I.LE[NNE[H

(23)

movements ﬁts best within the context of Trajan’s Parthian War (“2). In other words, it could have been that this particular cohors IIII Gallorum equitata was initially transferred to Thrace in c. 113, replacing a regiment from there which had been ordered further east for Trajan’s forthcoming campaign. Then, in c. 115/116, when Trajan began to realise how gravely he had underestimated the forces he required to hold his new provinces of Armenia, Mesopotamia, and Assyria (“3), it too was sent to the east, eventually ﬁnding itself in Cilicia some— time before 121.

A brief recapitulation of the major points will serve to end this study. To begin with, of the nine auxiliary cohorts named as being ‘absent on campaign’ on the new diploma for Moesia Superior for 115, at least three and perhaps four have left epigraphic records with formulae or other internal evidence showing they

were in Asia Minor at about this time, and so presumably in connection with

Trajan’s Parthian War : the [III Raetorum ; VII Breucorum equitata c.R. ; I Thra-cum Syriaca equitata; and (perhaps) the I Montanorum (equitata?). What is more, another inscription from Asia Minor with epigraphic features of the same approximate date reports the presence there of at least one other unit otherwise known to have been regularly based in Moesia Superior at the time: the ala Claudia Noua (miscellanea). Furthermore, it seems quite likely that service in Trajan’s Parthian War explains why four more units normally based in one or other of the European provinces have left epigraphic traces for their presence in Asia Minor: the ala I Flauia Augusta Britannica >< c.R. and the cohors I Campanorum c.R. from Pannonia Inferior; the cohors Lepidiana equitata bis

torquata c.R. from Moesia Inferior ; and the cohors IIII Gallorum equitata from

Moesia Inferior via Thrace. Consequently the conﬂuence of epigraphic evidence places us in a much better position to assess what supplementary (auxiliary) forces Trajan deployed for his Parthian enterprise. To this we might simply add that others will no doubt wish to explore more fully how certain of these texts report matters of broader interest with regard to the Roman army : speciﬁcally, how no less than four of the inscriptions discussed here refer to auxiliary soldiers travelling with their wives and other family members when on active duty ; and how three of them record ciues Romani who chose to serve in the nominally peregrine auxilia.

Bilkent University, Ankara. Julian BENNETT.

(112) The possibility that the [III Gallorum moved from Thrace to Cilicia at a later date, i.e., under Hadrian, can surely be discounted : we can be certain that when Hadrian transferred his ‘sacred armies’ from the east to Moesia he made certain that sufﬁcient units were left in the Anatolian provinces for their security ; and there is nothing in our sources to suggest the region needed re-garrisoning between 117 and 121.