The Context and Date of Priscus Fragment 6

Author(s): Brian Croke

Source: Classical Philology, Vol. 78, No. 4 (Oct., 1983), pp. 297-308
Published by: The University of Chicago Press
Stable URL: http://www.jstor.org/stable/269957
Accessed: 14/04/2009 04:18

Your use of the JSTOR archive indicates your acceptance of JSTOR's Terms and Conditions of Use, available at
http://www.jstor.org/page/info/about/policies/terms.jsp. JSTOR's Terms and Conditions of Use provides, in part, that unless
you have obtained prior permission, you may not download an entire issue of a journal or multiple copies of articles, and you
may use content in the JSTOR archive only for your personal, non-commercial use.

Please contact the publisher regarding any further use of this work. Publisher contact information may be obtained at
http://www.jstor.org/action/showPublisher?publisherCode=ucpress.

Each copy of any part of a JSTOR transmission must contain the same copyright notice that appears on the screen or printed
page of such transmission.

JSTOR is a not-for-profit organization founded in 1995 to build trusted digital archives for scholarship. We work with the
scholarly community to preserve their work and the materials they rely upon, and to build a common research platform that
promotes the discovery and use of these resources. For more information about JSTOR, please contact support@jstor.org.

The University of Chicago Press is collaborating with JSTOR to digitize, preserve and extend access to
Classical Philology.

http://www.jstor.org
 THE CONTEXT AND DATE OF PRISCUS FRAGMENT 6

BRIAN CROKE

A PERENNIALproblem in analyzing ancient texts preserved only in

fragments and excerpts is establishing their context and, especially
important in the case of historical texts, their date. The many
fragments of early Byzantine historians in particular have been exten-
sively used by modern scholars, but there is much scope for more precise
and rigorous analysis of them. In using Priscus of Panion, for example,
as a source for the 440s we rely largely on the excerpts preserved in two
compilations made for the tenth-century scholar-emperor, Constantine
Porphyrogenitus. They deal with embassies from the Roman imperial
court to various barbarian chiefs and officials (Excerpta de legationibus
Romanorum) and embassies from barbarians to Romans (Excerpta de
legationibus gentium).
Since the excerpts are confined to embassies and are torn from their
context in the original history of Priscus, dating individual excerpts is
primarily achieved by correlating the excerpts with other sources for the
440s, notably the annalistic chronicles (e.g., Marcellinus, Prosper, Chro-
nicon Paschale, Theophanes).2 A useful check on this method is to analyze
the Excerpts to see if they bear witness to a systematic and sequential
process of extraction on the part of the Byzantine editor. Taking the
fragments relating to the Hunnish wars in the 440s (i.e., 1-14), we find
that fragments 1, 4, and 8 are successive extracts from Excerpta de le-
gationibus Romanorum. Fragment 1 describes events in 435, fragment 4
events in 447, and fragment 8 recounts the embassy to Attila of Maximinus
and Priscus in 449.3 They are therefore in order; and this is made clearer
still by the next successive fragments of the Roman embassies: 13(449),
14(449), 18(452),24(456), 25(456), 28(461), 29(462), 32(464), 33(465), 40(467).
When we turn to the Excerpta de legationibus gentium we discover that
the compiler here engaged in the same process: that is, he excerpted

1. Ed. C. de Boor, Excerpta de legationibus (Berlin, 1903). To my knowledge there is no comprehensive

study of the purpose and editorial technique in the Excerpta (cf. H. Hunger, Die hochsprachliche profane
Literatur der Byzantiner, vol. 1 [Munich, 1978], pp. 361-62, 366). A measure of this neglect is the fact
that in his lengthy discussion of Constantine's literary output and patronage Arnold Toynbee (Constantine
Porphyrogenitus and his World [London, 1973], pp. 575-605) makes no mention of the Excerpta. Of
general interest in this regard is P. Brunt, "On Historical Fragments and Epitomes," CQ 30 (1980): 477-
94.
2. For Priscus and his history, see B. Baldwin, "Priscus of Panium," Byzantion 50 (1980): 18-61; and
R. C. Blockley, The Fragmentary Classicising Historians of the Later Roman Empire (Liverpool, 1981),
pp. 48-70, 113-23.
3. The dates used here are those of C. Miiller in Fragmenta Historicorum Graecorum, vol. 4 (Paris,
1851).

[? 1983 by The University of Chicago. All rights reserved]

0009-83 7X/83/7804-0002 $01.00

297
298 BRIAN CROKE

Priscus in the order in which the material appeared in Priscus' history,

or at least arranged the extracts in chronological order if they did not
appear that way in the original history. Fragment 2, which describes the
outbreak of the war in 441, is the first barbarian embassy excerpted. Then
follow fragments 3(442), 5(447/448), 6, 7(448), 12(448), 15(450), 16(450),
19(452),20(453), 21(453), 26(456), 27(460/466),30(462/463),31(464), 34(466),
35(466), 36(467), 38(467), 39(467), 41(468). Taken together, then, the ex-
cerpts of Priscus contained in Constantine's books on embassies clearly
indicate that they are preserved in chronological order. That this should
in fact be the case is suggested, in the absence of systematic research, by
the judgment of those scholars who have worked closely with both the
texts themselves, where such survive, and their excerpts,4 and is con-
firmed by some elementary analysis.5 At least the evidence is substantial
enough to give warning that any attempt to violate this sequence must
be very carefully considered.
Such an exception has been proposed recently by W. Bayless, who
would date fragment 6 to 441 precisely, "in spite of the fact that fragment
6 follows fragment 5 in the Excerpta de legationibus gentium."6It is the
purpose of this paper to examine more fully the validity of this exception
to the regular pattern of excerpts of Priscus preserved in Constantine's
books.

