BMCR 2008.12.31, Gavin Kelly, Ammianus Marcellinus

Gavin Kelly, Ammianus Marcellinus: The Allusive Historian. Cambridge
Classical Studies.  Cambridge/New York:  Cambridge University Press,
2008.  Pp. xi, 378.  ISBN 9780521842990.  $99.00.

Reviewed by Eric Fournier,  West Chester University of Pennsylvania
([email protected])
Word count:  2872 words
-------------------------------
To read a print-formatted version of this review, see
http://ccat.sas.upenn.edu/bmcr/2008/2008-12-31.html
To comment on this review, see
http://www.bmcreview.org/2008/12/20081231.html
-------------------------------

Table of Contents
(http://catdir.loc.gov/catdir/enhancements/fy0838/2008298030-t.html)

Gavin Kelly's Ammianus Marcellinus: The Allusive Historian is a
thought-provoking and original study of a key fourth-century author. It
makes a valuable contribution to the field of late antique studies and
of Ammianus in particular by focusing on the literary aspects of the
historian's text. Bridging the gap across disciplines, Kelly's book
should appeal equally to classicists and historians. In doing so, Kelly
follows in the footsteps of Averil Cameron, Elizabeth Clark, and others
who in turn have taken their cue for reading texts from continental
philosophers.[[1]] The result is an engaging interpretation of Ammianus
as a creative and erudite writer. By the same token, Kelly also creates
a via media in scholarship on Ammianus between the two dominant voices
of the last two decades, John Matthews and Timothy Barnes, the former
largely taking Ammianus at his word as a starting point for a larger
exploration of the world the historian lived in, the latter presenting
Ammianus as a writer of fiction whose historical account was
unreliable.[[2]] Instead, Kelly appraises Ammianus on his own ground,
as a writer of history according to the ancient literary conventions of
the genre, and not following (dubious) modern criteria of objectivity
and veracity.

A "considerably revised" (ix) version of Kelly's 2002 Oxford thesis,
the book is divided in two parts, each presenting one main argument.
Part one ("The Elusive Historian") exposes the limits of the
biographical approach which has so far dominated scholarship on
Ammianus. Kelly shows the conjectural aspect of our knowledge of
Ammianus's life ("(un)likely but unprovable" is a constant leitmotiv),
and instead insists on the writer's careful self-representation within
the narrative. Part two ("The Allusive Historian") aims at shifting the
paradigm from the author toward his text. Kelly's analysis focuses
especially on intertextual elements of Ammianus's account (external and
internal allusions, exempla, structural and thematic similarities).
Against previous scholars who assumed Ammianus to be unsophisticated
and who judged him of poor taste for his abuse of literary flourishes,
Kelly insists on his genius and the meaningful political nature of his
literary devices ("I am happy to believe that he wrote with anger and
partiality" [6]).

The book is divided in seven chapters -- three for the first part and
four for the second -- in addition to a short Epilogue (318-21), an
Appendix (322-31, on the "Sources for the tsunami of AD 365"), a
Bibliography of both editions and translations of Ammianus and modern
studies (332-55), a Subject Index (356-64), an Index of Modern Scholars
(365-6), and a very useful Index Locorum (367-78). The introduction,
mainly devoted to surveying previous scholarship on Ammianus, ends with
a brief comment on the dating of the work, which Kelly places in the
late 380s-early 390s. More controversial is his claim that the work was
published all at once. The problem is that Kelly does not discuss any
alternative: why would Ammianus have not written in separate
installments and publicly presented his work in progress before
publication, for example?

Chapter 1 ("The Bones of the Battlefield") introduces the main themes
of the book, and constitutes a constant point of reference for the rest
of the work, perhaps in a deliberate imitation of Ammianus's
intra-textual habit. Kelly discusses the battle Ad Salices (AM 31.7.16)
of 377 CE in which Gothic and Roman forces clashed inconclusively, as a
prelude to Adrianople (although Valens was not directly involved). This
is an excellent discussion of some of the main features of Ammianus's
intertextual characteristics, which include allusions, window allusion,
autopsy, "archaeological reconstruction" as a literary device, and
"twisted ecphrasis" (15-21). It is puzzling, however, why such terms
are here presented for the first time without any definition. This
seems to be a structural problem, since section one of chapter four is
entirely dedicated to defining such terms (166-175). The latter section
would have been best placed in the introduction, or at the beginning of
chapter one. Definitions aside, Ammianus's description of the battle Ad
Salices is shown to illustrate the span of the historian's allusions,
from classics such as Tacitus and Vergil, to contemporary writers like
Libanius (19-20). Another important theme of the book introduced by
Kelly's analysis of this passage is the political implications Ammianus
conveyed through his allusions, in this case a reference to
flesh-eating birds as an augur of Adrianople (24). The behavior of the
military commanders at Ad Salices, who did not bury the dead and
therefore impiously allowed the birds to devour them, sharply contrasts
with Julian' "pious and correct" behavior after the battle of
Strasbourg, in yet another omen of both later chapters of the book and
the disaster of Adrianople.

