How are we to talk, and write, about the origins and rise of Islam? It was a highly complex phenomenon that encompassed at one and the same time a military conquest that seized vast territories from the two “great powers” of the early seventh century, Sasanid Persia and the Later Roman (or Byzantine) Empire – completely destroying one, and seriously truncating the other in the process; the rapid crystallization of a new state originating in the Ḥijāz region of western Arabia, an area that had hitherto had no tradition of statecraft; the rise to power and prominence of people who until then had been only dimly visible in the historical record; the appearance of a new religion, Islam, and its new scripture, the Quran; and the emergence of Arabic (previously known only from inscriptions) as a new literary language that gradually spread through much of the Near East at the expense of Aramaic, Greek, Coptic, Latin and other languages that had long held sway. It involved the restructuring of political borders and domains in the Near East and, for over a century, the redirection of tax revenues to new areas, sending them now not to Constantinople and Ctesiphon or Iran, but rather to Medina or Damascus. Moreover, since the conquering elites hailed mainly from the towns of the Ḥijāz and incorporated the pastoral nomads of Arabia into their forces, it involved a hundred-year-long inversion of the usual relationship of power between the settled peoples and hitherto autonomous nomadic populations of Arabia and adjacent areas. It was, in short, a historical phenomenon that made sudden and dramatic changes in the world into which it came, one that shaped the course of the future decisively, with consequences that endure right up until today.Footnote ²

Such complex historical processes are by their nature rich and challenging to analyse, and a marked diversity of opinion on many aspects of them is to be expected, and indeed to be welcomed as indicative of a healthy climate of scholarly debate. But the scholarly and popular discussion of Islam's origins has long been hampered – even crippled – by the use of deeply entrenched conventional terminologies that are inappropriate to the historical realities we seek to understand. It is not just that we use “inappropriate” names for various phenomena; more serious is the fact that these engrained terminological habits inhibit our ability to conceptualize clearly the true nature of the phenomena associated with Islam's origins. Yet these misleading terminologies have seldom been challenged, and a serious discussion of them is overdue. The present essay aims to stimulate such a discussion.

The problem of what terminology we, as modern scholars, should use to describe Islam's origins is complicated by two factors. The first we might call the challenge of terminological conservatism or inertia. It affects both the seventh-century sources we must rely on to discuss Islam's origins, and the way we ourselves describe what we think was happening. In general, the appearance of any novel social or political group poses a challenge to those outsiders who first come into contact with and make mention of it. Precisely because the group is new, these first outside witnesses lack a clear terminology to describe and characterize it, or sometimes even to name it. In such cases, outside observers are wont to refer to it, or to its members, using established terms applied to categories to which members of the new group had formerly been known to belong; indeed, lacking a ready vocabulary, it is almost inevitable that these outside observers will fall back on familiar, pre-existing terminologies.Footnote ³ It may take considerable time for these “outsiders” to acquire a clearer sense of what the new group is all about, and to develop a more precise vocabulary to describe it; and even when they do, the pre-existing terminologies by which some members of the new group were first described, although inappropriate, may long continue to be applied, particularly if these pre-existing terms carry some form of ideological message (often pejorative) appealing to the writers.Footnote ⁴ A good example is found in the Syriac Apocalypse of Pseudo-Ephrem, composed most likely between 660 and 690, which describes the Sasanian empire as “Assyria”, even though the Assyrians had died out centuries before.Footnote ⁵

The second element that complicates modern attempts to describe Islam's origins is the fact that the movement begun by Muḥammad, like many new movements, required a considerable time to develop clear boundaries and to define itself.Footnote ⁶ It began, to be sure, with a strong religious impetus – as a movement of strict monotheism, insisting on pious behaviour and constant mindfulness of God, and perhaps filled with the conviction that the Last Judgement was soon to come; but these were ideas that had long histories in the Near East and by the seventh century ce were widespread there. Under these circumstances, it is hardly surprising that we should find considerable evidence that in its earliest years, the new movement included individuals belonging to older monotheistic traditions, such as Jews and Christians, within its new community. This initial lack of sharp definitional clarity obviously creates difficulties for modern scholars who wish to characterize what we traditionally call the “Islamic community” in its formative stage.Footnote ⁷ We shall discuss the definitional problem more fully below.

* * *

The easiest place to focus our reconsideration, perhaps, is with the expansion outside Arabia of the community established by Muḥammad. This process is generally called “the conquests”, a rubric that has long been used in works of Western scholarship to cover many of the facets associated with Islam's origins, including broad questions of statecraft and social and institutional change. These works of Western scholarship, going back almost a century and a half, enshrine in their titles the two poles in a debate over how best to characterize the conquests – whether as “Arab” or as “Islamic”. (Indeed, we may wish to question also whether the very term “conquest” is entirely appropriate to describe the events of Islam's origins, as it may lead us to overestimate the role of military action, and obscure in our minds the role of co-operation and collaboration between “conquerors” and “conquered”, but we can leave this issue aside here.Footnote ⁸)

In fact, both terms – “the Arab conquests” and “the Islamic conquests” – while used occasionally even in the eighteenth century, did not come to dominate discussion of the conquests until relatively recently; and both, as we shall see, are deeply problematic. Before 1900, European writings on what we today usually call the “rise of Islam” tended to use the terms “Mahometan” or “Saracen” to describe such events and institutions, and the people participating in them. Among these works was Simon Ockley's famous and highly influential The Conquest of Syria, Persia, and Ægypt, by the Saracens… (1708).Footnote ⁹ The term “Saracen” continued occasionally to be used even into the twentieth century – by that time, it seems, mostly by students of art and military history.Footnote ¹⁰ “Mahometan” or “Mohammedan” was also a fairly frequent descriptor until around 1900; among the latest titles to employ this term was H.A.R. Gibb's classic overview, Mohammedanism (1949).Footnote ¹¹ Dozens of works with the title Mohammedanism preceded it in the nineteenth and early twentieth centuries, by authors as varied as Reginald Bosworth Smith (Schoolmaster at Harrow, 1875), Sigismund W. Koelle (German missionary, 1889), D.S. Margoliouth (Laudian Professor of Arabic at Oxford, 1911), C. Snouck Hurgronje (Professor at the Leiden School for Colonial Civil Servants and later Leiden University, 1916), and many others.

