Tuesday, December 7, 2010


Version at BMCR home site

James Howard-Johnston, Witnesses to a World Crisis: Historians and Histories of the Middle East in the Seventh Century. Oxford/New York: Oxford University Press, 2010. Pp. xxxvi, 573. ISBN 9780199208593. $199.00.

Reviewed by Anthony Kaldellis, The Ohio State University

This book is a monumental achievement of scholarship. Howard-Johnston reassesses the reliability and uncovers the sources of every text that conveys primary information about the seventh-century wars between the Romans and Persians and then the Arabs and everyone else. The importance of this period cannot be overstated nor the difficulty that historians face in coping with inadequate and understudied sources written in six difficult languages. The results here are impressive, in fact spectacular. Howard-Johnston's approach is to establish baseline narratives and dates by relying on non-Muslim sources and then to check the Islamic tradition against them, squeezing data from every source to fill in the picture. It is a painstaking, thorough, methodical, and lucid analysis. In 530 pages I found no sentence that did not clearly say what it was meant to, and one typo (at 525, the date 556 should be 656). It has eleven excellent maps. The book concludes with a hundred-page narrative of events (chapters 14-16), summarizing the findings. This book, the result of a lifetime of research, will become the standard reference for future discussions of these wars.

I will note two problems. The author does not rehearse the history of the scholarship on each issue and the views of other scholars. It is sometimes unclear whether the conclusions are his own, whether they elaborate someone else's, or represent a consensus. He has done the research, but it is laborious to retrace his steps. His sparse engagement with other scholars is mostly limited to a small circle of colleagues and friends. Second, he has a tendency to infer authors' biographies from texts and feed these inferences back into his argument about their value, though this does not much affect his chronological and unabashedly positivistic analysis. (Finally, in the schizophrenic fashion of much Byzantine scholarship, he spells most names with respect to their original languages except for Byzantine names, which are subjected to the distortion of Latinization and Anglicization. I will not follow that fashion here.)

Chapter 1, on the career and corpus of Georgios of Pisidia, contributes the least to the argument. Its concern is to date his poems for Herakleios, which is useful, and uncover his outlook. But Howard-Johnston prefers the pathos of Georgios' religious poetry (29-30, 35) and does not confront the problems that historians face in extracting history from them. We also find hypothetical biography: a gap in his output suggests that he lost favor, which leads to a search for offending verses (32-34, 291; something about mastering desire); these are used to create more biography. Chapter 2 is on the Paschal Chronicle and the Chronicle to 724. Howard-Johnston uses official documents embedded in the Paschal Chronicle as benchmarks to identify other such documents in other texts, which pays dividends throughout the book. He argues that it was the patriarch Sergios who authored the account of the siege of 626, because there is a "deafening silence" about him in it, and that it was meant as a dispatch to Herakleios (46-48). Not everyone will agree with this. Howard-Johnston does not engage with Treadgold, for whom that account was written in the circle of the master of offices Bonos, who was also in charge of the City.1 Howard-Johnston backs down from his position when he says that it was not Sergios who wrote the account but someone in his circle: "it would be hard to envisage a man with his heavy responsibilities… summoning up the intellectual energy and reserving the time to… write a lucid narrative." But examples can be given from antiquity and the Middle Ages (e.g., Caesar). From the Syriac Chronicle to 724, he extracts an earlier Chronicle to 636 written by, or in the circle of, a priest Thomas, whose brother is mentioned as a victim of an Arab raid (64-65). This chronicle contributes dates and information, which Howard-Johnston tabulates at the end (as he does, helpfully, at the end of every chapter). He uses these dates to evaluate the chronology of later sources (69). Throughout the book he carefully discuss the shape, goals, and context of surviving texts before dissecting the sections that interest him and extracting lost sources. He does so methodically without losing the reader, despite the complexity of the analysis.

