The two books under review are read profitably together. Each deals with third (AH)/ninth (AD) century Islamic history, but located against a wider backdrop: de la Vaissière, the long history of Transoxanian elites to the close of that century; Tor, the emergence of paramilitary organization in the medieval Near East to the late fifth/eleventh century. Each of the books considers regions – Transoxania and eastern Iran – that often are given short shrift in the analysis of early Islamic history. The two books also advance our understanding of medieval Near Eastern military history. Of particular value is their way of relating military matters to society and politics. Finally, de la Vaissière and Tor are to be commended for taking on large-scale questions. Each does so by insisting on a serious rethinking of the question at hand.
De la Vaissière initially conceived his book as a compte-rendu (p. 9) of this reviewer's The Breaking of a Thousand Swords (Albany, 2001). The book is the more valuable for taking on a wider aim: an account of the integration of the Central Asian aristocracy into the third/ninth century Abbasid Empire and the subsequent introduction of the slave military system (the subject of my work). De la Vaissière writes in a deft, spare style and frames difficult arguments nicely. He relies on a range of written sources in Chinese, Sogdian, Persian and Arabic (but, here, mostly works in translation), alongside numismatic, epigraphical and archeological evidence.
His view of the formation of the third/ninth century Turkish regiments is sharply revisionist. These regiments, it is widely held, heralded the arrival of the slave military (mamlūk) institution to Near Eastern/Islamic society, a system of recruitment, training and activity that would reach its apogee with the Mamluks of late medieval Egypt. De la Vaissière challenges the argument, one associated with David Ayalon, Daniel Pipes and Patricia Crone, mainly on two grounds: that it fails to account for Central Asian influences upon the caliphate and that it gets wrong the dating and introduction of the mamlūk system. De la Vaissière is happy to recognize the formation of said system – the question lies in making sense of its initiation in the first Abbasid period.
The author brings together much evidence on Transoxanian aristocratic society from the pre-Islamic period to the reign of al-Ma'mūn (198/813–218/833). He is correct that historians of the medieval Islamic Near East have neglected this material, and he is right to see this as a failing of my discussion of the Samarran Turkish military (pp. 12–13). De la Vaissière insists that not only were the politics of Transoxania distinct from those of Sasanid Iran, especially Khurasan, but that the region was in greater contact with China and the Turkic steppe than usually is acknowledged. Distinct as well were the terms by which Transoxanian elites were integrated into Arab/Islamic society. A still-limited Arab presence east of the Oxus left intact the military and economic underpinnings of local authority into the early Abbasid period. Such was the state of things upon the arrival of al-Ma'mūn as governor of Khurasan (192/807–201/816). These same families backed his decision to challenge his brother (al-Amīn, r. 193/809–198/813) for the caliphate.
The discussion of Central Asian patterns up to the point of al-Ma'mūn's arrival takes up half the book. Central Asian historians are better equipped than this reviewer to assess de la Vaissière's points. But, as I will argue below, de la Vaissière can be a bit free with his evidence. His stress falls on several patterns. First, as noted, the survival of a local, self-confident Transoxanian elite; second, the presence of the chākar, a comitatus-like form of military service characteristic of Sogdian society, by this point well diffused across Central Asia and eastern Iran; and, finally, long interaction with Turkic steppe society. This entailed long-term Sogdian-Turkic alliances, employment of Turkish mercenaries and chākars (de la Vaissière finds no exclusively Turkic chākar units), and, finally, trade in Turkish slaves, used, however, solely for domestic and agricultural labor. Each of the patterns mattered much as regards al-Ma'mūn's bid for supremacy and the appearance of the Samarran Turkish regiments. What follows are comments, specific and general, on the second half of the book.
De la Vaissière is right to worry (p. 141) that scholarship on the early third/ninth century relies too heavily on al-Ṭabarī's information regarding the Central Asian and Turkish presence in Abbasid Iraq. A resident of Baghdad during the period, al-Ṭabarī is an invaluable source, all the more so that he provides extended descriptions of such events as the 251/865 siege of the city by Turkish and allied forces. But much remains to be understood of his views on the Central Asian and Turkish presence. There is much to suggest that, in the case of the Turks, he thought it noxious. De la Vaissière, it needs be said, is himself happy to rely on the historian when it suits his purposes.