Fragment 6 of Priscus deals with the inability of the Roman court to cope
with Attila's constant embassies following the ratification of a treaty, and
begins as follows:

When the peace was made Attila again sent ambassadors to the Eastern Romans
demanding the fugitives. And they, receiving these envoys and flattering them with

4. E.g., C. de Boor, "Die Excerptensammlungen des Porphyrogenitus,"Hermes 19 (1884): 123.

5. So much is already known in the case of Procopius (B. Rubin, "Prokopios von Kaisareia,"RE 23.1
[1957]: 389-90). A sample comparison of the text of Theophylact Simocatta (ed. De Boor) with the
Constantinian Excerpta de legationibus Romanorum et gentium (ed. De Boor) shows that the extracts
reflect strictly the order of the original text: Exc. de leg. Rom. 1 (221.14-222.15 De Boor) = 1.4.6-9
+ 1.6.1-3; 2 (222.16- 22) = 1.6.4-5; 3 (222.23-223.28) = 1.8.1-11; 4 (223.29-224.7) = 3.17.2-3; 5
(224.8-226.6) = 6.11.4-21; 6 (226.7-31) = 7.15.8-14; 7 (226.32-227.24) = 8.1.1-8; 8 (227.25-33) =
8.15.2,7; Exc. de leg. gent. 1 (477.21-478.23) = 1.2.6-1.4.1; 2 (478.24-479.3) = 1.6.5-7.2; 3
(479.4-480.25) = 1.15.1-14; 4 (480.26-481.10) = 3.15.5-10; 5 (481.11-483.17) = 4.10.5-11.11;
6(483.18-485.14) = 4.12.8-14.7; 7 (485.15-21) = 4.14.8-9; 8(485.22-33) = 5.2.4-6; 9(485.34-486.27)
= 5.2.7-3.11; 10 (486.28-31) = 5.11.9 + 5.13.1; 11 (486.32-487.2) = 5.16.4-6; 12 (487.3-7) =
6.3.1; 13(487.8-23) = 6.3.5-9; 14 (487.24-488.7) = 7.7.3-8; 15 (488.8-30) = 7.13.3-7; 16 (488.31-489.3)
= 8.4.1-2. The chronological order of Theophylact is also preserved by Photius in his book-by-book
summary (Bibl. Cod. 65), as it is for his summaries of Theophanes of Byzantium (Cod. 64), Procopius
of Caesarea (Cod. 63), and Candidus (Cod. 79). In addition, research on Photius' literary technique has
only clarified the fact that, like the Constantinian excerptor, Photius followed the order of the texts he
was excerpting (T. Hagg, Photius als Vermittler antiker Literatur: Untersuchungen zur Technik des
Referierens und Exzerpierens in der Bibliothek [Uppsala, 1975], pp. 197-98, with the specific examples
of the speeches of Himerius [p. 128] and Dio Chrysostom [p. 160]).
6. "The Chronology of Priscus Fragment 6," CP 74 (1979): 154.
 PRISCUS FRAGMENT6 299

very many gifts, sent them back again, saying that they had no fugitives. And again,
he sent other men. When they had transacted their business a third embassy arrived
and after it a fourth, for the barbarian, seeing clearly the Romans' liberality, which
they exercised through caution lest the peace treaties be broken, wished to benefit his
retinue. And so he sent them to the Romans, forming new excuses and finding new
pretexts. They gave ear to every order and obeyed the command of their master in
whatever he ordered. The Romans were not only wary of undertaking war on him,
but they feared the Parthians who were, it chanced, making preparations for war, the
Vandals who were troubling the sea coasts. . ..7

While this is normally taken to refer to events following the settlement

with Anatolius in 447, it is now proposed that the treaty discussed "would
be the one-year treaty of 441 mentioned by Count Marcellinus."8This
thesis is advanced to account for the references in Priscus to Persian
preparations for war and Vandal sea raids, which cannot apply, it is
argued, to 447 and immediately thereafter. Bayless aims for yet greater
precision and argues that "the only war which the Romans fought with
the Persians in the period began in 441": consequently, Priscus "is really
referring to conditions before the outbreak of the Persian war."9There
are, however, many unconsidered difficulties in this thesis which em-
phatically support the traditional date for the important fragment 6 of
Priscus.
First, Priscus is describing a time after the ratification of the peace
when no fewer than four successive legations from Attila were received
in Constantinople. The theme of his account is the powerlessness and
impoverishment of the imperial court, a situation exacerbated by each
new embassy. Priscus states further that the Romans capitulated easily
to the Huns because of the threats of the Vandals, Persians, Isaurians,
Saracens, and Ethiopians.10 If we take this trepidation to apply to the
period of the very first legation and allow sufficient time for the conclusion
of the Hunnish war in 441 and the ratification of the treaty, then we
might, on this construction, expect Priscus to be referring to a period
from late 441 onward.

