Friday, March 07, 2008

How the Islamic World came to be   posted by Razib @ 3/07/2008 05:12:00 PM

Last summer I read When Baghdad Ruled the Muslim World: The Rise And Fall of Islam's Greatest Dynasty, and this week I finished The Great Arab Conquests: How the Spread of Islam Changed the World We Live In, from the same author. In a strange twist the first book focuses on the Abbasid Caliphate, which flourished between 750-9501, while the second covers the period up to 750, the early Right Guided Caliphs and the Umayyads. I really don't see the logic in the order of publishing here, but no doubt there is some obscure reason why they came out in the sequence they did. Both books are a mix of social and narrative history, but The Great Arab Conquests tends to focus on the latter a great deal more than the former. The social context and dynamics are packed into the initial and last chapters, with most of the filling in the middle a litany of Arab names, battles, and obscure nations. On occasion the writing gets a little sloppy here and there, as if they were in a rush to get this book to press. If you know a fair amount about the history of the Islamic world the monotonous recitation of material you could find in Wikipedia might be a bit tiresome, but if you aren't as familiar with such details the book is a worthy introduction (though I do think the density of names, places and events might be veering into diminishing returns). My own interest in this period is driven by the fact that these early years of Islam have had a major impact on the rest of human history. Not only did it give rise to one of the major civilizational blocks of the modern world and finish off the long decline of late antiquity, but the early Arab polities served as the mediators of informational and economic exchange because of their geographical parameters.

Remember, in the few decades before 750 the whole region from what is today Uzbekistan to Spain was notionally under the aegis of one political entity. After 750 the Abbasids shifted their focus to the east and North Africa and Spain immediately went their own way, but even then the geographical range of the Muslim world-state was enormous. In the 14th century Ibn Battuta journeyed from West Africa to China, and his travels were aided by the Islamic international network which was in large part derived from the early formative period. But Battuta had to negotiate the frontiers of dozens of states within the Islamic world, up until 950 the Arab caliphs would have provided a more thoroughly integrated polity across which ideas and trade flowed freely. Not only was the Islamic world expansive, but the geometry of the political domain is of importance: it had contact with all the other major world civilizations very early on, and in many cases it was through Muslims that other societies learned of each other. I think this has resulted in one obvious dynamic: everyone has an opinion about Muslims and is in conflict with them because Muslims intersect with non-Muslims. In Central Asia the Muslims came into conflict with the Chinese. In the Eastern Mediterranean and along the Black Sea they came into conflict with the Orthodox. In Spain, Italy and France Muslims came into conflict with Latin Christians. In India Muslims came into conflict with Hindus. In Africa Muslims came into conflict with Christians in Ethiopia and Nubia, and pagans in the Niger river valley. Islam also later spread to Southeast Asia; Muslims came into conflict with the Spanish in the Philippines, converted the peoples of maritime Southeast Asia, and were major influences at the courts of the potentates who ruled in Thailand and Cambodia.2 In Power and Plenty the authors make the point that Islamic civilization was the only one in contact with all the others major regions at 1000 AD, and it served role as a conduit both across time (i.e., preserving some of the Greek works) and space. Paper was transmitted from China to the West via Islam, and Arabic numerals from India to the West.

So how exactly did this world empire emerge in one generation from the deserts of Arabia? That's the main reason I read The Great Arab Conquests. Unfortunately, there weren't many new insights, though the author does touch upon the historiography and its problems for the 7th century. As many of you know there is a revisionist school of thought which contends that much of the early history of Islam was fabricated, the shape of what we know about early conquests more accurately reflects the self-perceptions of 8th century Muslims, at the height of empire, as opposed to the realities of the conquerors who defeated Byzantium and Persia. The Great Arab Conquests pulls back from this sort of extreme revisionism, and I think this is a good thing to do. Sometimes amazing secret-history narratives are very attractive, but they aren't necessarily right. The fact that Herodotus wasn't lying about the origins of the Etruscans makes me tend to think that the discounting of ancient annals and histories by modern scholars is an overreaction, just as the initial skepticism about the history depicted in the Hebrew Bible was.3 The Great Arab Conquests for example does not repeat the model of some scholars that the early Arabs were actually from the North Arabian lands, and that a Hijazi origin was a later myth.

