Since his earlier study on Byzantium and the Decline of Rome, [1] Walter Kaegi developed an interest for Late Roman Africa as well as for Byzantine confrontation with the Islamic conquest, as with his Byzantium and the Early Islamic Conquest. [2] It is but natural that these two perspectives combined result in the present regional study of the process which deprived the Empire from a rich province, which compensated partly for the loss of Egypt. It claims to build an "impartial comparative framework to sort through identity politics" and therefore implicitly to be above from the main trends of Byzantine Africa historiography outlined in Chapter 2: colonialism seeing the French as restorers of civilization in Barbary and successors of the Romans, anti-colonialism reactions and their deficiencies, orientalism of the English-speaking literature. "Impartiality" or "objectivity" is a worthy but unattainable ideal and Kaegi will probably be considered in the future as a representative of the American post-colonialist approach. As explained in Chapter 1, he relies on integrating Byzantine and Arabic sources for a study of the Byzantine collapse explained by the interaction of religious and military history. It represents a worthy "effort to reread Arabic, Byzantine and Latin literary texts...in the light of the latest research in several disciplines: history, religion, archaeology, numismatics and...toponymics" (9). In spite of this intention, neither archaeology--as exemplified by the absence of Anna Leone, [3] and the complete overlook of archaeological evidence for the survival of Byzantine presence in the Balearics in the eighth century--nor sigillography nor numismatics (see below) play a significant role in the narrative which is mainly about military history, interpreted in the religious context which divisiveness deeply affected the Byzantine resistance capabilities.

One welcomes, however, this documented narrative and interpretation of the varying phases of the Islamic conquest of Byzantine Africa from the battle of Sbeitla in 647 to the fall of Carthage in 695 and its aftermath. Several assets contribute to its contents: the full account taken of the latest study of Islamic sources, and notably the use of the late Yves Modéran's research or the yet unpublished PhDs of Jonathan Conant and of Mohammed Ben Abbès, [4] the integration of Byzantium's Eastern history and religious history and above all the consideration of geographic constraints, outlined in Chapter 3, for their influence on logistics and military strategies.

The author reads the collapse of Byzantine Empire in Africa primarily as a failure to harness the resources of Africa, both human and financial, stemming mainly from cultural tensions. "The challenge was one of organization, mobilization, coordination, morale and willpower. But the Byzantine failed to tap the local manpower...and financial ressources to accomplish that...The ecclesiastical politics, the crisis of the imperial legitimacy (Heraclian dynasty), fear of internal military unrest, fiscal pressure, and gaps between Greek and Latin culture, between Greeks, Latin, and autochtonous populations, and even the lengthy and tenuous communications envenomed the relations, and created impediments to any viable defense" (154). The turning point of the conflict was certainly the creation of the Muslim navy, allowing the conqueror to take control of the coast, solving much of the logistical problems of a conquest of North Africa--see for example for importance of the navy in the "second thrust" in the 660s (179-182)--and the steps taken to insure the alliance of some local Berber warriors. The tribal military structure of Early Islam was apparently more apt to integrate the Berber fighting force, like the Persian elite cavalry before, as exemplified by the 10,000 warriors from Tripolitania following 'Uqba as early as the 670s (184).

Even if he stresses the cultural dimension of the problem of the collapse of Byzantine power in Africa, Kaegi's main interest lies in military history and he focuses on the now fashionable (but nonetheless extremely important) aspects of this field of study: the strategic culture of the Byzantine empire (recently illustrated by Luttwak), intelligence gathering, the "face of battle" and the impact of logistic constraints on strategy. On this problem, see ultimately Haldon. [5] This approach was already stressed in his previous biography of Heraclius, [6] that detailed at length the emperor's campaigns in the East. But in the African context, this ambition is checked by the paucity of sources and defining them as "suboptimal at best" (125) is an understatement. Consequently he sometimes get lost in speculation and part of his argument could appears redundant (even on details, as p. 192 about the magister militum John).