With Chapter 2 ("The Adventures of Ammianus"), Kelly picks up his first
main task, essentially a negative one (as he himself admits at 108), to
demolish any appearances of certainty one might have regarding the life
and career of Ammianus. He first discards the approach of psychohistory
because the autobiographical passages, he argues, are carefully crafted
from a literary point of view and designed to reflect the historian's
experiences as he wished his reader to view them. This chapter mainly
analyzes the function of such accounts and argues that they
strengthened Ammianus's credibility by creating an "authoritative
persona" (33). In order to do so, Kelly reviews three problematic
aspects of such accounts which have traditionally led scholars to doubt
Ammianus's veracity, his bias toward his commander Ursicinus (44-53),
the embellishments included in his narrative of the campaign of 359
(53-4), and literary allusions (to the Trojan War within his famous
account of the siege of Amida, for example [55-61]). While Kelly
explains away such problems convincingly, he more importantly concludes
that veracity "is the wrong question."

The rest of the chapter supports his contention that Ammianus's
exhibition of autopsy (personal testimony) bolsters his authority as a
historian (66). Two examples support this argument, the passage in
which Ammianus claimed to have seen the Persian army fifty miles away
(18.6.20-3, 18.7.1-2) and Ammianus's account of the great tsunami of
365 CE (26.10.15-9). In the first case, it has long been pointed out
that Ammianus must have had eagle eyes to see at such a distance. Kelly
defends Ammianus by unfolding this passage as an ecphrasis, a
combination of the eyewitness account with the historian's account. The
latter consisted in a "pardonable exaggeration" (86), a reference to
Herodotus, a classic whose authority increased Ammianus's credibility
(87). In the second case (the tsunami, which restates Kelly's previous
publication on the topic[[3]]), autopsy not only lends credence to a
marvelous occurrence and marvelous association with previous
literature, it is a way for Ammianus to present his account as superior
to others (92-3). Moreover, Kelly argues that this passage's depiction
of a ship inland, which Ammianus claimed to have seen, "flung around
and yawning apart from long decay" from the giant wave, constitutes a
metaphor for the state, evoking the consequence of Julian's death for
the Roman state, against Christian writers who interpreted it as a sign
of God's anger with Julian's reversal of the triumph of the Church
(96). For Kelly, this piece is a mise en abyme representing Ammianus's
main argument, because it illustrates the incoming fall of the Empire
at Adrianople (99).

While Kelly has obviously read Ammianus very closely and presents
mostly very sound interpretations, he also occasionally seems to be
carried away and argues for readings that are a bit of a stretch (as
Kelly himself anticipates, at 77: this "may seem to some to be
over-interpretation of a few linguistic oddities.") Chapter 2 includes
two such forced readings. In the first case, Kelly claims that the
autobiographical passages occur especially in contexts of solitude and
danger. Solitude, we are asked to believe, is a "metaphor for
[Ammianus'] loneliness in the field of history. By showing himself
alone, he suggests that his historical testimony is unique" (72). Even
more questionably, Kelly argues that similarity of language between
Ammianus's account of Julian (16.1.3: fides integra rerum) and his
account of a captured Persian spy who had "revealed the full truth of
things" (18.6.16: pandit rerum integram fidem) makes the latter a
"potential historian" (75) and a historien manque/ (76). Another
passage (18.5.7, cited by Kelly at 76), seems to depict the spy as a
potential informant, the likely source of Ammianus's account of events
in the Persian Empire. Additionally, a less literary inclined reader
might well conclude that the spy "pandit rerum integram fidem" as a
result of undergoing torture precisely so that he would reveal the
truth.