Chapter 3 is on the History of Khosrov (pseudo-Sebeos), in Armenian. This was written by a churchman who, if we look for the historian-in-the-history, is identified with a "bit-part player," a bishop who appears in an episode of 653/4. "Who else would have remembered what was said on that occasion?" (73) His narrative reaches to 655. Howard-Johnston identifies the documents and sources that lay behind this history, including the Persian "Book of Lords." The virtue of pseudo-Sebeos is that he did not tamper with his sources or try to fuse them into a stylistically coherent narrative, and was "remarkably unpartisan" (79). He thereby becomes a major source against which to check other accounts, especially Islamic ones. "Had the text not survived, it would be virtually impossible to reconstruct on solid foundations the early history of Islam" (100). Chapter 4 is on the Armenian History to 682 and the Syriac Khuzistan Chronicle. The first is "disinterred" from the tenth-century History of Movses Daskhurants'i, and consists of four clusters (108-109). I was not convinced that these clusters make up a single text or that they must be attributed to one author. Here the problems of hypothetical biography become acute. Howard-Johnston must postulate multiple editorial phases to explain confusions and supplements this with conjecture: the author "may have been killed in the troubled period following the Khazar attack of 685, since his silence about it is hard to explain otherwise" (113). This hypothetical death then explains the unevenness of the dossier, verging on a circular argument. Other possibilities are relegated to a note (123 n. 33) with no reference to the scholarship. But these problems do not affect the main purpose, to extract usable information (whoever the author). This text enables him to date the battle of Qadisiyya: 6 January 638 (116-117), against which other sources are corrected. Howard-Johnston has now established his chronological framework, and other texts will provide supplementary details (starting with the Khuzistan Chronicle in this chapter) or be corrected.

Chapter 5 deals with less informative Roman sources. There is no reason to doubt that Ioannes of Antioch was from Antioch just because his work focuses on Rome and Constantinople (140). Mariev's different reconstruction of this author appeared too late for inclusion here.2 Regarding Theophylaktos, Howard-Johnston holds to a date ca. 630 for the composition of the History and sees it as triumphalist (143-146, 292; cf. 146). Yet Efthymiades will argue that 4.13.13 refers to Arab expansion, suggesting a date ca. 640 (and reads the work as pessimistic).3 Howard-Johnston discusses Theodoros Synkellos, saints' lives, the Miracles of St Demetrios, the Doctrina Jacobi, and the Maximos dossier, and in Chapter 6 surveys Strategios on Jerusalem, saints' lives, the works of Sophronios of Jerusalem, the Maronite Chronicle (which proves that the death of Ali cannot be placed after 658) (177), and John of Nikiu (favoring Coptic for the original) (185). Some "nuggets" come from these texts. Howard-Johnston omits archaeology. As he is looking mostly for dates, that is understandable, but mass graves weaken his argument (against Strategios) that the Persians showed restraint in Jerusalem (166).4

Chapter 7, which reconstructs the history of Theophilos of Edessa, is a tour de force. Theophilos covered the years 590-750 and lies behind the Chronicle to 1234 and Michael the Syrian (via Dionysios of Tel-Mahre), Theophanes the Confessor (in Greek), and the chronicle of "Agapius" (in Arabic). This much was known (though Howard-Johnston does not discuss the scholarship). What he offers is a new, methodical, and painstaking reconstruction of that source and a distillation of its information and limitations (few dates, many fanciful stories). Chapter 8 focuses on Nikephoros' Breviarium, elaborating on Mango's view that it was more a literary exercise than researched history. Mango admitted that it cannot be dated but saw it as early. Howard-Johnston engages in more speculative biography: it must have been an early work, in part because of its terminus (242-243), which enables further Quellenforschung: "it would have been quite out of character for the young Nicephorus to have engaged in the complex editorial task of integrating material from a number of distinct genres" (247). This must have been done by a single source, which Nikephoros rewrote. This theory could be correct, but the reasoning is fragile. Howard-Johnston assumes that historians wrote as close to the terminus of their works as they can be placed and that they become fuller as they go along, though he knows many examples where this did not occur (e.g., 55, 106, 192, 198; in fact, Byzantine historians avoided the recent past). One cannot base conclusions on the "character" of Nikephoros, which is unknown, and shifting the problems in his account of the early seventh century (250) onto another source (the putative second continuer of Ioannes of Antioch), a source closer to the events, does not resolve matters. Some distortions have to be attributed to Nikephoros after all (251, 256, 264: "ham-handed").