This major input of the book is however tarnished by the disregard of some key-problems or sources of evidence which are either simply discarded or reduced to succinct bibliographical references. Let us focus here on seals and coins, our field of expertise. For instance pp. 116-117 where a little group of lead seals of Byzantine military officials is discussed. The first one mentions Iohannis, cubicularius, imperialis spatharius and magister militum Byzacenae; this same individual issued seals with the same legend in Greek, still unpublished but whose inscription was given by V. Laurent. [6] The second seal (mentioned on p. 240 with no reference) is that of Theoctistus, praepositus, baiulus atque magister militum per Numidiam. One can add the Greek seal of Tigranios, koubikoularios, basilikos chartoularios [kai] stratelatou Noumidias and the Latin one of Leontius, magister militum Byzacenae, [7] documents not mentioned in the book. Three of these seals belong to eunuchs placed at the head of regional armies. Their dating is of paramount importance because, if some proved to be seventh century documents, they would constitute the only documentary evidence for the organization of the Byzantine army in Africa at this time. Therefore this problem deserved more than, at best, a brief footnote explaining that major experts do not agree (p. 118 n. 6, the longest discussion of the problem). Especially striking is the absence of these bullae (simple mention of John p. 196, n. 82) from the discussion on the possible administrative and military reforms in Africa facing the Muslim onslaught. The criteria for dating these early seals are far from precise and reliable but should have been discussed, even if to ultimately dismiss their testimony. The epigraphic criterion is tricky because of the lack of well-dated material for this period, and results in two different proposals dating this group of military seals either to the late sixth century (John Nesbitt) or to the end of the seventh century (Werner Seibt). Tigranios's seal matches the epigraphy of other seventh century specimen quite clearly. Furthermore, the strange final "A" on his seal points also in the same direction: its only parallel appears on Sicilian coins from Constans II onwards. Furthermore, Seibt's argument (the use of dative which appears later than the sixth century when invocative monograms on seals compelled using this case instead of the genitive) is quite compelling even if the dative can be found already under Constans II, lowering slightly the t.p.q. Invocations to the Virgin on the seals of officials (like the above mentioned Iohannis) seems a seventh century evolution, maybe linked with the affirmation of the Mother of God as the protector of the capital city.

Leaving aside the intricacies of sigillography, we can also observe that during the seventh century, especially from the reign of Constans II, the role of the members of the cubiculum became steadily more important in the West, as exemplified by Sicily and Italy, where they wielded even military positions as exarchs of Ravenna and then strategoi of Sicily. In this context, we should not be surprised to find at least three members of this palatine institution commanding in Africa. Clearly, even in the sixth century, eunuchs happened to command armies in the West, but that became more systematic under the last Heraclids and the joint exercise of the functions of basilikos chartoularios and magister militum, see Tigranios as it is quite striking. [8] The only seals of basilikoi chartoularioi in Italy are dated to the seventh century and many more seals of seventh century members of the cubiculum are preserved in Carthage. So these seals reflect clearly the abandonment of the title of dux for regional commander, in favor of magister militum and its Greek equivalent stratelates at this time. This process stems from the use of awarding an honorary diploma of magister militum to duces in order to raise their actual rank. [9] The reform was well under way at the end of the sixth century, but the new official title for regional commander was apparently not yet magister militum by then. Under Heraclius, the commander of Numidia Peter still made use of the title doux on his lead seal, [10] the only official document we can rely on. It can be supplemented with the inscription (ca. 646) mentioning the Armenian duke John of Tigisis (70). The seals evidence is certainly open to discussion but needed consideration since they could throw an original light on the organization of the Byzantine army in Africa during the conquest. Even if they belong to the end of the sixth century and not to the seventh as we believe, the evolution in titles, which they show, remains of great importance because it has also been considered to derive from a military reform enhancing the military effectiveness of duces, [11] another point worth discussing when assessing the military might of the African province. To conclude with the military organization, Kaegi's use of the Liber pontificalis to assert the existence of a mobile African field army loyal to the Heraclids at the end of the reign of Constans II seems a little far-fetched. [12] The sentence (referred to but not cited)--Et perrexit exercitus Italiae per partes Histriae, alii per partes Campaniae, necnon et alii per partes Sardiniae Africae--seems to identify quite clearly the troops marching against the usurper (Mizizios) with units of the Italian army. That some of them operated in Africa is an important point to analyze the Byzantine strategy and Africa's own military power. An even less favorable reading would be to consider that the last mentioned Italian unit came only from Sardinia, the second term referring to the fact that Sardinia belonged administratively to Africa. Whatever the interpretation, the Liber pontificalis fails to mention an African army whose political choices could reflect African sensibilities on the matter. Another text that should have entailed a more thorough discussion is the famous iussio of Justinian II of 687. [13] This text mentions the troops of Sardinia before the African unit, an interesting hierarchical evolution that could be linked with the contemporary opening of the mint of Sardinia. The same source also mentions the unit of Septem, so it comes as a surprise to read p. 256 that Byzantium had lost this city at the very beginning of Constans II's reign. [14] This unit could have been transferred to Sardinia but it preserved its own identity and the fact could have been commented upon. The author is well aware of this text and its complexities but it would have been worthwhile to explain more clearly the basis of his interpretative choices.