The adjectives “Mahometan” (or “Muhammadan”) and “Saracen” gradually fell out of favour after about 1900, however, and both sound quaint and outmoded to modern ears. They were gradually replaced in book titles by the terms “Arab conquests” and “Islamic conquests” (or “Muslim conquests”), which now dominate our discourse on the expansion associated with the rise of Islam. Not a few recent works hedge on this terminology, and use either the hybrid form “Arab-Islamic conquests”, or oscillate between using “Arab conquests” and “Islamic (or Muslim) conquests” from paragraph to paragraph and even from sentence to sentence, treating them essentially as synonyms.Footnote ¹² Given the dominance of these two terms in modern discourse on Islam's origins, it is worth devoting a few paragraphs to tracing the respective trajectories of each, before considering their suitability.

* * *

Of the two terms, “Islamic conquests” is the more recent, evidently first coming into wide circulation in Western languages only around the mid-twentieth century.Footnote ¹³ It has long been recognized that the Islamic narrative sources of the eighth–tenth centuries (and later) presented the origins of Islam, including the conquests, in an idealized light. One of the main genres of this Islamic historiographical tradition, which crystallized more than a century after the events of the conquests themselves, was the theme of futūḥ, which offers a salvation-historical vision of the expansion as an expression of God's will. The futūḥ literature saw the events of the expansion, depicted as resounding victories by the Muslims against overwhelming odds, as a quasi-miraculous process and evidence of God's favour for Muḥammad and the community of his followers.Footnote ¹⁴ It has thus recently been argued, and rightly so, that use of the term “Islamic conquests” unwittingly reflects these idealizing tendencies of the later Islamic sources, and should therefore be avoided because it tends to overemphasize the religious component of the expansion, minimizing or obscuring in the process the mundane factors – the crass lust for power, plunder, and property – that, rather than religious ideals, very likely motivated many of the conquerors.Footnote ¹⁵

There is also, however, another, and even stronger, reason not to refer to the expansion of the new community as the “Islamic conquests”: it is because seventh-century documents – that is, sources contemporary with the events of the expansion itself – do not describe them in this way; nor do they call the people who led the conquests “Muslims”. Only later Islamic sources speak in this way. Let us review briefly the evidence from the seventh century ce, the century in which the conquests took place, starting with the Arabic sources and then moving on to consider evidence from contemporary non-Arabic (and non-Muslim) sources.

The seventh-century writings produced by the conquerors in Arabic do employ the words islām and muslim, but neither word figures prominently until after the seventh century. More importantly, as we shall see, in the seventh century Arabic sources islām and muslim do not yet mean “Islam” as a reified religion, as it would later come to be understood, or “a Muslim” as an adherent of this faith.

One text in which the words islām and muslim are used is the Quran, one of the earliest extant sources hailing from the new community established by Muḥammad.Footnote ¹⁶ The date of the crystallization of the Quran as a set (or nearly set) text has been the subject of intense debate since the late 1970s, but it now seems fairly clear that the basic text – at least its “skeleton” of consonants – was largely fixed by the end of the seventh century ce.Footnote ¹⁷ This being so, the Quran can thus be used as evidence for how various terms, including muslim and islām, were used in the seventh century by the community established by Muḥammad. What is most striking is that the words muslim and islām are used relatively infrequently in the Quran, and in particular are not used as terms of address for the Quran's original audience. Rather, the Quran consistently addresses its audience as mu'minūn, “believers”, a term that is also overwhelmingly more frequent in the Quran than muslim. Footnote ¹⁸ Moreover, the meaning of muslim in the Quran is quite consistently “one who submits to God's will”.Footnote ¹⁹ It does not yet have the meaning of an adherent of a particular religious confession (i.e. a “Muslim”, as we say today), a meaning it would acquire only later. This can be seen most clearly in passages such as Q. 3: 67, in which the word muslim is applied (as an adjective) to the patriarch Abraham, who is called ḥanīfan musliman, “a monotheist who has submitted himself to God”. This meaning of muslim in the Quran may be a continuation of pre-Islamic usage, since in a number of pre-Islamic Arabian inscriptions we find MSLM used as a personal name, presumably meaning “someone committed to God” (or perhaps to some pagan deity).Footnote ²⁰

Turning to the Quran's use of the word islām, the first thing to note is that it occurs only a total of eight times in the whole text; and in five of these cases, islām is clearly a verbal noun referring to an individual's act of submission to God's will. It thus conforms to the text's use of the word muslim, “one who submits to God's will”.Footnote ²¹ The only exceptions are three verses (Q. 3: 19, 3: 85, and 5: 3) in which the Quran speaks of “the religion of Islam” (dīn al-islām) apparently in a reified manner, but these three verses appear to be anomalous and may be alterations of, or interpolations into, the original Quran text made at a slightly later date, perhaps around 700 ce, when what we now know as Islam was beginning to coalesce as a distinct religious confession from the original Believers’ movement.Footnote ²² In any case, even if we take these three verses as original, the marginal status of the term islām in the Quran, particularly in this meaning, seems quite clear given its rarity. Far more central to Quranic discourse, both in frequency and in import, as already noted, are terms related to the concept of mu'min, “believer”.Footnote ²³ It is for this reason that it seems most appropriate to call the community founded by Muḥammad the “Believers’ movement”.