As regards the chākars, it would be perverse, given the weight of the evidence, not to see its presence in Abbasid Iraq in the form of the Shākirīya corps. Others, including Christopher Beckwith, have made the same argument. De la Vaissière's argument contributes by collecting much more evidence on the institution's Sogdian origins, and possible influences on Abbasid military life. The latter question remains unresolved: the book's argument is to see the influence as crucial to the introduction of the Turkish regiments. But de la Vaissière is uncertain as to how long it retained its distinctive Central Asian character (p. 189). Also, as he points out, throughout the Samarra period the Shākirīya corps stood entirely apart from the Turkish regiments (with, eventually, its own dīwān). More generally on Central Asian patterns at work in Samarra: de la Vaissière is often insistent on the flow of such influences in the one direction but seems less willing to see a Near Eastern (Iraqi) and/or Abbasid imperial imprint upon the Central Asian presence.
Finally, to de la Vaissière's argument regarding the Turkish military of Samarra, framed as the central contribution of the book. It is woven of many strands. He wants to see, for example, al-Mùtaṣim's half-Sogdian mother, Mārida, as behind the decision to rely on Turko-Sogdian regiments (pp. 169–71). The debate is an old one. De la Vaissière does not indicate that her designation as a Kufan muwallada suggests that she was turned over to slave traders at a young age – which begs the question of her ability to know enough of the Sogdian past to educate her apparently semi-literate son in such matters.
De la Vaissière is also adamant that Ashinās, a key figure of the Samarran high command, was of noble origin (citing Chinese sources, he makes the case that “Ashinās” was the name used on the Turkish steppe, one redolent of Western Turkic glory). This, he argues, is but further evidence of an Abbasid reliance on a Sogdian-Turkic elite under whose command al-Mùtaṣim placed his Turkish troops. He rests his case in part on al-Ma
ūdī's description of the wedding of Ashinās’ daughter and the son of al-Afshīn, insisting that it can only show that Ashinās was of high noble background prior to his capture and sale into Abbasid service. Compelling stuff, to be sure, but de la Vaissière might better weigh the counter-evidence: that the Murūj al-dhahab is given to “literary” embellishment (far more, at any rate, than the histories of al-Ṭabarī and al-Yàqūbī); that the Arabic sources refer to Ashinās consistently by the one name (without title or patronymic), which suggests that whatever his origins at birth, these carried little weight in Abbasid Iraq; and, finally, that it was not unusual for Abbasid clients to rise from relatively modest origins to high office.
De la Vaissière's main discussion turns on the more difficult question of the origins and service of the Turkish regiments as a whole. He argues that the Turkish soldiers were salaried fighters, acquired as adult prisoners of war, placed under the command of a Sogdian-Turkic nobility, and incorporated into Abbasid military service with the use of both carrot (salaries, housing, uniforms) and stick (modes of punishment). He sees nothing to indicate that the Turks were acquired as youngsters then shaped/trained into skilled soldiers loyal only to the caliphate, this being the hallmark of the mamlūk system. In sum, he dismisses outright the relevance of the “mamlūk paradigm” to the case of the Samarran Turks. He adds, in making his own case, that there was no further acquisition of Turkish recruits after al-Mùtaṣim, devoting much space (pp. 238–53) to prove that the Turkish presence consisted throughout the Samarra period of a “closed” community populated solely by the adult recruits, their (slave) wives, and two generations of offspring. All growth was internal; none was the result of further acquisitions. (De la Vaissière says little of the women assigned to the Turkish soldiers, apart from appearing to suggest that they were Turks, something not indicated in the unique report by al-Yàqūbī).
De la Vaissière challenges the usefulness of the mamlūk paradigm in understanding the history of the Samarran Turks: this is all to the good. But questions remain. That the sources say nothing of a training system in Samarra is to argue from silence. More to the point is that some manner of training system had to have been in place for the second and third generation Turks. All armies drill. And, on the point of the Samarran Turks as prisoners of war (slaves only at the point of their transfer to Abbasid care): yes, perhaps, but no references exist of, say, collective manumission. One also has to solve the problem of al-Ṭabarī's consistent reference to the non-elite Turks as mawālī.
And what of the standing of the Turkish rank and file? De la Vaissière says almost nothing of hierarchy and the distribution of authority within the Turkish ranks (a subject about which al-Yàqūbī and al-Ṭabarī appear to say a good deal). One takes from his discussion that the Samarran Turks were somehow as one in serving the Abbasid state, and, thus, upon the assassination of al-Mutawakkil (247/861), and during the subsequent upheaval that marks the later Samarra period, they stood shoulder to shoulder. De la Vaissière describes the upheaval – his brief comments concern the events of 251/865–866 – in terms of a fatal rift dividing the Turks (led now by a second generation command) from the Transoxanian/Sogdian and Khurasani elite (who rally to the Ṭāhirid governor of Baghdad). This is to ignore many indications that the upheaval had much to do with divisions among the Turks themselves, of a factional sort, on the one hand, and on the other, of the kind that divided commanders and ordinary troops. The indications, in other words, are of a rank and file serving on very particular terms.