The role of Africa, and more broadly speaking of the Western provinces, in the finances of the Byzantine Empire under the Heraclids is truly but too briefly acknowledged: "The Byzantines needed to defend Africa because of the important budgetary contribution of North African revenues" (155). Kaegi offers a quick survey of the various interpretations of the role of the kommerkiaroi, but fails to underscore that the evolution leading from officers controlling some frontier trade posts to major financial/fiscal operators (whatever their exact role was...) began in Africa, after 615, probably in 618/619, precisely when Egypt was conquered by the Persians, and that the first use of kommerkion on official document appears also in the West, on Sicilian seals, in the 650s. [15] Neither is there an analysis of the curious seals of chartoularioi tou sitonikou found in Carthage in the seventh century and purportedly linked to the provisioning of Constantinople, [16] a potential new proof of the importance of Africa for the Heraclids and of the diverging interests between the rich province and the central region of the Empire. Conversely, the book could have discussed the papyri attesting of levies for the Muslim fleet used in the kourson tes Afrikes, the most striking evidence of the importance of Egyptian resources for the conquest. [17]

Kaegi's use of numismatic sources is also surprising and exemplified by the choice to illustrate a majority of coins from the Constantinople mint as a gallery of imperial images and only two coins from the productive African mint in Carthage, one solidus of Justinian II dated 695/6 and one silver coin of Constans II with PAX on reverse. The author is not aware that such an inscription already featured on earlier coins of Maurice, when the religious and political context in Africa was rather troubleless and can just have been copied from this tradition. He proposes that it was destined to celebrate the restoration of political unity after the death of the usurper Gregory, a major character in the narrative. But Gregory himself is a shadowy figure: if he opposed Constans II, why did the mint of Carthage never stopped striking coins in the name of the grand-son of Heraclius, as demonstrated by the indictionnaly dated series of solidi covering the whole period assigned to Gregory's rebellion?

