Introduction

There are between six and seven million Arab Christians in the Middle East today (1995). They represent 6.3 per cent of the whole population of the countries among which they are spread: Egypt, Lebanon, Syria, Iraq, Jordan, Israel, and Palestine. This is more or less the same percentage the Ottomans found almost 500 years ago, when they wiped out the Mameluke and Persian dynasties which shared control of the region. However, this does not mean that the ratio of Christians to Muslims is permanently fixed. On the contrary, after four centuries of Ottoman rule, during which an unexpected resurgence took place, Christianity reverted to the downward trend it had followed during the first millennium after the Hegira. In the twentieth century Islam has advanced at a pace comparable to that of the pre-Ottoman era. However, the mechanisms which have influenced this modern trend belong to a different kind from those of the pre-Ottoman era; all are linked to modern political, economic, and demographic processes.

The study of the demography of Christians in the Arab East involves looking at the frequency of a particular characteristic - religious affiliation - in the population as a whole, as well as describing the evolution of this characteristic over a very long period of time and over a very wide area. Essentially it means studying Islamization, bearing in mind that the term is used here simply to mean the increase in the percentage of Muslims. There are four processes which may have led to this: conversion from one religion to another, sometimes of individuals, sometimes of whole groups; massacres and exoduses with a religious connotation - both exceptional - which result in the substitution of one population by another; the fusion of population groups of different religions which, through mixed marriages, produces a second generation of one religion; differential population growth, where different birth and death rates influence the speed of population growth according to religious affiliation.

Each of these processes has its own time scale. As demographic processes inevitably occur over long periods of time and follow a rhythm of their own, they do not generally fit in with the important moments of factual history. On the other hand they do coincide with the longer periods of social and economic history or of important geopolitical relations. The explanations for demographic processes therefore lie in general contexts rather than in specific events in history.



1. The Change to a Position of Minority, from the Hegira to the Mamelukes

At the birth of Islam in the seventh century, Christianity had become the faith of the vast majority of the population of the Fertile Crescent and of Egypt, although compromises were made with the ancient beliefs, and although the spread of the faith undoubtedly varied according to time and places. When the population of the Arab East was counted for the first time, in the second half of the sixteenth century, the process of Islamization seemed to be coming to an end, as had happened three centuries earlier in the Maghreb. The number of Christians in the Fertile Crescent was then slightly less than 10 per cent of the whole population. ¹ The figure was very possibly the same in Egypt too. ² There is no direct information about the population of the first ten centuries of Islam and even less on the evolution of the religious composition of that population. The sources, which are incidental, fragmentary, and disparate, provide information which is only indirectly linked to demography: taxation, military recruitment, production, trade, and settlement. Besides the natural margin of uncertainty of such sources, there is also that of their numerical ratio to the population. Furthermore these sources, invariably intermittent, give at the very most a few isolated traces, and never a complete series of facts. To describe how 90 per cent of Eastern Christians became Muslims we can therefore only resort to hypotheses.

I will begin by destroying a myth: that of an invasion by an Arab Muslim population. Frequently referred to by some Maronites in their claims to descend from the ancient Phoenicians, or by some Copts in their claims to descend from the Egyptian Pharaohs, this myth infers that Christians in Lebanon or in Egypt are of a different ethnic descent from their Muslim countrymen. Religion is quite an important distinguishing feature in the East today. There are moments when this distinction is felt as a real difference. It is easier then to accept these differences if they can be explained by foreign descent. The myth is thus founded by modern experience, and not by demographic history. The Arab peninsula has never, in any period, had a large enough population to produce an emigration which could account for more than a tiny minority of the present population.