Chapter 3 completes Kelly's first objective, to show "The Limits of
Biography." Here, Kelly asserts that Ammianus lived in Rome during the
late 380s, where he wrote and presented his work (109). Concerning his
controversial city of birth, Kelly insists that the debate should be
focused away from Libanius' Ep. 1063 (addressed to a fellow Antiochene
writer in Rome named Marcellinus), and argues that there is sufficient
"cumulative evidence" that Ammianus was, if not a citizen, at least a
resident of Antioch (114: oral traditions [23.5.3, 22.12.8]; 115:
first-person accounts [19.8.12, 25.10.1, 29.1-2]) both as a youth and
later in his life. Kelly refuses to take sides on the Antioch debate,
instead reminding his reader that late antique members of the elite
were mobile and would have had many homes. As with the question of
veracity, Kelly points out that efforts to determine whether or not
Ammianus was a citizen of Antioch are misguided, instead rightly
focusing on the importance of the city in the historian's work (116).
Section 2 of this chapter tackles the question of Ammianus's social
status and career, and Kelly points out that military or administrative
origins for his family would fit well with Antioch and Ammianus's
bilingualism (122). Only speculation is possible for his later years,
with one certainty: that he moved to Rome around 375-385 (130).
Although Kelly repeatedly supports Barnes's argument that Ammianus
disliked Christianity, Kelly discards the latter's suggestion that our
historian was an apostate (130: "pure and unsupported speculation.")
Regarding his social rank, Kelly concludes inevitably that he was uir
perfectissimus, the highest equestrian rank, which was conferred to a
protector domesticus upon retirement (131). In a third section devoted
to Rome, Kelly concludes that Ammianus's Roman digressions (14.6 and
28.4) were not autobiographical and that he was not expelled from Rome
(133-4). His high status as well as the support of the influential
Hypatius would militate against such a view (136). In a fourth section
on Ammianus's personal relationships, Kelly argues that political
opinions predominated over personal bias (142-54).

Chapter 4 ("Ammianus' Intertextuality") opens the second part of the
book and is the most debatable. In addition to coming too late to be
useful, as already mentioned, it is also oddly organized. After his
presentation of definitions and the historiography of his topic, Kelly
introduces a catalogue of seven types of allusion (opposites, divided
allusions, punning alterations, alterations of context, glossing,
window allusions and exemplary) which he has identified in Ammianus.
The problem is that the presentation mixes examples of each category
with a narrative that discusses it later or before the relevant
example, without obvious identification of which is related to what.
For example, Kelly. does not appear to discuss the last category,
"exemplary," while the last paragraph introduces a previously
unidentified allusion to Julian's own Caesares that Kelly presents as a
way for Ammianus to "correct" his source (213). It remains unclear what
type of allusion the latter represents.

Chapter 5 ("Sources") argues against a distinction between "source
criticism" (attempting to identify the author's source of information)
and intertextuality (a more interactive relationship with the original
text) in Ammianus through four case studies (The Obelisk; The tsunami
again; Herodian; Eutropius). A comparison between Ammianus 17.4.12-15
and ILS 736, both concerned with an obelisk of Rome, reveals that our
historian emphasized Constantine's impiety in transferring the sacred
stone to Rome whereas the (now lost) inscription lauded Constantius II
for the achievement. This is a case where Ammianus had read the
inscription but was "correcting" it by presenting his own version as
the most accurate one, an example of the historian's manipulation  of
(or what Kelly considers "engaging" with) a text considered as his
"source" by modern scholars. On the tsunami, Kelly follows Bruno
Bleckmann in dentifying Ammianus's source for this event as an
"anonymous Church historian of homoean opinions" (230).[[4]]
Considering Ammianus's otherwise willful ignorance of Christianity,
this is a surprising conclusion indeed! On the other hand, since
Ammianus appears to have been widely read, perhaps it should not
surprise us so much. If Bleckmann is right, this is a good example of
Ammianus's wide reading interests and the way he transformed his source
to present his account with "authority and literary superiority" (231).
Kelly's section on Ammianus's use of Herodian is very well argued and
convincing (232-7). The last section of this chapter on Eutropius is
intriguing, and for good reason. Kelly shows that Ammianus used the
breviarist, a source he did not need to use for events he perhaps
witnessed himself, as an "intertext" (text with which Ammianus engaged
[244]). In Kelly's view, Ammianus exploited Eutropius to cast doubt on
the circumstances of Jovian's death by using the breviarist's version
of the same event. But Ammianus also adds an allusion to Scipio
Aemilianus' death by poisoning from Cicero's Pro Milone. Kelly
concludes that "Ammianus' intention was clearly to hint, without in any
way taking responsibility for the suggestion, that Jovian was poisoned"
(245). More importantly, Eutropius' text helped Ammianus illustrate
Jovian's insignificance, with Ammianus outdoing Eutropius in brevity
and presenting his version of events as the official one, only to
allude to a somber possibility.