These historical details aside, the author basically presents a model where the Arabs pulled an "inside straight" when it came to timing. Most of you might know that the Byzantines and Persians were engaged in the "World War" of their age for a generation which ended a few years before the Muslim break-out; but there were some interesting details which I think need highlighting. At the maximum extent the Sassanid Empire in the early 7th century pushed all the way to the shores opposite Constantinople and into Egypt and Yemen. After the victories of Heraclius the Persians evacuated their new conquests. But an important point is that much of Syria had been under Byzantine rule for only a few years after one generation of Persian rule when the Arab conquests began if our chronologies are correct. I think this is a critical insight; obviously a modern nationalist sensibility is totally inappropriate to project back to the 7th century. The Syrian lands which the Arabs conquered were populated by Aramaic and Greek speakers of various Christian sects as loggerheads as well as a large number of Jews and numerous Christian Arabs on the margins. But even if loyalty to the Byzantine state is something that would be plausible, remember that for most residents a greater portion of their lives had likely been spent under Persian rule. Additionally, the large Jewish minority in these regions are attested to have been sympathetic to the Persians (there were Persian Jews associated with the Sassanid armies), and some of the early annals seem to indicate that they also welcomed the Arabs. The local non-Chalcedonian Christians were also generally unsympathetic toward their Byzantine overlords, while even loyalists seemed to see the recent events more as the hand of God temporarily punishing lax and heretical Christians more than a world-changing event that would alter the path of history.

With 1,400 years of hindsight we can see that what happened in Syria in the 630s was of great significance. The Arab conquest resulted in the spread of their language from the Atlantic to the Tigris, and the long and slow decline of non-European Christianity. The individuals alive during the 630s could plausibly enter into a conversation with aged pagan philosophers who had retreated to Alexandria in their youth as well as the founders of early Islamic orthodoxy in their old age.4 But the patchy records from non-Muslims of the period suggests that proximate realities loomed much larger in the factors which affected their course of action, response and impressions. The anti-Imperial Monophysite faction which was driven from Alexandria before the Arab conquest has been depicted by some as welcoming the Muslims. This is probably an exaggeration, but it seems that the record does suggest that many Christians who were perceived as heretical by Constantinople saw recent events as simply the hand of God evening the scales of justice. Across much of Iran local rulers entered into truces with the Muslim invaders as the Sassanid royal family fled to the east. In parts of northern Iran Muslims could not enter without the expressed permission of the local non-Muslim potentate, who secured his rule through payment of tribute to the Caliph. In the 7th century Arab Muslims were a small rentier class of warriors overlain atop preexistent institutions and societies. Greek remained the bureaucratic language of the western half of the empire, while Persian dominated the east. Correspondence of non-Muslims during this period doesn't even address Islamic rule with great detail; each subculture was rather self-sufficient. John of Damascus, the last of the Church Fathers, was a minister of the Umayyad Caliph partly because the non-military aspects of the early Islamic Empire was rather simply an extension of previous dispensation.

Which brings me the point about words like "Muslim" or "Islamic." How appropriate are they? In the 10th century they seem to be pretty appropriate; we know what they mean. With the introduction of paper there was apparently a boom in writing during the 9th century, which along with the emergence of a critical mass of Arab speaking Muslims within the Muslim Empire, means that our picture of 850 is far better than that of 750. I think this explains why The Great Arab Conquests feels more patchy than When Baghdad Ruled the Muslim World, which was built upon a much more fully fleshed out world because of the closeness and continuity of the documentary sources. The Umayyad interlude is to some extent a forgettable phase in the eyes of Muslims because of its reputed non-Islamic nature. When we say that someone was Christian in 2nd century Rome, I know what that means, but when we say someone was Muslim in 7th century Damascus, I think that the term itself is less clearly defined and implies far less. Like any religion, Islamic grew organically within a particular cultural context. But its particular circumstances were important in shaping the path of that development. To be short about it the early phase was one defined by the dominance of Arab ethnicity as opposed to Muslim religion. Christian Arab tribes were allowed to join the early Islamic armies with full rights, while non-Arab converts to Islam might remain second class citizens. There were also cases of intra-Islamic conflict where non-Arabs were executed and Arabs were imposed a monetary penalty as punishment for their rebellion. Racism and ethnic chauvinism have always been implicit features of the Islamic world (e.g., see descriptions of black Africans by Arab geographers), but during the first few decades it manifested in a very explicit manner. Not only were Arabs superior to non-Arabs, but particular lineages (e.g., the Quraysh) were superior to others (e.g., recruits from Yemen). During the Umayyad period Arabs lived as a military caste in their own cities, and despite the putative mercantile background of many of the early Muslims they were most definitely rentiers who extracted tax from the vast numbers of non-Muslims. Public nudity at the court of the Umayyads or the patronage of pagan-themed mosaics by powerful Muslims seems peculiar or heterodox only when we use our point of reference as that of the Abbasids, who reflect a far different cultural sensibility and likely an Islamic religiosity far closer to contemporary norms than that of the 7th century.5