Another very early source from Muḥammad's community is the so-called “Constitution of Medina” (alternatively called “the umma document”, “the ṣaḥīfa”, or “kitāb al-Madīna”). Its text survives only in later narrative sources of the eighth and ninth centuries ce, but it occurs in almost identical form in several different later sources. Moreover, its style is so archaic, and its content so at odds with the norms of the later Islamic community, that all scholars, even the most sceptical, agree that it must represent a transcription of an actual early document – an agreement drawn up between Muḥammad and his Meccan followers on the one hand, and the people of Yathrib (later Medina) on the other, establishing the political order of the new community.Footnote ²⁴ This document, like the Quran, uses both the terms mu'min (“believer”) and muslim, but again, it is clear that mu'min is the dominant term for the new community as a whole. Muslim is used only twice, and its use is consistent with what we see in the Quran, where it means “one who submits himself to God”.

Other seventh-century Arabic documents produced by the new community reinforce the impression gained from the Quran and the “Constitution of Medina” that these people thought of themselves first and foremost as a community of Believers (mu'minūn). Roughly two dozen inscriptions and other early documents from the seventh century that refer to the leader of the community exist, and they invariably refer to him as amīr al-mu'minīn, “Commander of the Believers”.Footnote ²⁵ Most of the earliest dated documents produced by the conquerors provide no era, just the year, but those few that provide an era give dates in the form sanat X sanat qaḍā’ al-mu'minīn, “year X, year of the jurisdiction of the Believers”.Footnote ²⁶

Just as strikingly, we know of no surviving inscription, coin, or papyrus document produced by the conquerors in the period before 690 ce that refers to islām or in which the members of the community refer to themselves as muslims.Footnote ²⁷ The first document to do so is part of the interior inscriptions of the Dome of the Rock in Jerusalem, dated to 71/691–2. These inscriptions focus mainly on asserting God's oneness and rejecting the idea that Jesus is God or God's son; but they also include the phrase inna al-dīn ‘ind allāh al-islām, “In truth, religion with God is al-islām”.Footnote ²⁸ It thus seems that the word islām first begins to be used in the sense of a distinct religious confession – Islam – only in the last decade of the seventh century, when the first wave of conquests was already over. The next dated instance occurs in a graffito from Wādī al-Gharra in eastern Jordan, dated to 107/726, which mentions rabb al-muslimīn, “Lord of the Muslims” (or “of those who submit themselves to God”?).Footnote ²⁹

It is possible, of course, given the dearth of surviving documentation from the early community of Believers, that the early conquerors actually did call themselves “Muslims”, and did think of themselves as advancing a distinct new religion, “Islam”, but that by chance no evidence has survived. While this is possible in principle, the silence of the Arabic sources on this point seems more than simply an accident in view of the fact that the non-Arabic sources from the seventh century also do not describe the conquerors in ways that suggest that they called themselves “Muslims”. Let us turn to consider these non-Arabic sources.Footnote ³⁰

When the community first established by Muḥammad expanded outside Arabia in the middle and later decades of the seventh century ce, its members came into contact with Christians, Jews, Zoroastrians and others, who sometimes left reports about their first contact with and impressions of the Believers. These reports were set down in the languages favoured by these communities: Coptic, Armenian, and especially Greek and Syriac. (Contemporary Middle Persian and Hebrew or Jewish Aramaic sources, from the Zoroastrian and Jewish communities, are unfortunately virtually non-existent for the seventh century.)

Among the seventh-century non-Arabic sources, those written in Greek employ several different terms to refer to the conquerors: arabes, sarakenoi, Agarenoi and Magaritai (the latter two words being renderings of the Arabic word muhājirūn, “emigrants” and probably meaning something like “settler-soldiers” in the conquered territories).Footnote ³¹ Seventh-century Syriac texts refer to the invaders as mhaggrāyē (another rendering of Arabic muhājirūn) and ṭayyāyē (a complicated word that will be further discussed below,Footnote ³² but usually meaning “invaders from the desert”). In other words, early sources written in languages of the communities overtaken by the new movement do not describe them using any word that appears to be a transcription of the Arabic words muslim or islām. This suggests that the conquerors were not known as “Muslims” in the seventh century by the people they came to rule, and probably did not yet refer to themselves in this way.

The implications of all this should by now be clear: as historians, we should no longer speak of “Islam” or “Muslims” as being present in the early days of the community, at least until shortly before the year 700 ce, when the term islām first begins to become visible in documents as the name of a distinct monotheistic confession.Footnote ³³ To speak simply of “Islam” as being present in the earliest years of the community, or to speak of the early Believers as “Muslims”, is blatantly anachronistic and misleading, akin to speaking of Jesus and his apostles as being already “Christians” or adhering to “Christianity”.Footnote ³⁴ Doing so conveys the false impression that the early Believers’ movement was already the same as the later Islam that emerged from it, even though the Believers’ movement seems to have been open to pious Jews, Christians and other monotheists, and perhaps even to others, in ways that Islam ultimately was not.Footnote ³⁵ The transition from a community that identified itself as Believers, mu'minūn, to one of Muslims, muslimūn, was essentially a process of drawing definitive boundaries separating those who accepted the Quran as God's word, and Muḥammad as God's prophet, from all other monotheists, in particular from Christians and Jews.Footnote ³⁶

Similarly we cannot, as responsible historians, continue to speak of the “Islamic conquests”, and should consider the titles of all earlier works that do soFootnote ³⁷ as reflections of an earlier, outmoded form of expression and conceptualization that – like the older terms “Saracen” and “Mahometan” – we have now outgrown.

* * *

If speaking of the “Islamic conquests” is no longer acceptable for the reasons just given, what about the term “Arab conquests”, which is the other phrase usually used to refer to these events? And can we speak of the conquerors simply as “Arabs”?