De la Vaissière's discussion takes a final turn: the upheaval in question marked a failure of “the Samarra project”. The experiment of bringing together a body of soldiers acquired from the Central Asian steppe and organized under a Sogdian-Turkic command, collapsed because of the distances separating that elite from its home territories and the insurmountable challenges of sustaining the experiment in an “alien” political and fiscal environment. But, rather than jettison the idea altogether, the Abbasid state decided on a novel solution. It remained committed to the idea of a Turkish military. The Turks, after all, were the best available fighters (de la Vaissière ignores the references to Turkish cavalry fighting haplessly against the Zanj). It had to redefine, however, the terms of their service. The solution was to turn to what modern scholars know as the mamlūk system, a step he attributes to al-Muwaffaq. The evidence? A series of references, including a passage from the Pandname (attributed to Sebüktegin); the so-called “budget” of al-Mùtaḍid from Hilāl al-Ṣābi’; and al-Ma
ūdī's reference to young recruits (ghilmān ṣighār) drilling the company of Yàqūb ibn al-Layth. De la Vaissière places much weight on the first of these items – as evidence for the presence of schools specializing in the military training of slave boys although the text makes no such claim – as he does that of al-Ma
ūdī (which is more helpful).
De la Vaissière tries, in effect, to strike a balance. On the one hand, he is aware of the many questions surrounding the usefulness of the mamlūk model as regards the Samarran Turks and finds, to his credit, that it falls short of the mark. On the other hand, he seeks to preserve that same model by moving its moment of inception to the fourth/tenth century. He has spelled out, and done so with clarity, a difficult argument. But one might consider the following. First, there is an alternative explanation for the decision to continue employing Turks, this despite Samarra's mayhem, one that lies with the careers of Bughā the Elder and his male offspring, notably Mūsā ibn Bughā. The extended and, for the most part, loyal service of these offspring indicated that the project initiated in Samarra was worth preserving. This is to argue with de la Vaissière's assertion that al-Muwaffaq introduced a novel approach to service by Turkish fighters. The second suggestion is that before further embrace of the mamlūk model, historians make a more concerted effort to systematically assess and compare the cases whereby Turkish slave troops were employed by dynasties of the Islamic Near East.
The thesis of D.G. Tor's volume is clear: modern scholarship has badly misunderstood the `ayyārūn of medieval Islamic history. The first culprit was Theodor Nöldeke, whose understanding of the `ayyārūn as “brigands” was drawn too hastily from too narrow a range of late (i.e. post-fifth/eleventh) sources. The idea, of the `ayyār as a lawless character remained unchanged through the work of Lambton, Bosworth and Lapidus, mitigated only in part by Cahen, Mottahedeh and Paul, with their identification of a religious element at work in `ayyār activity and the likelihood of ties to respectable urban society. Further work (Cahen and Zakeri) has sought an Iranian/Sasanid origin of the `ayyār phenomenon, a view that Tor dismisses. The definition “brigand”, with its obvious connotations, persists, a problem worsened by a general failure to see an evolution in the term's meaning from its apparent early third/ninth century origins through the fifth/eleventh century, when new elements of “chivalry” and Sufi affiliation came into play.
In order to right this wrong, Tor turns first (pp. 27–34) to classical Arabic lexicography, but `ayyār fails to turn up in (any) Arabic texts before the early third/ninth century. Tor dubs it muwallad. Later dictionaries, projecting backward, assign it a proper “classical” origin, but the question of its actual provenance remains. More to the point, the lexicographers do little to shed light on the significance of the `ayyārūn to contemporary Near Eastern/Islamic society. This is the task that Tor takes on.
The `ayyār, as a discernible social type, emerged from a “culture of volunteer warfare for Islam” (p. 84) of the early third/ninth century developed along the Byzantine frontier and in eastern Iran, particularly Sistan. The movement sprang from a commitment to the complementary duties of al-amr bi'l-màrūf and jihād that persisted in Islamic circles following the conclusion of the major Arab/Islamic conquests. By the early Abbasid period, a new generation of ascetic warriors appeared, with Ibn al-Mubārak (d. 181/797) leading the pack. These mutaṭawwi`a were committed variously to an ideology of pious struggle; the fight against the Byzantines and “Turks” across the frontier and the Kharijites at home; and a stubborn resistance to caliphal authority. Tor sees the “volunteers” as fiercely independent, militantly Sunni, and a source of private, non-state military force.