So how did we get from there to here? In the year 600 Arabic was spoken only in Arabia as well as amongst communities in the Levant and Mesopotamia which were ethnically Arab. The dominant religion in North Africa, Egypt, Syria and Iraq was Christianity, in Iran it was Zoroastrianism, while Central Asia was a melting pot of Zoroastrian, Buddhist and shamanistic influences. In the year 1000 Arabic was the dominant language of high society from Morocco to the Tigris, and Islam was the dominant religion from the Atlantic to the borders of India. What happened? As noted above it seems that the initial conquests were a function of a coincidence of historical events; the Muslim Empire was not even the first Arab one, that of Zenobia's was. But the persistence, well, that's a different thing. If you are a Muslim you could posit the hand of God, but if not, I don't think there's a very good explanation at this point why the Arabs didn't go the way of the Mongols and get absorbed. One model is that Islam as it existed in 630 was a compelling force in keeping the Arabs distinct and giving the conquered folk an identity to assimilate to, but I think though there likely was some religious motive during the initial years it seems implausible that Islam was wholly formed out of the Arabian desert (many normal aspects of Islam today, such as a sharp Shia-Sunni divide or Sharia are attested to have emerged over the centuries). The circumstantial evidence strongly suggests that the outlines of the contemporary religion developed in situ within the Islamic Empire over a few decades, not de novo in the deserts of ARabia.

By the 8th century the Christians of the Muslim of Empire were less sanguine about their new rulers. John of Damascus still referred to Muslims as a heretical sect of Christianity at this point, so a total crystallization that Islam was a new rival world religion had not developed (from the orthodox Christian perspective, whatever your orthodoxy happens to be, heresies come, and heresies go, and during the first century or so I think it is understandable why Christian intellectuals assumed that Islam was a passing fad to go the way of Arianism or Manicheanism). After all, the majority of the residents of the Islamic Empire were still Christians and Zoroastrians! It seems that Islam was a sect of the Arabs and their clients, the mawalis. For centuries prior to the rise of the Arab Empire Chalcedonians ruled over non-Chalcedonian populations in the Byzantine Empire, while in Persia Zoroastrianism had remained an ethnic religion and Mesopotamia was predominantly Christian and Jewish (Ctesiphon, the winter capital of the Sassanian Shahs did not have a fire temple, but it did have churches and synagogues). In The Rise of Western Christendom Peter Brown notes that the increase in conversions to Islam started occurring when Arabic replaced Greek & Aramaic as the dominant bureaucratic languages. Tensions between the Arab Muslims and their subjects are apparent in massive rebellions by Berbers and Copts early in the 8th century. During the second siege of Constantinople Coptic sailors crossed the lines and defected to the Byzantines; at this point the early 7th century resentments between Chalcedonians and non-Chalcedonians were distant memories. The Abassid revolt against the Umayyads is often depicted as that of mawalis against the Syrian Arab regime, but it seems likely that it also tapped into chafing of the Zoroastrian Iranian majority against Muslim rule. Iran remained predominantly non-Muslim into the 10th century, and in the last decades of the Umayyad Caliphate the Arab armies finally conquered the vassal non-Muslim states in the northern highlands.