7. Priscus frag. 6 (FHG 4:76). The translation is that of C. D. Gordon, The Age of Attila (Ann Arbor,
1966), p. 68.
8. Bayless, "Chronology,"p. 155.
9. Ibid. Bayless arrives at this conclusion by suggesting that E. A. Thompson and others inferred a
treaty of 441 "from Priscus' statement in fragment 6 that the Persians were preparing for war" (ibid.).
Rather, Thompson and others reasonably deduce the one-year treaty from the ambiguous and compact
statement of Marcellinus (s.a. 441.1, ed. Mommsen, Chron. min., 2:80): "Persae, Saraceni, Tzanni,
Isauri, Hunni finibus suis egressi Romanorum sola vastaverunt. missi sunt contra hos Anatolius et Aspar
magistri militiae pacemque cum his unius anni fecerunt." Although this passage has been taken to mean
that Anatolius and Aspar together concluded a one-year truce with the Persians (0. Seeck, "Flavius
Ardaburius Aspar," RE 2 [1896]: 608, but not Geschichte des Untergangs der antiken Welt 6 [Stuttgart,
1920], p. 292) or, conversely, that the Huns with whom Aspar negotiated were those in the Caucasus
(A. Demandt, "magister militum," RE supp. 12 [1970]: 749, cf. 742), it is apparent that the one-year
treaties were those concluded by Aspar with the Huns (0. Maenchen-Helfen, The World of the Huns
[Berkeley and Los Angeles, 1973], p. 110) and Anatolius with the Persians.
10. Frag. 6 (FHG 4:76).
300 BRIAN CROKE

The Persian war, however, was certainly over by June 441.11 So if

fragment 6 refers to the period before the Persian war was concluded, it
means that the invasion of the Huns must have occurred, and been
resolved, much earlier in the year. We know that on 6 March 441 the
large fleet under Areobindus, which had been assembled to do battle
with the Vandals, had not yet sailed. 12 In other words, since it is scarcely
likely that the Huns would have invaded when the full strength of the
imperial army was still in Europe, the Hunnish war must have begun
after 6 March 441. Hence, on Bayless' reconstruction, we must believe
that before the conflict with Persia was terminated (i.e., about May), and
after 6 March, the Huns invaded Illyricum, were beaten off, a one-year
treaty was reached, and at least the first of a series of embassies went to
the eastern capital. The compression of all this action into March/April
441, while not absolutely impossible, results in a very tight fit.
More damaging to Bayless' case is his proposal that all the Hunnish
embassies took place not only before the conclusion of the Persian war
but before it even broke out. In doing so, of course, he is only following
Priscus' statement that the embassies occurred while the Parthians "were,
it chanced, making preparations for war" (HapOvaiov ?v rrapao-K'v)
rvyX6avovraq).'3Although the war was not concluded until some time in
mid-441, the Persian king Isdigerdes II (438-457) had actually mobilized
his troops the previous year, or even as early as 439. The Persian war
was well under way in 440.1'4Consequently, in the period after the Hun-
nish war of 441 the Persians were demobilizing or had already decided
on peace. Because the Persian war began in 440, they could only be
described as "preparing for war" in early 440/late 439 and certainly not
in early/mid-441. This problem, which clearly suggests that the events
described in fragment 6 cannot be dated to 441, is overlooked by Bayless.
A still more serious objection to assigning fragment 6 to 441 is its general
theme of Roman foreboding and incapacity. In general, the court of Theo-
dosius had been able to treat the Huns with relative disdain in the early
440s. For example, before the outbreak of war in 442 Theodosius' court

11. NTh. 5. 3 (26 June 441), with Maenchen-Helfen, Huns, p. 109. The statement of Theodoret HE
5. 37 that the Persians attacked only when the Romans "were occupied with other wars" is often taken
(e.g., by Maenchen-Helfen, ibid.) to refer to 441; but it describes circumstances in 421/422 (Seeck,
Geschichte, p. 85; B. Croke, "Evidence for the Hun Invasion of Thrace in A.D. 422," GRBS 18 [1977]:
349-52; and K. Holum, "Pulcheria's Crusade A.D. 421-22 and the Ideology of Imperial Victory," GRBS
18 [1977]: 167-71).
12. NTh. 7. 4.
13. Frag. 6 (FHG 4:76).
14. J. B. Bury, History of the Later Roman Empire2 (London, 1923), vol. 2, p. 6; E. Stein, Histoire
du Bas-Empire (Paris, 1959), vol. 1, pp. 291-92. The decisive evidence is E. Vartabed, Hist. Arm. 1
(Collection des Historiens de 1'Armenie, ed. V. Langlois, vol. 2 [Paris, 1869], p. 184), who says that "in
the second year of the reign of Izdegerd II," that is, 439-440, "il fondit sur le pays des Grecs [Romans],
s'avanca jusqu'a la ville de Medzpin [Nisibis] et devasta, en les saccageant, plusieurs villes appartenant
aux Romains. . . . Cependant le bienheureux empereur Theodose . . . lui envoya des sommes d'argent
considerables par un personnage appele Anadol [Anatolius] qui etait son general en Orient"; cf. Moses
of Chorene, Hist. Arm. 3. 67 (Langlois, p. 172), who dates the invasion to the time of the death of the
Armenian patriarch Sahag, "at the beginning of Izdegerd's second year"; also referred to in T. Noldeke,
Geschichte der Perser und Araber aus der arabischen Chronik des Tabari iibersetzt (Leiden, 1879), p.
116, and Michael the Syrian, Chron. 8. 14.
 PRISCUS FRAGMENT6 301