If the connection between military volunteerism and the `ayyārūn is hazy for the early period, then matters grow clearer with the appearance of the Saffārids in Sistan. Tor devotes four chapters – the heart of the book – to the dynasty's history. Its founder, Yàqūb ibn al-Layth (d. 265/879), “the most renowned and successful of all ninth-century `ayyārs” (p. 34) was, in sum, a product of the mutaṭawwi` movement. Tor brings together much evidence to demonstrate his commitment to the ideals of asceticism, the unity of Islamic society, and jihād. At no point does Tor broach the subject of charismatic authority: given the relative brevity of the Ṣaffārid enterprise following Yàqūb's death, it seems correct to ask if his heroic example might not have mattered as much as the ideals to which he so closely clung.
She also argues (pp. 135–47) for Yàqūb’s close ties to “staunchly Sunni” `ulamā', suggesting that the emergence of a later Sunni quietism obscures a pattern of third/ninth century activism among “radical proto-Sunnis” (p. 151), the followers of Ibn Ḥanbal chief among them. These ideals would lead Yàqūb to march eventually on Baghdad but not, Tor argues, to challenge Abbasid rule per se; rather, it was to bring a more capable candidate to power (al-Muwaffaq). These same ideals would oblige the supporters of the dynasty to turn their backs on Yàqūb's brother and successor, `Amr ibn al-Layth (d. 289/902), held, as he was, to have strayed from the righteous path.
Tor concludes the long discussion of the Ṣaffārids with several points. First, that the `ayyārūn functioned as “volunteer holy warriors” (p. 229) from the early third/ninth century into the late fifth/eleventh century; second, that they constituted a “well-connected elite paramilitary force” (p. 35), with discernible and respectable ties to elements across Abbasid society, notably the urban `ulamā’; and, finally, that even late and hostile sources demonstrate an awareness of the `ayyārūn as such. The result is a much different picture of the `ayyār phenomenon than that provided by previous scholarship. Tor finishes up (Chapters 7 and 8) by drawing out two further characteristics of the `ayyār. By the fourth/tenth century, these pious warriors came to be associated with a specific code of conduct. Tor argues that the sources for the period came to see it as a code of chivalry. And, finally, in her view, the sources also make clear a close association of the `ayyārūn with Sufi values and an emerging network of Sufi associations.
Tor uses a rich body of evidence. She insists throughout that a distinction be made between pro- and anti-Ṣaffārid sources. Generally speaking, the Persian sources, “with the sole exception of the Tārīkh-ī Sīstān, are uniformly more hostile toward the Ṣaffārids than are the Arabic ones” (p. 91). But things get murky: Ibn Khallikān is hostile to Yàqūb, but he also preserves copious amounts of “positive material” from other, presumably pro-Ṣaffārid sources, evidence which begs the question of said hostility. Likewise Tor's use of the Tārīkh-ī Sīstān: she thinks much of the book but subjects it to rather little interrogation (see p. 179). Related is Tor's uneven treatment of the numerous akhbār that make up her central chapters. The one khabar is to be read for what it “reflects”, the next for what it describes but it is often unclear as to where distinctions of the sort lie. Tor appears to have enviable amounts of evidence with which to make her case, but one wishes it were sorted out more clearly.
To close, a brief comment on the value of these books to researchers. De la Vaissière's book provides maps, illustrations, a bibliography and an index. But why such a shallow index? The most significant toponyms (e.g. Samarqand, Samarra) and personal names (e.g. Ma'mūn, Mùtaṣim) are ignored, as are all topical entries: the chākars of Central Asia (and Samarra); the institution of the mamlūk; archery by Central Asian fighters; Sogdian dress (e.g. the ornate belts that figure as evidence for Central Asian influences); and so on. More egregious still is the decision, in the case of Tor's book, to leave off producing an index altogether. This is a dense, crowded book, full of references to all manner of individuals, institutions, locales, and patterns of conduct, but who would know? One indication that the book did not receive careful editing: an apparent footnote occurs in the body of the book (p. 45), and text apparently meant to appear at the end of Chapter 4 is missing (p. 157 ends in mid-sentence).