But at this point a critical threshold had been passed. Enough mawalis now existed which identified with the Arab Islamic regime that overthrowing the new order was now more difficult to do simply by pitting the demographic advantage of the dhimmis against their Muslim overlords. That advantage was fast disappearing, and in key sectors like the professional military the Muslim stranglehold was overwhelming. The quietism of Christians during the first centuries of conquest was due to the fact that little changed in their day to day lives (there is some evidence that their tax burden initially decreased), but by the 8th century they were now fast becoming a marginalized sector who could not reverse the direction of the reaction which was inducing cultural change.

Finally, I want to include a deterministic observation. So far I've focused on the contingent events which framed the Arab conquest, and the inexplicable (from a non-Muslim perspective) robustness and evolution of the Arab sacred ideology into a world religion. But why did Iran remain non-Arab while Egypt did not? I think the explanation for the Arabicization of Fertile Crescent is pretty easy: this region was already Semitic speaking, and Aramaic and Arabic are relatively close. The switch from Aramaic to Arabic was like the switch from Akkadian and other assorted ancient Semitic dialects. But Egypt was dominated by Coptic, which is very different from Arabic. I think geography explains it, mediated through the circumstances of conquest. Remember, Egypt is pretty accessible along the axis of the Nile, and the Arab conquest was very rapid. In summary, the Arab armies defeated the Byzantines, and became the new head of the snake. In contrast, Iran was conquered piece-by-piece, and large expanses remained under non-Muslim rule for nearly a century after the initial nominal conquest. Some regions, such as Tabaristan, remained under the domination of Zoroastrian potentates until the 9th century, two centuries after conquest. There was no Coptic equivalent, unless you count the functionaries of the Christian church. Iranian high culture persisted I believe because the fragmented topography of the land made absolute and immediate rapid conquest impossible; in contrast, the Arabs managed to take Egypt as a whole without any mediators to the peasantry (the Coptic hierarchy would be horizontal alternatives in this model). By the time the Iranian elite was predominantly Muslim, Arab rule was collapsing, and with it any necessity for Arabicization. Ferdowsi is an exemplar of this non-Arab Iranian Islamic counter-culture which persisted because of the support given over the centuries of Arab domination by non-Muslim Iranian princes in the localities.6 I think this model explains to some extent the persistence of Berber dialects in the highlands of North Africa; the contrast with the Iranian example being that the Berber dialects were never independent vessels of high culture and so have been losing ground to Arabic until the present.

Balancing inevitable fixed dynamics and path dependent contingent factors is difficult. Unfortunately an understanding of humanity requires knowledge of what comes before, and there's a whole lot of messy data to comprehend, and the subtleties of historiography to keep in mind. The the early centuries of Islam are fascinating, and critical to understanding human history as a whole because of Islamic civilization's relationship with the societies which came before, as well as the many with which it interacted. And of course, the Islamic Question is one which we are having to grapple with today.

1 - The caliphate lasted quite a bit longer than 950, but beyond that point they were puppets for the most part.

2 - Most people do not know, for example, that for a short time the Khmer monarchy was under Muslim influence. But the Therevada Buddhist character of Khmer society was robust and deep-rooted enough that this only resulted in the overthrow of the Muslim faction because of popular discontent.

3 - This is not to say that the Hebrew Bible, or Herodotus, are accurate histories. But, it is one think to look at a work skeptically, but another to dismiss it entirely as totally unreliable.

4 - After the closing of the Academy of Athens the pagan non-Christian intellectual tradition did not immediately go extinct. Some relocated to redoubts of paganism such as Haran in Syria, while others relocated to the cosmopolitan metropolis of Alexandria. There are reports of pagan teachers as late as 600, so it seems entirely likely that some survived to see the fall of the East Roman provinces to the Muslim armies, spanning the culture of the ancient classical West with Islam.

5 - The standard model is that the Abassids, centered on Baghdad, were more influenced by Iranian motifs and models while the Umayyads, centered on Damascus, were shaped by the Hellenic currents dominant in their locus of power.

6 - The lowlands of southwest Iran are Arabic speaking. I think geographic accessibility likely explains this, though this region seems to have been predominantly Christian, and so quite likely to have been Semitic, at conquest in any case.