The term “Arab conquests” seems first to have come into use in the middle of the nineteenth century,Footnote ³⁸ and became the dominant term in the early twentieth century – not coincidentally, I think, just as the modern concept of Arab nationalism was beginning to take root among the people we (therefore) call Arabs today, and a time when nationalist conceptions (including concepts of “race”) dominated Western thought in general. The earlier view that the conquests were an ethnic “Arab” movement is nicely captured in the opening paragraph of the Dutch Orientalist M.J. de Goeje's study of the conquest of Syria, originally published in 1864:

This passage, with its melodramatic reference to the imagined agonies of a “homeland, groaning under a foreign yoke”, and the implication that the conquest was virtually a liberation, is redolent of the romantic pathos of nationalist rhetoric. Classic nationalist thought saw all of history as shaped by the deep and often silent or unarticulated movements of “nations” – races, peoples – and saw “nations” as enduring ontological realities, each having distinctive physical, intellectual and moral qualities, rather than as constructed social and political communities that formed in particular historical circumstances, evolved over time, and eventually dissipated.Footnote ⁴⁰ To nationalist thinkers, nations were “hard”, unchanging and unchangeable elements of the natural world, akin to a particular species of animal or kind of mineral. Nations in their view were not mere associations of individuals bound together by malleable culture and transitory ideas, and there could be no question of “fuzzy edges”; one either was of a particular nation or one was not – unless, of course, one was a “half-breed”, of mixed parentage, the product of “miscegenation”.

The nationalist conceptualization of history has proven very durable, and was still current in scholarship into the mid-twentieth century (and, indeed, is still found in some circles even today): consider this quote, appearing in 1950 in the widely used survey by the eminent historian Bernard Lewis, The Arabs in History: “Originally the great conquests were an expansion not of Islam but of the Arab nation, driven by the pressure of over-population in its native peninsula to seek an outlet in the neighboring countries. It is one of the series of migrations which carried the Semites time and again into the Fertile Crescent and beyond”.Footnote ⁴¹ This is much more measured in tone than de Goeje, but it still has the unmistakeable stamp of classic nationalist thought in its reference to “the Arab nation”. The image of successive waves of Semitic migrants coming over millennia from Arabia to submerge the surrounding lands of the Near East – an image my esteemed teacher, John H. Marks, once sardonically dismissed as “this bubbling well of Semites, somewhere deep in the middle of Arabia”Footnote ⁴² – was part of the furniture of nationalist ideology in its perceptions of the Near Eastern region and its history.

What the above-quoted passages and many others like themFootnote ⁴³ do, however – besides attributing the impetus for the conquests to the movement of the “Arab nation” – is to downplay, or sometimes eliminate entirely, religion as a motive force for the expansion. Sometimes religion (“Islam”) sneaks in by being linked to “Arab” identity, as reflected in the last phrase of the quote from de Goeje, above, but in such cases the implication is that the religious impetus was really secondary, merely a cover for the “real” force behind the expansion, which was the working-out in history of the “national will” of the “Arab nation”, striving for self-realization.

Such overtly racialist-nationalist conceptualizations of history fell out of favour in the West following the Second World War, the horrific atrocities of which made the danger of blind acceptance of race-based views distressingly clear.Footnote ⁴⁴ In spite of this, however, scholars and others have continued to speak of the “Arab conquests”, reifying the existence of an “Arab nation” back into the seventh century and beyond. By the middle of the twentieth century, the term “Arab conquests” had become commonplace, and a quick review suggests that it has actually surged in popularity in titles of books published since about 2000.Footnote ⁴⁵ The fact that the term “Arab conquest” seems to be increasing its hold in popular usage at the expense of “Islamic conquest” suggests that people today are more inclined to view the conquest as the product not so much of religious as of pragmatic, perhaps “ethnic”, factors.

But, just how appropriate is the term “Arab conquest” as a descriptor for the expansion of the community established by Muḥammad? Its use could only be justified in one of two ways. The first would be if the conquerors referred to themselves as Arabs, or described their movement as an Arab movement – what social theorists would call an “emic” description. In this case, they might be aptly classified at least as what Smith terms an ethnie, a named human community sharing myths of common ancestry, common history, and common culture (such as language).Footnote ⁴⁶ The second basis on which we might refer to the “Arab conquest” would be to assume that, even if the conquerors did not call themselves Arabs, they nevertheless belonged to a category which we today identify as “Arabs”, and which exists ontologically, even in the absence of the term – an “etic”, rather than an “emic” concept.Footnote ⁴⁷ The latter notion, however, is precisely an echo of classic nationalist thought, the assumption that the “Arabs” are a “nation”, a natural and eternal division of humankind. Since, however, all national identities are historical constructs, not ontological realities,Footnote ⁴⁸ this second assumption must be rejected. Use of the term “Arab conquest”, then, can only be valid if the conquerors described themselves as Arabs or their movement as an Arab movement. But, as we shall see, there is virtually no evidence to support this claim, and much that goes against it.Footnote ⁴⁹

Let us begin with the Arabic sources. In the conquerors’ own early inscriptions and other writings from the seventh century, which are in the Arabic language, it is striking that they never refer to themselves in them as “Arabs”. As we have seen in our earlier discussion, the conquerors seem to conceive of themselves primarily in religious terms, as Believers (mu'minūn). They refer to their leader as “commander of the Believers” (amīr al-mu'minīn) and to their rule as “the jurisdiction of the Believers” (qaḍā’ al-mu'minīn); any reference to an Arab identity is strikingly absent.