responded to the threats of the Hunnish king, Attila, by refusing to

surrender Hunnish refugees living among the Romans. In fact, they were
prepared to do battle with the Huns, wrote Theodosius, and would send
ambassadors to dissolve their agreements.'s Furthermore, when the war
with the Huns was concluded in 441, the imperial court worked out a
satisfactory arrangement with the Vandals and recalled the imperial army
from Sicily, to bolster the forces fighting the Huns in Illyricumand Thrace.16
In 443 the Romans were purposefully building up defenses on all frontiers
as well as refurbishing the Danubian flotilla and camps.17
Consequently, the Romans were in a relatively robust and confident
position in 441/442, in contrast to their position after the crushing losses
to the Huns in 447.18 In fragment 5 Priscus records the impact of the
Roman defeat and the crippling terms imposed by the Huns. He describes
the levy of new taxes, how some senators put their furniture and wives'
jewelry up for sale, and how others preferred suicide, either by the noose
or starvation. In addition, Priscus comments that the Romans pretended
to agree voluntarily to such harsh terms, while they were actually com-
pelled to concur by fear and necessity.19Clearly the tone of fragment 6,
describing an exhausted and bankrupt Roman state, pressed in on all
sides and fearful of its enemies, most closely accords with the situation
after 447, not that in 441. This observation is reinforced by a consideration
of the fragment's position in the excerpts of Priscus.
It was noted above that the order of the Constantinian excerpts relating
to the Hunnish invasions reflects the chronological order of Priscus' work.
On this assumption fragment 6 would belong between fragment 5 (the
terms of Anatolius' treaty and its impact) and fragment 7 (the embassy
of Edeco), while the date of fragment 5 would provide the terminus post
quem for fragment 6. It is vital, therefore, to establish the date of fragment
5.

II

Although some scholars in the past have dated fragment 5 to 442 or 443,
it is now very firmly fixed in 447/448.20 The two main criteria for this
date are the geography of the events described and the identity of the
ambassadors involved-Anatolius and Theodulus. Let these be consid-
ered in turn.
First, it must be pointed out that the only chronological indication
provided in fragment 5 is the phrase "after the battle in the Chersonese"

15. Priscus frag. 3 (FHG 4:73).

16. Theophanes A.M. 5942 (p. 102 De Boor).
17. NTh. 24.
18. For which, see Maenchen-Helfen, Huns, pp. 117-25, ascribing the upsurge in Hunnish power
after 445 to the personal ascendancy of Attila.
19. Frag. 5 (FHG 4:74).
20. Maenchen-Helfen, Huns, pp. 111-25; Seeck, Geschichte, pp. 291-94.
302 BRIAN CROKE

Trjv gv Xeppovro-n
(pjrLEd tLaxrv). In 441 the Huns invaded Illyricum only,21
and in 442 broke into northern Thrace as well.22 Only in the "Great
Invasion" of 447, however, after the Chersonese "Long Wall" had been
damaged in the earthquake of 26 January, did the Huns penetrate Thrace
and move into the Chersonese.23 Indeed, much of the province of Europa,
in which the Chersonese was incorporated, was leveled in the invasion.24
"After the battle in the Chersonese" can therefore only refer to events in
447 and later.
Second, the ambassador Anatolius was, as magister militum, occupied
in the East against the Persians in 440/441.25 In 442 he was in Edessa,
where he provided a silver reliquary for the bones of the apostle Thomas.26
He was still magister militum in the East on 28 January 443.27 Anatolius
had been stationed in the East since 433, in addition to a previous posting
in 420, and was only recalled to Constantinople to take up the post of
magister militum praesentalis in 447.28 The mere fact of Anatolius' pres-
ence in Thrace, acting on behalf of the emperor, presupposes a date of
447 or later. Finally, Theodulus is designated in fragment 5 as magister
militum per Thracias.29 He only succeeded to this position in 447, when
his predecessor, Arnegisclus, was killed soon after the outbreak of the
Hunnish invasion.30
In 447/448 therefore, according to fragment 5, Anatolius is sent on an
embassy to the Hun court. Attila subsequently concludes an agreement
(o-v,8z3do-8es) with Anatolius, but makes plain, in voicing his dissatisfaction
at the Romans' retention of Hun prisoners and refugees, that he will not
ratify the treaty (ra? rTq eiphvrpl (jvv06Ka') unless the prisoners are sur-
rendered or ransomed.31 Fragment 6 begins with the statement that the
treaty had been made and then goes on to describe the procession of