Chapter 9, on Theophanes, provides a comprehensive and useful treatment of a long-standing problem. Renewing an argument from 1994, Howard-Johnston argues that his coverage of the last phase of the Roman-Persian war was based on a lost history by Georgios of Pisidia, which was based on imperial dispatches into which Georgios added poetic interludes (286-295). To anyone familiar with late Roman historiography this sounds crazy, yet the steps by which he reaches this conclusion are solid. If there is a flaw in the reasoning, I do not see it. For the period 685-720, Theophanes relied on the lost history of Traianos (299-309), who is about to receive a longer treatment elsewhere.5 Chapter 10 surveys the nuggets in later texts, the History of the Patriarchs of Alexandria; Chronicle of Seert; Annals of Eutychios; and Iranian "Book of Lords," which is not subjected to in-depth analysis. Howard-Johnston's aloofness works against him: Pourshariati has demonstrated that there is more history to be found in the Iranian tradition and reconstructs events differently on that basis.6

Having established a baseline chronology, Howard-Johnston turns in chapters 11 and 12 to the Islamic traditions about Mohammad and the Arab conquests. His approach is somewhat paradoxical. His method in the book is to rely on non-Muslim sources and correct the Muslim tradition based on them, but when he reaches the Muslim sources he opposes the most skeptical Islamist scholarship, which denies reliability to Muslim sources. He argues that they built upon a reliable basis, though distortions entered their narrative regarding the dates of the conquest, aspects of the Prophet's life, the failures at Constantinople, and the death of Ali. He believes that the conquests were Muslim rather than Arab. Others who are qualified to evaluate this material will have to weigh in here. Chapter 13 surveys all the texts to prepare for a new narrative of the period. It seems that Howard-Johnston never met a source he didn't love, but now he indulges in a caricature of classicizing historiography (the image is old, I won't repeat it here). In the seventh century, "history became too serious to be treated as a mere branch of literature" (420); "truth mattered more than literary display" (423). Quite the contrary: one Prokopios, in exchange for most of these sources, would have given far more truth.

This review is biased in favor of historiography, avoiding Howard-Johnston's arguments about chronology and historical reconstruction, which are the meat of his book but cannot be evaluated here. As a historian, I know that he has done something quite extraordinary. No review can do justice to a book that will become a game-changer in many fields, that has explained so much that was so obscure, and has provided a new reading of one of the most crucial periods. Witnesses will facilitate discussion among disciplines that are connected but tend to break down along linguistic lines. Howard-Johnston is to be commended.


1.   W. Treadgold, The Early Byzantine Historians (New York, 2007) 342.
2.   D. Mariev, Ioannis Antiocheni Fragmenta (New York and Berlin, 2008).
3.   S. Efthymiades, 'A Historian and his Tragic Hero: A Literary Reading of Theophylact Simokatta's Ecumenical History,' in R. Macrides, ed., History as Literature in Byzantium (Ashgate 2011) 167-183, here 177-178.
4.   G. Avni, 'The Persian Conquest of Jerusalem (614 C.E.) – An Archaeological Assessment,' Bulletin of the American Schools of Oriental Research 357 (2010) 35-48.
5.   W. Treadgold, 'Trajan the Patrician, Nicephorus, and Theophanes,' in D. Bumazhnov, ed., Bibel, Byzanz und christlicher Orient (Louvain, 2010) forthcoming.
6.   P. Pourshariati, Decline and Fall of the Sasanian Empire: The Sasanian-Parthian Confederacy and the Arab Conquest of Iran (London and New York, 2008).

No comments:

Post a Comment