In the Quran, too, as we have seen, the term “Believers” is overwhelmingly the term used for those to whom the preaching is addressed, its original target audience. The Quran does, however, describe itself in several passages as qur’ānan ‘arabiyyan, usually translated as “an Arabic Quran” (or “Arabic recitation”), which might be taken to imply an “Arab” identity, if only by suggesting that the text was tapping into a sense of common identity shared by those who spoke what we today call the Arabic language, and was claiming to be a recitation (qur’ān) in that language. But it is not clear that “an Arabic recitation” is the correct translation of the phrase qur’ānan ‘arabiyyan.Footnote ⁵⁰ The root ʿ-r-b from which the word ʿarabī is derived carries the meaning “to be or make [linguistically] intelligible”, as seen in the related verb aʿraba, “to express clearly”.Footnote ⁵¹ We might, then, render the phrase qur’ānan ʿarabiyyan not as “an Arabic Quran”, but rather as “a clear Quran” or, more colloquially, “a recitation in plain speech”.Footnote ⁵² A couple of other verses also contrast lisān ʿarabī, “an ʿarabī tongue” or “language”, with the word aʿjamī (Q. 16: 103; 41: 44), which basically means something unclear. The notion that the Quranic word ʿarabī may refer simply to clarity of expression and may not actually refer to a group of people known as “Arabs” or to their speech is reinforced by another Quranic verse (Q. 13: 37), which states wa-kadhālika anzalnā-hu ḥukman ‘arabiyyan…, “And so We revealed it as an ʿarabī judgement …”. It seems apparent that in this verse, ḥukman ‘arabiyyan cannot mean “an Arab[ic] judgement”, which is nonsensical (even though most translators insist on rendering it so), but means rather a judgement the import of which is clear, an unequivocal judgement. All considered, the Quran's use of the word ʿarabī seems more likely to be linked to notions of clarity of expression than to ideas of ethnic identity.

We might even question the degree to which a single, common Arabic language was available in the early seventh century, rather than a plethora of related Arabic dialects, some of them possibly mutually unintelligible;Footnote ⁵³ what we call “Classical Arabic” was constructed in the centuries after the conquest as a kind of lingua franca, largely a hybrid of dialects of North Arabia and Quranic usage itself. It has never been anyone's mother tongue.

Moreover, not once does the Quran address Muḥammad's community or its members as “Arabs”, or refer to them even in passing using such terms. In a few verses, the Quran mentions the aʿrāb (a plural meaning “pastoral nomads”), but those passages convey a pejorative sense, because the way of life of these non-settled people hindered them from the regular observance of prayer and other obligatory rituals. This makes it seem even less likely that “Arab” was a word that the community founded by Muhammad would have wanted to apply to itself.

Turning to the non-Arabic sources from the seventh century that refer to the conquerors and their conquests, we find a similar picture.

In Syriac sources from the seventh century, the conquerors, as we have seen, are termed either mhaggrāyē or ṭayyāyē, neither of which can be taken as evidence that the conquerors called themselves “Arabs”. The Syriac word ṭayyāyē was derived from the name of the nomadic tribe of Ṭayyi’, which migrated northwards from southern Arabia to the fringes of the Fertile Crescent in the early centuries ce. The tribe's name came to be adopted by Syriac-writing authors as the general term for all camel-herding nomads from deep in the steppes east and south of Syria and Mesopotamia.Footnote ⁵⁴ These ṭayyāyē or “invading desert people” are mentioned with some frequency in the Syriac literature of the fourth, fifth, and sixth centuries as a disruptive element because of their habit of raiding the settled districts of Syria and Mesopotamia. Syriac ṭayyāyē is thus a classic “outsiders’ term”, a general designation that Syriac authors came to apply to many different groups (for example, to nomads of different tribes), not just to members of the tribe of Ṭayyi’. These authors were either not aware of, or did not care to notice or comment on, the distinctions between different tribes or groups of people whom they lumped together under the term ṭayyāyē.Footnote ⁵⁵ What mattered, rather, was the perception that these ṭayyāyē came from a world beyond, and different from, the world familiar to the Syriac authors. That familiar world encompassed especially the towns, villages, monasteries, and settled agricultural districts and adjacent steppes of Mesopotamia and Syria-Palestine, where various Aramaic dialects akin to Syriac were spoken, as well as the Greek-speaking minorities in many of the larger towns and monasteries.

By describing the conquering Believers as ṭayyāyē, then, the seventh-century Syriac authors were using a traditional term for intruders coming from the desert into their familiar world. Nothing about this word suggests that these newcomers called themselves “Arabs”: certainly the word ṭayyāyē itself cannot be seen as an effort to mimic such a word in Syriac. Yet modern scholars of Syriac literature with distressing frequency translate the term ṭayyāyē simply as “Arabs”. By doing so, they suggest to the reader that these desert people may have shared a common identity as “Arabs”. But this is anachronistically to inject a nationalist conception of modern vintage into these Late Antique texts.Footnote ⁵⁶ The eminent French Syriacist J.-B. Chabot (1860–1948), who edited and translated (usually into Latin) a large number of basic Syriac texts in the Corpus Scriptorum Christianorum Orientalium (CSCO) series, may have started this tradition, because he routinely made this equivalence in his many Latin translations of the Syriac texts. That Chabot should have done so is hardly surprising, given that he was formed as a scholar in the late nineteenth century, the heyday of European nationalist thought, and no doubt subscribed – unconsciously, and perhaps consciously – to the notion that the “Arabs” were a “nation” of eternal standing, who had “always been there”. Not all scholars of Syriac followed this pattern, however; Chabot's contemporary, the Syriacist E. W. Brooks (1863–1955), for example, consistently transliterates the word ṭayyāyē in his Latin renderings in the CSCO series, rather than translating it.Footnote ⁵⁷ Nevertheless, many scholars today still insist on rendering ṭayyāyē simply as “Arabs”,Footnote ⁵⁸ in the process conveying the notion, whether unwittingly or intentionally, that the early Believers may have thought of themselves as “Arabs”.