21. Marcell. comes, s.a. 441. 3 (ed. Mommsen, Chron. min., 2:80): "Hunnorum reges numerosis
suorum cum milibus Illyricum irruerunt: Naisum, Singidunum aliasque civitates oppidaque Illyrici
plurima exciderunt."
22. Marcell. comes, s.a. 442. 2 (ed. Mommsen, Chron. min., 2:81): "Bleda et Attila fratres multa-
rumque gentium reges Illyricum Thraciamque depopulati sunt"; Prosper 1345 s.a. 442 (ed. Mommsen,
Chron. min., 1:479): "Chunis Thracias et Illyricum saeva populatione vastantibus exercitus, qui in Sicilia
morabatur, ad defensionem Orientalium provinciarum revertit."
23. Evagrius HE 1. 17.
24. Marcell. comes, s.a. 447. 2 (ed. Mommsen, Chron. min., 2:82): "Ingens bellum et priore maius
per Attilam regem nostris inflictum paene totam Europam excisis invasisque civitatibus atque castellis
conrasit."
25. Marcell. comes, s.a. 441. 1 (ed. Mommsen, Chron. min., 2:80), with J. R. Martindale, Proso-
pography of the Later Roman Empire II (Cambridge, 1980), p. 85 ("Fl. Anatolius 10"). For the vexed
question of Anatolius' tenure of this post, see Holum, "Pulcheria's Crusade," p. 169 with nn. 13 and 66.
26. Chron. Edess., s.a. 753 (=A.D. 442); PLRE 11:85.
27. CJ 1. 46. 3.
28. Theod. Ep. 45; PLRE 11:84.
29. Priscus frag. 5 (FHG 4:76), with Demandt, RE Supp. 12 (1970): 745.
30. Since Theodulus was already magister militum at the time of the embassy of Senator (Priscus
frag. 4 [FHG 4:74]), which must postdate 445 (Maenchen-Helfen, Huns, p. 119), and since Arnegisclus
is recorded as magister militum per Thracias early in 447 (Jordanes Rom. 331; Chronicon Paschale [ed.
Dindorf] 586. 5) when he was killed, then Arnegisclus must have succeeded the Vandal John whom he
killed in 441 (Marcell. comes, s.a. 441. 2 [ed. Mommsen, Chron. min., 2:80]). This means that Theodulus
succeeded Arnegisclus, despite "Theodulus 2" PLRE II: 1105-6, which assumes a date of 443 for the
Peace of Anatolius.
31. Frag. 5 (FHG 4:75).
 PRISCUS FRAGMENT6 303

embassies to the Romans and the pliancy of the imperial court in acceding
to each new request.32It must firmly be stated that fragment 6 fits perfectly
between the final ratification of the peace and the embassy of Edeco to
Constantinople in 448 recounted in fragment 7. It would seem, then, that
the excerpt was taken from Priscus' account of the aftermath of the
disastrous invasion in 447.
There are therefore only two possible explanations for fragment 6. (1)
If we acknowledge, as Bayless does, that fragment 5 can only be dated
to 447/448,33 and if we accept, as Bayless also does, the usual assumption
that the fragments of Priscus are preserved in chronological order, then
the events described in fragment 6 must follow those in fragment 5. They
are therefore to be dated to 447/448 or just after, and require explication
in the light of events in the late 440s, when successive Hunnish embassies
did in fact visit Constantinople. (2) If, on the other hand, it is insisted
that fragment 6 can only be dated to 441 (while fragment 5 can only be
dated to 447/448), then we are obliged to regard fragment 6 as an exception
to the sequential order of the fragments.
We have already seen that this latter explanation is almost completely
invalidated because of the strict sequence of the fragments, the difficulties
in placing fragment 6 in 441, and the continuity between fragment 5,
which can only be dated to 447/448, and fragment 6. Such an explanation
becomes even more problematic if the details which are claimed to apply
exclusively to 441 can be shown to fit circumstances in 447 and afterwards.
Consequently, it is necessary to consider the specific details contained in
fragment 6.

III

The core of Bayless' argument is his contention that the picture of bel-
ligerent Persians and Vandals only makes sense before 442. A close ex-
amination of Priscus' statement, that the Romans were submissive because
they were being threatened by other enemies, exposes the imprecision of
the historian's description. He continues: "[The Romans] were not only
wary of undertaking war on him [Attila], but they also feared the Par-
thians, who were, it chanced, making preparations for war, the Vandals,
who were troubling the sea coasts (ra KraraO&Xarrcav raparrov7ra), the
Isaurians, who had set out on banditry (r7rvXorTrsiav), the Saracens, who
were overrunning the eastern part of the empire (rr)v wt KTaraCTpeXovra),
and the united Ethiopian races (rd AtiOLOTLKd&
o Ov7 oUavvto-'rdleva).Being
humbled they danced attendance on Attila and strove to meet other races
with military power, mustering their forces and appointing generals."34
We hear only vaguely of Persians preparing for war, Vandals raiding
32. Frag. 6 (FHG 4:76).
33. "Chronology," p. 154 (thereby contradicting his earlier support for the date of 443: "The Treaty
with the Huns of 443," AJP 97 [1976]: 176-80).
34. Frag. 6 (FHG 4:76), translated by Gordon, Attila, pp. 68-69.
304 BRIAN CROKE