Furthermore, in such a protracted war as was the decade long struggle against the Muslims, the financial strength of the province was of paramount importance and the coinage is one of our best proxy to assess this important parameter. An analysis of the estimates offered by H. Ben Slimène Ben Abbès would have been useful. [18] Even if they must be used with caution, they hint at a massive increase in gold coins production in the first years of Constans II, an element to take into consideration when analyzing both the causes or context of Gregory's stance against Constantinople and the military capacity of Byzantine Africa. Later, the apparent upsurge in the production in 654/5 and 657/8 could tentatively be linked to the chronology of the restoration of imperial power in Tripolitania, as proposed by Constantin Zuckerman in his analysis of the career of Narses Kamsarakan (as in note [9]). Unpublished calculations regarding Sicily confirms the increased monetary production during these same years. Another aspect of monetary history deserving attention is the debasement process affecting Italian, Sicilian and Sardinian Imperial coinage around 700. It testifies to a huge financial effort to contain Muslim expansion. If the Muslim armies continued their expansion towards the West and ultimately to Spain it was also because they were unable to cross to Italy, because of this mobilization. This element could have been taken into account in the discussion that concludes rightly that "it would be false and excessively Eurocentric to imagine that the Muslim conquest of North Africa was a conscious part of any long-term plan on their part to invade and conquer Spain" (260). In the following consideration of the aftermath of the fall of Carthage the first Muslim coinage of North Africa would have been worth some consideration. This well known series of numismatists has not attracted much attention from historians although, together with the Arab-Byzantine coinage of the early Umayyad period in the East, [19] it is one of the most fascinating phenomenon in the history of the Muslim conquest, with gold coins sporting Muslim religious formulas in Latin script and produced with the same reduced diameter and thick flans as the seventh century issues from Carthage. It implies probably continuity in mint management, the desire to engage with the local population in their own language and monetary tradition, a well-known phenomenon of the aftermath of conquest in all periods. Of particular interest is the very brutal debasement experienced by the first series, that contrasts with the preserved quality of the last Imperial coinage in Africa. Could it be a sign that up to the end the Carthaginian mint benefited from bullion send from the East? The choice of Latin religious formula occurred here after the adoption of the fully Arabic model for the Oriental dinars which could not be imposed immediately on the African population. Here, the minimalist formulation is reminiscent of the "basic article of faith" exposed in the so-called Constitution of Medina and devised to allow a collaboration between exponent of different faiths. [20] This apparent tolerance to local sensibilities can also be looked for in the strange copper coinage sporting fish or the head of Hercules that was produced in Tangiers and Tlemcen.

To conclude what was intended as a constructive review, this book does not provide yet the much needed full assessment of Byzantine Africa and a "new Diehl" for the twenty-first century remains to be written, commendably by a team of researchers. But W. Kaegi offers a welcome insight into the cultural context of the struggle and an updated contribution of the military side of such an enterprise--see the analysis of the African drill found in the Strategikon of the Emperor Maurice (135-140)--confronting sources from both sides and North African history with that of the Byzantine East and must be thanked for it.

--------

Notes:

1. Walter Kaegi, Byzantium and the Decline of Rome (Princeton: Princeton University Press, 1968).

2. Walter Kaegi, Byzantium and the Early Islamic Conquests (Cambridge: Cambridge University Press, 1992).

3. Anna Leone, Changing Townscapes in North Africa from Late Antiquity to the Arab Conquest (Bari: Edipuglia, 2007). The otherwise extremely rich bibliography is now to be supplemented by P. Reynolds, Hispania and the Roman Mediterranean, AD 100-700: Ceramics and Trade (London: Duckworth, 2010).

4. Yves Modéran, Les Maures et l'Afrique romaine (Rome: Ecole française de Rome, 2003); Jonathan Conant, "Staying Roman: Vandals, Moors and Byzantines in Late Antique North Africa" (PhD diss., Harvard University, 2004; Mohammed Ben Abbès, "L'Afrique byzantine face à la conquête arabe: Recherche sur le VIIIe siècle en Afrique du Nord," (PhD diss., Université Paris X-Nanterre, 2004).

5. For example, J. Haldon, General Issues in the Study of Medieval Logistics: Sources, Problems and Methodologies (Leiden: Brill, 2006), 36.

6. Walter, Kaegi, Heraclius, Emperor of Byzantium (Cambridge: Cambridge University Press, 2003).

7. V. Laurent, Les sceaux byzantins du Médaillier Vatican (Città del Vaticano: Biblioteca apostolica vaticana, 1962), 91, n. 1.