Chapter 6 ("The Exemplary Historian") analyzes Ammianus's abundant use
of exempla, which contrast with other types of allusions by their
visibility and the manner in which they test the reader's knowledge of
history (258). Not only did Ammianus use exempla in the typical
--intertextual -- manner, using the past to interpret the present, but
Kelly also argues that he used the present to interpret the future by
creating his own precedents (260). An interesting feature of exempla is
the unique potential they offer for making subtle attacks. Here, Kelly
presents another case that seems a bit stretched, but which yields a
very plausible and fascinating conclusion. At 21.16.11-12, Ammianus
compares Constantius II to the usurper Avidius Cassius (175 CE) and
(negatively) to Marcus Aurelius. For Kelly, the geographical map of the
events associated with this comparison can be applied to Julian's
usurpation, because Marcus was in Illyricum (like Julian) and Cassius
usurped power in Syria (where Constantius was when Julian usurped
power). According to Kelly's reading, therefore, Ammianus really meant
to compare Julian to Marcus Aurelius, by contrast with Constantius who
resembled the usurper Cassius, which incidentally legitimated Julian's
reign (280-1). A section on Gallus closes this chapter, in which Kelly
argues that Ammianus's villains do not recognize exempla when they
occur, in contrast to Julian who did (287), thereby showing their
ignorance and the paradigmatic nature of history.

Building on the notion that Ammianus created exempla to interpret the
future, the very brief Chapter 7 ("Julian's Monument") explores the
role of Julian as an exemplum for books 26-31 (296-7, 304). Kelly
argues that Ammianus tailored his portrait of Gordian III as a model
for Julian, that the battle of Strasbourg becomes the standard against
which to judge post-Julianic rulers, and that Ammianus manipulated the
chronology of contemporary events to suggest that Constantius II had
learned from Julian's exemplum (300-2). Ammianus used Julian as a
standard for later rulers, whereas the posthumous treatment of
Constantius II is limited to two "unedifying" examples, both concerning
the insignificant Jovian  (305). Valentinian failed to match up with
Julian's achievements, therefore Ammianus compared him negatively to
the good emperors (Trajan, Antoninus Pius and Marcus Aurelius) whereas
Julian had been presented as their equal (308-10). Valens' ignorance
contrasts with Julian's learning, which explains his incompetence in
handling justice (311). Adrianople, Valens' other failure, also inverts
Julian's success (313). Kelly notes "impressive [verbal] parallels"
between the two accounts (17.1.7 and 31.7.16), and underlines how
Adrianople also functions retroactively to glorify Julian's behavior at
Strasbourg, "the node of the work" (314-6). There are too few
references in this chapter, however, so it is difficult to follow the
argument in detail. Several times the reader is referred to the
author's "reckoning" of a literary occurrence, without elaboration,
footnote, table, or full discussion in the text (301, 304, 306). In a
similar fashion, the previous chapter would have gained from including
a table of exempla in an appendix (esp. 265-6).

In the end, this is a very valuable study that should appeal to a
variety of scholars: late antique historians, scholars of Ammianus,
Latinists, classicists interested in historiography and the use of
classical texts in the late Empire, students of rhetoric and
literature. Kelly's ongoing attention to modern scholarship would make
the book a good reading companion to Ammianus for graduate seminars,
and his presentation of most significant passages in translation makes
it accessible to students with little knowledge of Latin. Modern
students of Ammianus have now a third pole of attraction to consider,
after navigating between Matthews and Barnes for over a decade.


------------------
Notes:


1.   A. Cameron, "Redrawing the Map: Early Christian Territory after
Foucault [Review of Histoire de la sexualite/ 2-3]," JRS 76 (1986),
266-71; eadem, History as Text: The Writing of Ancient History (London,
1989); eadem, Christianity and the Rhetoric of Empire. The Development
of Christian Discourse (Berkeley, 1991); E.A. Clark, History, Theory,
Text: Historians and the Linguistic Turn (Cambridge, MA, 2004).

2.   J.F. Matthews, The Roman Empire of Ammianus Marcellinus (London,
1989); T.D. Barnes, Ammianus Marcellinus and the Representation of
Historical Reality (Ithaca, NY, 1998).

3.   G.A.J. Kelly, "Ammianus and the Great Tsunami," JRS 94 (2004),
141-67.

4.   B. Bleckmann, "Vom Tsunami von 365 zum Mimas-Orakel: Ammianus
Marcellinus als Zeithistoriker und die Spaetgriechische Tradition," in
J. den Boeft, et al. (eds.), Ammianus after Julian: The Reign of
Valentinian and Valens in Books 26-31 of the Res Gestae  (Leiden,
2007), 7-31.