The confusion generated by rendering Syriac ṭayyāyē as “Arabs” is highlighted when we consider another word used in Syriac, the word ʿarab or ‘arbāyā (plural ʿarbāyē) – a word clearly cognate with the Arabic word ʿarab. In Syriac texts, this word refers to the semi-nomadic peoples who inhabited the steppe regions (sometimes also called ʿArab) that were adjacent to the towns and settled districts of Syria and Mesopotamia. These semi-nomadic peoples had close economic, social and political ties with the towns of Syria and Mesopotamia; unlike the alien ṭayyāyē, the ʿarbāyē were not periodic interlopers, but rather a regular and integral part of the social and cultural world to which the Syriac authors belonged. The linkage of these populations of ʿarbāyē to the settled areas near which they lived is reflected in the appointment by cities like Edessa of a functionary called the shallīṭā d-ʿarab (in Greek Arabarchos) the “ruler of the ʿarab”, whose job was to monitor and probably to defend the semi-nomadic population that had regular ties to a city.Footnote ⁵⁹ A passage in the sixth-century Ecclesiastical Chronicle of Zacharias Rhetor notes that in the pre-Islamic period, the fortress of Dara was built to protect the “land of the ʿarbāyē” from the raids of the Persians and the ṭayyāyē,Footnote ⁶⁰ making clear that this region and its population of ʿarbāyē were considered part of the Roman/Byzantine sphere, and that they needed to be guarded from the disruption caused either by Persian attack or raiders coming from the desert. Above all, this passage makes clear that the ʿarbāyē – who are often described in translations as “Arabs” – were different from the ṭayyāyē, who like the Persians were outsiders and enemies. To translate the term ṭayyāyē also as “Arabs”, then, lumps together two different groups of people whom the Syriac sources carefully distinguish.Footnote ⁶¹

Syriac sources, then, do not suggest that there was an operative “Arab” identity in the years preceding the rise of Islam; rather, the conquerors were described by Syriac authors using the term ṭayyāyē traditionally applied to marauding interlopers, usually nomadic, a classic case of terminological conservatism. Their use of this term tells us only that Syriac writers of the seventh century saw the invaders as in some ways similar to the nomadic desert-dwellers who had long interfered in the cultured, Christianized late Roman world of Mesopotamia and Syria, particularly because they originated in the same regions from which the traditional ṭayyāyē had come. But it tells us nothing of the actual nature of the movement to which these people belonged, what motivated their incursion, how they conceived of themselves, or what they called themselves.

In Greek sources from the seventh century, the conquerors are sometimes referred to as arabes, another classic “outsider term” that Greek authors had used for centuries before the time of Muḥammad to refer to the nomads of the steppes east of the settled areas of greater Syria. After about the fourth century ce, Greek authors also began to use the word arabes to refer to the inhabitants of the new Roman Provincia Arabia (the former Nabataean kingdom, occupied by Rome in 106 and incorporated into the Empire).Footnote ⁶² Use by Greek authors of the word arabes to describe the early seventh-century invaders, in other words, represents another case of terminological conservatism. Given the fact that the newcomers included some people whom Greek authors would traditionally have considered arabes – such as camel-herding nomads or people coming from far to the east and south – their calling the conquerors arabes cannot be taken as evidence that these people called or thought of themselves as “Arabs”. And as we have seen above, arabes is only one term used by the early Greek sources to describe the newcomers, and neither of the other terms – sarakēnoi, Agarenoi – suggests that the conquerors called themselves “Arabs”.

In sum, the non-Arabic sources from the Fertile Crescent region, both Syriac and Greek, like the contemporary Arabic sources, offer no support for the notion that the invaders called themselves, or thought of themselves, as “Arabs”.

But what of the claim that a nascent “Arab” identity was already in the process of crystallizing in the centuries before the rise of Islam?Footnote ⁶³ If one accepts this view, then it seems natural enough that the people who organized the conquests from Arabia in the seventh century were also “Arabs”. Related to this is the notion that the rise of Islam was part of a process of ethnogenesis by which the “Arabs” crystallized as a people, in a manner resembling the coalescence of various Gothic peoples on the western frontiers of the collapsing Roman Empire – the Ostrogoths, Lombards, Allemanni, and others.Footnote ⁶⁴

The problem with this hypothesis is that there is no firm evidence that any group of people of the Arabian peninsula or its peripheries called themselves “Arabs” before the rise of Islam. A handful of inscriptions in some form of Arabic language and dating to the centuries before the rise of Islam have been found, but as Fisher has noted, it is difficult to know in most cases whether Arabic is being used as a “conscious ethnic marker” or is simply being used.Footnote ⁶⁵ None of these inscriptions – including those from Zabad (512 ce), Jabal Says (528 ce), Ḥarrān (568 ce), and Umm al-Jimāl (undated, prob. 6th cent.) – refer to a group called “Arabs”. The one possible exception, the famous Namāra inscription of 328 ce, discovered on the fringes of the Jabal al-Durūz, may only apparently refer to “Arabs”. Written in the Arabic language but in the Nabataean alphabet, in its first line it calls the dedicatee mlk l-ʿrb klh, translated by some as “king of all the Arabs”, and considered by them to be evidence for a broad “Arab” identity. But since the Nabataean language did not include the sound “ḡ”, its alphabet (like the later Arabic alphabet) used the sign for ʿayn when rendering the Arabic ḡayn, so this phrase may just as well be read as mlk l-ḡrb klh, “king of all the west” (or of a district called “the west”).Footnote ⁶⁶ Or, it may refer to a district called al-ʿArab,Footnote ⁶⁷ perhaps that same “land of the ʿarbāyē” we saw referred to in Syriac texts. Moreover, we find no further inscriptions referring to “Arabs” in the 300 years between the Namāra inscription of the early fourth century and the rise of Islam in the early seventh. The claim that the Namāra inscription reveals the existence of a nascent sense of “Arab identity” is, therefore, quite open to doubt; and there seems to be no other substantial evidence to support the idea. Particularly important is the absence of reference to “Arab” identity in the surviving pre-Islamic poetry, which if it offers any support for a broader collective identity seems to advance that of Maʿadd.Footnote ⁶⁸ The Quran's reference to itself as qur’ānan ʿarabiyyan is sometimes adduced as additional evidence,Footnote ⁶⁹ but as we have already seen, the interpretation of this phrase as reflecting an ethnic identity is also highly questionable.Footnote ⁷⁰