coastlines, Isaurians and Saracens plundering the eastern provinces, Ethi-

opians posing a threat, and the Romans appointing generals to organize
their forces. This account (which indeed offers little in the way of chron-
ological precision) is alleged by Bayless to correspond to that described
in the chronicle of Marcellinus for 441: "The Persians, Saracens, Tzanni,
Isaurians and Huns left their own territories and plundered the land of
the Romans. Anatolius and Aspar, Masters of the Soldiery, were sent
against them, and made peace with them for one year."35
Yet, in view of the arguments already advanced, the essential question
is not whether fragment 6 can be made to fit the account of Marcellinus
for 441, but whether the situation depicted by Priscus can be shown to
have existed in the otherwise sparsely documented period in and after
447. To consider fully this possibility it is necessary to describe and analyze
the relations in the late 440s between the Roman empire and the nations
that, according to Priscus, were threatening it.
The mobile and elusive desert tribes of Egypt and Arabia were always
difficult for the Romans to combat and control fully. The empire was
constantly at war with them, although these conflicts rarely appear in
our sources. We know, however, that in the latter part of his reign Theo-
dosius II had been forced to divide the province of Thebais into a lower
and upper half and take the uncommon step of uniting the civil and
military administration in a single official.36 Such action presupposes a
troublesome province and suggests that the Nobadae and Blemmyes had
been causing persistent problems for the emperor. It was at this time
(449-50) that the deposed patriarch of Constantinople, Nestorius, com-
plained of attacks by the Blemmyes and Nobadae near his place of exile.37
Soon after the accession of Marcian in 450 the Nobadae and Blemmyes
attacked and were defeated by the dux Florus. They agreed to seek a
peace treaty. Maximinus and Priscus journeyed to Egypt, and the treaty
was settled, but Maximinus died, whereupon the barbarians revoked the
agreement and began plundering anew.38
Adjacent to the Nobadae and Blemmyes, and frequently involved with
them, lay the kingdom of the Ethiopians, or Axumites, in a remote but
commercially strategic corner of the classical world. Our fragmentary
sources for the fifth century leave us in abysmal ignorance about events

35. Marcell. comes, s.a. 441. 1 (ed. Mommsen, Chron. min., 2:80).
36. Bury, History, 2: 237-38.
37. Evagrius HE 1. 7. It was perhaps at this stage that the Roman general Germanus ("Germanus
3," PLRE 11:505) was successful, as recounted in the fragmentary Blemyomachia of P. Berol. 5003 (most
accessible in D. Page [ed.], Select Papyri, vol. 3 [London and Cambridge, Mass., 1941], no. 142, pp.
590-94, with the three fragments in L. S. B. MacCoull, "Papyrus Fragments from the Monastery of
Phoebammon," Proc. 16th Int. Cong. Pap. [Chico, 1981]: 491-98); but it has been suggested recently
that the poem refers to an earlier campaign and was written by Olympiodorus of Thebes (E. Livrea,
"Chi e l'autore della Blemyomachia (P. Berol. 5003)?," Prometheus 2 [1976]: 97-123; cf. Anonymifortasse
Olympiodori Thebani Blemyomachia (P. Berol. 5003), Beitr. Klass. Philol. 101 [Meisenheim am Glan,
1978]).
38. Priscus frag. 21 (FHG 4:100); Jordanes Rom. 333. Priscus was at Alexandria during the following
year (Evagrius HE 2. 15).
 PRISCUS FRAGMENT6 305

in Ethiopia.39 When our materials become fuller and more varied, how-
ever, by the reigns of Justin I and Justinian, the Ethiopians and their
neighbors across the straits in the Yemen-the Himyarites-come into
clearer view, having been recently incorporated into the Christian em-
pire.40 The whole region appears constantly turbulent, as local factions
of pagans, Jews, and Christians (both monophysite and orthodox) vie to
assert control. It is not difficult to envision the mid-fifth-century (447/
448) forerunners of these generations displaying similar behavior and
thereby threatening the imperial government.
Some trade irregularity or the controversial religious policies of the
court in the mid-fifth century could easily have provoked rebellion and
dissent in a volatile Ethiopia within the doctrinal orbit of Alexandria,
just as they did in Egypt. The silence of our extant sources for Ethiopia
in the 440s does not necessarily indicate that any trouble not recorded
was "not serious enough."41 (It should be emphasized, too, that even if
fragment 6 of Priscus is to be dated to 441, there is still no corroborative
evidence for its statement that the Ethiopians were in revolt in 441.
Certainly Marcellinus does not include Ethiopians among the tribes ha-
rassing the Romans in that year; nor does Priscus mention the Tzani.
The argumentum ex silentio is especially untrustworthy when it is applied
selectively.) In fact, given Theodosius' division of the province of Thebais
and the situation which came to a head about 450, we have good reason
to suppose that the restless tribes of upper Egypt and Ethiopia were a
threat to the Roman administration at precisely this time. At least the
Himyarites were engaged in a devastating raid on Roman territory, per-
haps against Ethiopia and Egypt, in 450.42
Another intractable region where civil and military power were united
was Isauria.43 Ammianus was quick to point out what a fickle lot the
Isaurians were, at peace one minute, raiding nearby provinces the next.44