8. Tigranios: J. Zographopoulos, "Die byzantinischen Bleisiegel aus Karthago," Studies in Byzantine Sigillography 9 (2006): 86. It reads Tigranios, koubikoularios, basilikos chartoularios stratou Noumidias which is to be emendated to Tigranios, koubikoularios, basilikos chartoularios [kai] stratelatou Noumidias because the basilikoi chartoularioi are linked to the cubiculum, not to provincial armies. Leontius: J. Nesbitt et N. Oikonomides, Catalogue of Byzantine Seals at Dumbarton Oaks and in the Fogg Museum of Art: Volume I, Italy, North of the Balkans, North of the Black Sea (Washington DC: Dumbarton Oaks Research Library and Collection, 1991), No. 6.2.

9. C. Zuckerman, "La haute hiérarchie militaire en Afrique byzantine," Antiquité tardive 10 (2002): 172, and remarks by N. Duval, pp. 46-48.

10. For Peter's career, C. Zuckerman, "La haute hiérarchie militaire en Afrique byzantine," 173-174, and remarks by N. Duval, pp. 48-51. V. Laurent originally published this seal in V. Laurent, "Une effigie inédite de saint Augustin sur le sceau du duc byzantin de Numidie Pierre," Cahiers de Byrsa 2 (1952): 87-93, but offered a revised analysis in V. Laurent, Les sceaux byzantins du Médaillier Vatican, 92.

11. See C. Zuckerman, "La haute hiérarchie militaire en Afrique byzantine," 171–172, and remarks by N. Duval, pp. 47-48; the above mentioned duke John offers a good example of this evolution.

12. Liber pontificalis, ed. L. Duchesne (2nd ed., Paris: E. de Boccard, 1955-1957), vol. 1, 346.

13. Acta conciliorum oecumenicorum. Series secunda: Vol. 2, Part 2: Concilium universale Constantinopolitanum tertium, ed. Rudolf Riedinger (Berlin: Walter de Gruyter, 1992), 886.

14. On the connection between Africa and Spain, see M. Vallejo Girvés, "Byzantine Spain and the African Exarchate: An Administrative Perspective," Jahrbuch d. Österr. Byzantinistik 49 (1999): 13-23.

15. Vivien Prigent, "La Sicile de Constant II: l'apport des sources sigillographiques," in La Sicile, de Byzance à l'Islam, ed. A. Nef and V. Prigent (Paris: De Boccard, 2010): 157-187.

16. Unpublished seals cited by J.-Cl. Cheynet, "Un aspect du ravitaillement de Constantinople aux Xe/XIe siècles d'après quelques sceaux d'hôrreiarioi," in Studies in byzantine sigillography, ed. N. Oikonomides (Vol. 6; Washington DC: Dumbarton Oaks Research Library and Collection, 1999), 1-26, here p. 5.

17. See for example, N. Gonis, "Arabs, Monks, and Taxes: Notes on Documents from Deir el-Bala'izah," Zeitschrift für Papyrologie und Epigraphik 148 (2004), 213-224, especially pp. 218-219, offering a new edition of an order of payment in favor of sailors, with relative bibliography.

18. H. Ben Slimène Ben Abbès, "La production de la monnaie d'or en Afrique byzantine au VIIe siècle: étude statistique," in L'Africa romana: le ricchezze dell'Africa: risorse, produzioni, scambi: atti del XVII Convegno di studio, Sevilla, 14-17 dicembre 2006, ed. J. González, P. Ruggeri, C. Vismara, R. Zucca (4 vols., Roma : Carocci editore, 2008), 1151-1164.

19. For the state of the art, see C. Foss, Arab-Byzantine Coins: An Introduction, with a Catalogue of the Dumbarton Oaks Collection (Washington DC: Washington DC: Dumbarton Oaks Research Library and Collection, 2008).

20. See R. Hoyland, "New Documentary Texts and the Early Islamic State," Bulletin of the School of Oriental and African Studies 69/3 (2006): 395-416, 409-410.