We do find some emphasis on the “Arabness” of the conquerors, however, in the later Islamic narrative sources. While the main emphasis in those sources, when they describe the conquests, is on their “Islamicness”, with many reports describing the “Muslims” battling the “Unbelievers/mushrikūn”, in ninth- and tenth-century chronicles such as those of al-Ṭabarī (d. 923 ce) we also find reports about key battles in which the two sides are described as being, on the one hand, “the Arabs” (al-ʿarab) and on the other, either “the Persians” (al-furs) or “the Byzantines/Romans” (al-rūm).Footnote ⁷¹ By the time the reports in these sources were being compiled two or more centuries after the conquests, in other words, some concept of “Arabness” or of a broader “Arab identity” transcending individual tribes was available, and was being introduced by transmitters of conquest reports into the stories they told.Footnote ⁷²

The import of our discussion should now be clear: the conquerors in the seventh century, although coming from Arabia, did not call themselves “Arabs”, and there is little, if any, evidence that there existed a widespread “Arab” identity before Islam. Just as we cannot blithely speak of an “Islamic conquest”, or for that matter, of “Islam” or “Muslims” during the seventh century, then, we must also strive to avoid speaking of an “Arab conquest”. To do so is to conjure up a spurious category of “Arabs” many centuries ago through a facile projection of later medieval and modern nationalist concepts that did not exist in the seventh century. Besides being anachronistic, moreover, the term “Arab conquest” misrepresents the character of the conquest movement by masking its religious impetus, reflected in the fact that in the seventh century, the conquerors referred to themselves as mu'minūn, “Believers”.

* * *

But if these terms – “Islam”, “Muslim”, “Arab” – that dominate the existing literature about Islamic origins are henceforth to be avoided in speaking about the seventh century, how should scholars talk about these events, including the conquests?

Here, viewing the conquests as consisting of stages or phases, rather than being all of one piece, may help. What we so easily call “the conquests” was, after all, a prolonged process of expansion and, apparently, of political integration, lasting over a century, so the historian might a priori expect the process to have changed in character over time. The later Islamic sources that provide many reports about the conquests – especially about the heroic early days – do not explicitly break the conquests down into phases: the salvation-historical agenda of these sources does not really allow for such a breakdown, because in their view God's will, like God himself, is one. Nevertheless, the historian who grapples with the various reports about the conquests and the eventual emergence of Islam senses that the early stages of the process were different from what followed. At the start, during the first 25 years or so (the first generation) after the death of the prophet, the conquests seem to have been a manifestation of what the anthropologist Victor Turner called a state of “communitas”, an intense period of ideologically driven fervour that is not yet institutionalized, at least not fully.Footnote ⁷³ Later, however – perhaps already in the days of Muʿāwiya (r. 660–680), and certainly by the time of ʿAbd al-Malik (r. 685–705) and his son al-Walīd (r. 705–715) – the conquests had become virtually an official industry of the Umayyad regime.

These two stages should probably be identified as distinct phenomena, and analysed separately; and we may also choose to use different terminology to refer to each of them. The conquests of the second phase may well have lacked almost completely the kind of ideological energy that drove the first phase, and hence could be seen to have a totally different character. What had begun, it seems, as some form of enthusiastic ideological movement, had by then morphed into a more or less routine process of state expansion, a practice that was carefully organized, like tax collection, and evidently very lucrative for the rulers.

The second stage of the conquests is not only easier to conceptualize and describe, it is also easier to find a satisfactory terminology to characterize it: since it began and continued under the later Umayyad dynasty, we might simply call them the “Umayyad conquests” or “conquests (or expansion) of the Umayyad state” or “Umayyad regime”. Indeed, since Islam in its classical and modern sense as a distinct monotheistic confession was apparently beginning to emerge during the second stage of the conquests, we could even call them the “Islamic conquests”.

Much more difficult is finding a satisfactory term for the first phase of the conquests, during the seventh century, or indeed for any of the events associated with the early developments in the community during that century. For the early conquests, the challenge boils down to whether we wish to see a religious, or a secular (ethnic?) impetus as the primary driving force. This tension between religious and secular understandings of the early phases of the expansion, as we have seen, underlies the long history of terminological ambivalence in modern scholarly presentations, our waffling between either “Islamic” or “Arab”. Almost 40 years ago, Crone and Cook proposed “Hagarism”, which has the advantage that it is derived from a term actually used in some early Greek sources to refer to the conquerors – Agarēnoi.Footnote ⁷⁴ This term, as noted above, had long been used by authors writing in Greek to refer to nomads from the steppes east of Syria. But, since for these Greek-speaking Christian authors it was a reference to the conquerors’ supposed descent from Abraham via his slave-girl Hagar, it always bore a pejorative sense in their hands.Footnote ⁷⁵ Furthermore, the conquerors never refer to themselves as descendants of Hagar. For both these reasons, “Hagarenes” does not seem an ideal term for the historian to use when wishing to identify the conquerors. In any case, it has not been widely adopted by scholars. It is virtually certain that this term was a reflection of the Arabic word muhājirūn, which the conquerors are said to have used to refer to themselves (although, except for one instance in the “Constitution of Medina”, we know this only from later Islamic sources, not from contemporary Arabic ones – as far as I know, no seventh-century papyrus or inscription in Arabic uses the term).Footnote ⁷⁶