39. Bury, History, 2: 322: "We lose sight of [Ethiopia] for about a century and a half" (i.e., 350-
500). Our knowledge has not advanced since Bury's day. Note the gap in A. H. M. Jones and E. Monroe,
A History of Ethiopia (Oxford, 1956; repr. 1978), pp. 18-19; D. Buxton, The Abyssinians (London,
1970), pp. 41-42; and E. Ullendorff, The Ethiopians3 (Oxford, 1973), p. 53 ("We lack detailed docu-
mentation with regard to the following 100-150 years" [375-525]). For the rise of the Axumite kingdom
in the early fifth century, see F. Altheim and R. Stiehl, "Die Anfange des Konigreichs Aksum," Klio 42
(1964): 181-94.
40. A thorough account can be found in A. Vasiliev, Justin the First (Washington, D.C., 1950), pp.
283-302; see also Bury, History, 2: 322-27; Stein, Histoire, 2: 101-5.
41. Bayless, "Chronology," p. 155.
42. Y. M. Kobishchanov, Axum, tr. L. T. Kapitanoff and ed. J. W. Michels (University Park, Penn-
sylvania and London, 1979), p. 79. The late 440s may also provide the background for the recently
discovered inscriptions of the Axumite king Ezana II indicating conflicts with the Nobadae (SEG 26
[1976/77], no. 1813, p. 410) and Blemmyes (E. Bernand, "Nouvelles versions de la campagne du roi
Ezana contre les Bedja," ZPE 45 [1982]: 105-14); as well as for the papyrus letter indicating warfare
between the Blemmyes and Nobadae (T. C. Skeat, "A letter from the King of the Blemmyes to the King
of the Noubades," JEA 63 [1977]: 159-70, though a late-sixth-century date is argued by V. Christides,
"Ethnic Movements in Southern Egypt and Northern Sudan: Blemmyes-Beja in Late Antique and Early
Arab Egypt until 707 A.D.," Listy Filologicke 103 [1980]: 129-43). None of these documents can yet be
dated with absolute certainty.
43. A. H. M. Jones, The Later Roman Empire (Oxford, 1964), pp. 281, 609.
44. 14. 2. 1.
306 BRIAN CROKE

Although we hear of no direct trouble in Isauria in the mid/late 440s, it

is clear that some special circumstances had brought Zeno to Constan-
tinople and promoted him promptly. Furthermore, it was probably at
precisely this time (447-448), when he is attested in the military service
of Zeno, that Rufus captured and killed the Isaurian Balbinus, who had
raided certain cities in neighboring Cilicia.45By 449/450 Zeno was back
in Isauria, having become disgruntled over the Roman capitulation to
successive Hunnish embassies, and was apparently planning a revolt
against Theodosius.46 Maximinus was sent against the Isaurians with
military support.47 The Saracens, too, forever troublesome, were also
active in the period 447-450. At least we hear soon after of the general
Dorotheus campaigning against them in Palestine48and Ardaburius fight-
ing against them at Damascus, and of the ambassador Maximinus and
his friend Priscus arranging peace with them.49
There is, it seems, indirect evidence from the late 440s to vindicate
Priscus' statement in fragment 6 that the Romans had reason to fear the
Ethiopians, Saracens, and Isaurians in the period, of unspecified duration,
when four successive embassies were received in Constantinople from
Attila. In addition, we witness in the late 440s a flurry of military ap-
pointments to deal with various threats-Anatolius, Zeno, Ardaburius,
Rufus-just as Priscus notes. It is also worth emphasizing that our de-
tailed information about these particular trouble spots derives entirely
from Priscus, who had himself been involved in negotiations in each area
and who incorporated his adventures into his history. Were it not for
Priscus, then, we would not be in a position to appreciate the pressure
on the Romans in 447/448, precipitated by the demands of the Huns.
There remains the status of Rome's relations with the Persians and
Vandals in the late 440s. Since Priscus apparently had no dealings with
them, we are less well informed about these nations, dependent as we
are on the Excerpta de legationibus for our fragments of Priscus. It is
assumed by Bayless that the Romans had nothing to fear from either the
Persians or the Vandals after 442. Yet there is the testimony of Jordanes,
in a passage apparently derived from Priscus, that both the Vandals and
Persians were threatening the imperial realm early in the reign of Marcian
(450-57), who resolved peace with them.50What we really need to know,
however, is whether there had been such threats in the changed situation
of the late 440s, after the Romans had been so humiliated by the Huns.
Priscus, it will be observed, does not say that the Persians were pre-
paring to do battle against the Romans, but that when the imperial court
was being put under pressure "the Persians were, it chanced, preparing
for war." It so happens that in the late 440s the Persians were engaged

45. Malalas, 345. 8-11 (ed. Dindorf), with "Rufus 1," PLRE 11:958-59.
46. "Zenon 6," PLRE 11:1199-1200.
47. John of Antioch frag. 199. 1 (FHG 4:613).
48. Nic. Call. HE 15. 9; "Dorotheus 7," PLRE 11:377-78.
49. Priscus frag. 20 (FHG 4:100).
50. Rom. 33. For Jordanes' use of Priscus, see Mommsen, Jordanes (MGH:AA,5:1), pp. 34-37.
 PRISCUS FRAGMENT6 307

in their continuing war with the Caucasian Huns, the Ephthalites:5' Pris-
cus could be referring to that. Alternatively, his description appears to
fit neatly the situation in Armenia, for it is against the Armenians that
the Persians were preparing for war precisely in the late 440s. The Chris-
tian Armenians had been ruled by Persian governors since 438 and con-
stantly harassed by proselytizing Zoroastrians. Events reached a climax
in 449, when the edict of the marzpan Miht Narse required the forced
conversion of Armenians. The Christian Armenians resisted fiercely and
many suffered martyrdom. Meanwhile, in 449/450 the Armenians sent to
the Christian emperor in Constantinople for help. Theodosius was preoc-
cupied with the other enemies52 and declined to support them.53 The
Persians mobilized, and Armenian resistance soon faded after the defeat
of Vardan.54
The Vandal threat, too, is only described by Priscus as "coastal raids."
Such raids could simply be confined to small islands, or be otherwise
relatively harmless. Nonetheless, such raids would in themselves signify
the potential danger of the Vandals. It is true that there is no explicit
evidence for Vandal harassment of Roman territory between 442 and 455,
except for a raid along the coast of Gallaecia in 445.55 Yet this very
exception, a local raid reported only by a local chronicler, exposes the
danger of assuming that, because no direct evidence survives, no raids
took place. One is permitted to wonder how many other Vandal raids in
the later 440s went unreported or remain unknown because of the ex-
tremely patchy extant sources for the period. Consequently, we cannot
assert categorically and with complete confidence that the Vandals were
not seen as a threat in the late 440s, despite the treaty of 442. Certainly,
neither the mention of Persian preparations for war nor Vandal razzias
is strong enough, in the light of other difficulties, to support unequivocally
a date of 441 for fragment 6 of Priscus.
Moreover, it should be stressed that there is evidence in the period after
447 for the stream of Hunnish embassies to Constantinople taking ad-
vantage of the straitened circumstances of the Romans. Marcellinus de-
scribes how in 448 a severe fire at Constantinople destroyed the Troad
porticoes and the towers of certain gates. He goes on to say that these
were restored to their pristine condition by the City Prefect, Antiochus,
at a time when "Attila's envoys were demanding from Theodosius the
tribute formerly agreed on."56This fits most comfortably the circumstances
51. Bury, History, 2: 5, n. 7.
52. With the Huns at peace, this may indicate the threats of other enemies-Saracens, Ethiopians,
Isaurians.
53. For the sources (mainly Armenian), see Bury, History, 2: 5-6; Stein, Histoire, 1: 353. For the
background, see S. Nigosian, "Zoroastrianism in Fifth-Century Armenia," Studies in Religion 7 (1978):
425-34.
54. "Vardan," PLRE II: 1150-51.
55. Hydatius Chron. 131, s.a. 445 (ed. Mommsen, Chron. min., 2:24): "Vandali navibus Turonio in
litore Gallaeciae repente advecti familias capiunt plurimorum"; cf. C. Courtois, Les Vandales et l'Afrique
(Paris, 1955), p. 194, and E. A. Thompson, "The End of Roman Spain, Part 2," Nott. Med. Studs. 21
(1977): 24.
56. Marcell. comes, s.a. 447. 2, 3. (ed. Mommsen, Chron. min., 2:83): ". . legatis Attilae a Theodosio
depectas olim pecunias flagitantibus."
308 BRIAN CROKE

outlined in fragment 6 of Priscus and is another indication that the frag-

ment concerns not the early but the late 440s. There is no such corrob-
orative evidence for Priscus' picture of successive Hunnish embassies after
the invasions of 441 and 442. This is hardly surprising, since the treaties
following those invasions seem to have been readily acceded to by the
Huns at a time when the Romans still held the upper hand.
To conclude. In 441 the Huns under their kings Bleda and Attila in-
vaded Roman Illyricum. Their progress was arrested and a one-year truce
concluded by Aspar. The following year they invaded the Roman domain
again, this time reaching into Thrace. At this stage the Romans were still
in a strong enough position to cope with the Hunnish incursions. By 447,
however, their defensive capability had declined, and Attila's hordes were
able to devastate the European provinces. The aftermath of this invasion
and its settlement is reflected in fragment 6 of Priscus, where the Romans
are characterized as impotent, impecunious, and threatened by enemies
on all sides. Since the contents of fragment 6 can reasonably be explained
in the light of events in the late 440s, it is therefore to be dated to 447/
448 and is located logically between the peace settlements of Anatolius
(fragment 5) and the embassy of Edeco (fragment 7). So, in excerpting
the accounts of barbarian embassies contained in the history of Priscus,
Constantine Porphyrogenitus' editor naturally followed the chronological
order of the original work. Fragment 6 of Priscus should not be considered
the exception to this process after all.

Macquarie University