Roughly 20 years ago, the author proposed that Muḥammad's early community be called a community of “believers” (mu'minūn), a term that is also used in early sources.Footnote ⁷⁷ As we have seen, the Quran consistently and with overwhelming frequency addresses the people of Muḥammad's original community as mu'minūn, and this term also appears repeatedly in seventh-century Arabic inscriptions and papyri, particularly where they refer to the leader of the movement as amīr al-mu'minīn, “commander of the believers”, beginning already with Muʿāwiya (r. 660–680). This latter term is also seen in semi-translation in the writings of the conquered peoples: as amīrā d-mhaymenē in Syriac, amira tōn pistōn in Greek, and as amir-i wruishnikan in Pahlavi/Middle Persian (the latter on coins), in all cases best translated as “commander of the believers”.Footnote ⁷⁸

Whether the identity of the earliest conquerors as mu'minūn, “believers”, meant that their earliest community was open not only to followers of the Quran but also to righteous Christians and Jews, as I have argued elsewhere,Footnote ⁷⁹ continues to be debated; but that the conquerors conceived of themselves as “believers” can hardly be called into question, and the word suggests a fundamentally religious identity, whatever its exact character may have been.

Religious and material motivations very probably worked in tandem in the new movement, however. All movements that win a widespread following must appeal to many kinds of people, drawing supporters through a variety of incentives. For the Believers’ movement, it was not merely a question of the raw appeal of purely religious ideas, although that was surely part of it. Some “high-minded” people were doubtless drawn to the movement's purely religious ideas of divine unity and its ideals of pious behaviour, or were filled with fear of the Last Judgement; but the movement also offered tantalizing prospects for those who were basically motivated by material interests, who lusted after new lands to settle or properties to hold, after the plunder and rapine made possible by unsettled circumstances, or who simply longed for adventure and new horizons. Moreover, the religious ideals could serve as a justification – or cover – for the pursuit of material advantages, and the worldly success the conquerors enjoyed could be seen as affirmation of the truth of the religious message. We can, thus, perhaps best speak of an amalgam of mutually reinforcing secular and religious motivations on the part of the individuals drawn into the movement. But even in such an amalgam, there must be some central ideological component to the movement that provides the focus around which the other incentives are structured.Footnote ⁸⁰ Without a guiding ideological component, such a movement could not sustain itself, but would either collapse with the first setback, or quickly exhaust itself as the purely pragmatic motivations of the many participants led them to pursue their own separate interests helter-skelter. And in this case, the central ideological component seems to have been religious.

What then, was the nature of the religious impulse that lay at the root of the early conquests of the believers/mu'minūn and the rapid expansion of their embryonic state? The Quran's insistence on monotheism, on the idea of a Last Judgement and the concomitant need to live righteously in accordance with God's revealed law, and on the primacy of a revealed scripture, may have been powerfully articulated, but they were already well-worn notions in the Near East of the early seventh century, ideas that had hitherto caused most of those attuned to them, especially Christians, to embark on monastic renunciation of the world, rather than militant conquest of it. Considerable recent (and some not so recent) research, however, has pointed out the eschatological tone of many passages in the Quran and emphasized the likelihood that the early community of believers was motivated in part by intense apocalyptic enthusiasm – perhaps seeing themselves as participants in the very events of the anticipated “End-Time”, the creation of a new God-guided realm, which spurred them to expansive conquest so that they could “inherit the earth” God had granted them.Footnote ⁸¹ The fact that the conquerors, as we have seen, established for themselves near the beginning of their movement a new dating era – what early documents refer to as sanat qaḍā’ al-mu'minīn, “years in the jurisdiction of the believers”, and what later Islamic (and Islamicizing) sources refer to as the era of the hijra – suggests (but of course cannot prove) that the movement had an eschatological component.Footnote ⁸²

But this brings us back again to the nagging question: what are we to call this expansion, this first phase of the conquests that burst forth from Arabia in the early seventh century? What term can we use for this seventh-century conquest that does not anachronistically imply that it was motivated by a reified religious identity (Islam) or a politically articulated ethnic identity (Arabs), for neither of which there is meaningful early evidence and that, in each case, crystallized only at a later time? What label can we find that also does justice to the likely range of motivations among those actually involved, from intense pietism and eschatological foreboding to the basest opportunism and lust? “Hagarism” seems too rooted in a pejorative “outsider's view”; “Believers’ movement” sounds to some too vague, not sufficiently militant, too uniformly religious. Peter Webb has recently suggested – perhaps only facetiously, however – that we might call the conquests “Maʿaddite”, since the name Maʿadd is, in fact, found in both pre-Islamic Arabian inscriptions and early Arabic poetry as a term for a broad collective identity of Arabians, only later to be replaced by “Arab”.Footnote ⁸³ Perhaps, since the conquests issued from the Arabian peninsula and were led by people coming from Arabia, we should call them the “Arabian conquests” or the “Ḥijāzī conquests”?Footnote ⁸⁴ Or perhaps, since the conquests seem to have been led from the start by people from the tribe of Quraysh, we should call them the “Quraysh conquests”, akin to the way we might speak of conquests by the Habsburgs or the Hohenstaufens? But then, no sources – internal or external – refer to the conquerors or conquest explicitly in this manner, either. Perhaps the most cogent solution yet proposed is that of Azīz Al-Azmeh, who has suggested “Paleo-Islam” and “Paleo-Muslim”,Footnote ⁸⁵ from which we might also conjure the variant “Proto-Islamic”. Something like this has the advantage that it conveys the sense that the conquests lay at the root of what comes to be called Islam, while implying that the Islamic community with its firmly defined boundaries was still some distance in the future. But do these variants adequately communicate the likelihood that the community founded by Muḥammad seems in its early years to have included some Jews (as reflected in the Constitution of Medina) and Christians (as seen in the early Umayyad period)? The question remains open.