Like many other conquests, that which came from Arabia in the seventh century involved small military groups attacking whole populations. On their arrival the conquering soldiers nevertheless proclaimed the division of the world into two parts: lands already acquired and placed under the control of the Muslim State, and the remainder of the universe, which still had to be conquered. Everywhere it established itself, Muslim power naturally tried to keep the peace and to guarantee economic prosperity. So instead of direct attacks on the religious faith of the people they had conquered, which represented one of the most deep-rooted features of their identity, Islam drew up a code to define the status of those who professed Christianity or Judaism, the monotheistic religions from which it claimed to have descended. The coexistence of different religions was enshrined in law. Among the practices instituted by this code was a per capita tax paid by Christians and Jews. At the outset the Muslim State used this to fund the war. The preservation of non-Muslim populations under its jurisdiction thus became, paradoxically, a means for its territorial expansion. Political conquest advanced much more rapidly than spiritual conquest. However, the status accorded to Christians and Jews turned out to be ambivalent. The recognition of their existence allowed their communities to survive; at the same time the establishment of certain discrepancies encouraged them to join Islam. This change was most probably a slow, gradual process, apart from a few waves of mass conversions, particularly at the end of periods of religious tension caused by the international situation. This was undoubtedly the case during the reign of the first Mamelukes, whose arrival ended the difficult period of the Crusades.

The disparity in numbers of the conquerors and the populations they conquered soon set another mechanism in motion, the fusion of different populations. The soldiers, preachers, and merchants who first brought Islam out of Arabia formed a predominantly male population which was too small to allow marriages solely within the group. Mixed marriages were necessary for its survival. Here Muslim law instituted an important asymmetry: whatever the combination of the parents' religions, the offspring would always be born Muslim. In fact children belong to the religion of their father. The marriage of a Muslim man to a Christian woman thus produces Muslim offspring. Symmetric marriages, between a non-Muslim man and a Muslim woman, are forbidden, and require the man's prior conversion to Islam. If we suppose that marriages occurred regardless of religion, that is at random, and that this process of fusion was the only factor, Islamization can be said to have followed a logistic (S-shaped) curve. Although slow to begin with, it accelerated later, until the Muslims began to number half of the population, after which it slowed down again. A simulation shows that over the period in question, covering a thousand years, mixed marriages alone, and the conversion required for men in half of these cases, ³ would have sufficed to create a proportion of Muslims of around 90 per cent. Of course marriages do not occur at random, but follow all sorts of rules from the social and cultural to geographical proximity. This does not prevent the demographer from seeing this simple arithmetic of mixed marriages as an essential factor in mass Islamization.

There are no certain indications of the other two processes, the substitution of one population to another by violence, or differential population growth. There were undoubtedly periodic massacres, as under Caliph al-Hakim in Cairo (996-1021), but none was large enough to be considered even local exterminations. Some groups were also driven out of their territories, but they were not always Christians. On the contrary the most famous example, and a much debated one, is the exile of the Shi'ites of Kesrouan in Jabal 'Amil. Punished by the Mamelukes for their dissent from orthodox Islam, and possibly also for their recent openness towards the Frankish kingdoms, the Shi'ites left the field free for the influx of Maronites from the Qadisha. This was perhaps the first time, though not the last, that a Muslim power was to defend itself in this way from the geographical expansion of a Christian population. As for natural demographic processes, there may have been slower or faster population growth according to different religions, but there is no evidence to support even the least hypothesis on this account. As an aside it is worth remembering the poet al-Djahiz (ninth century), who observed that Christians 'fill the earth', put their prolific numbers down to their practice of monogamy. The remark would be very apt today, when Muslim marriages have proved to be more unstable than Christian ones, because they have been undermined by divorce and polygamy.



2. A Change of Tendency under the Ottomans

Before conquering Syria and Egypt, the Ottoman Sultans reigned over a mostly European and Christian Empire. In their Balkan possessions and in Istanbul they devised, and then tried out, a new system for the coexistence of different religions, which they extended to their Arab provinces. They replaced the series of rules which had previously governed all relations between Christians, or Jews, and Muslim society and power with the millet, which gave legal recognition to the main non-Muslim communities. In their private lives - marriage, inheritance, family roles, education, health, and so on - individuals were now subject to an authority belonging to their religion, to whom the Sultan of Istanbul delegated power in all matters pertaining to the millet. This system of mediation sanctioned the communities, which were composed at least of stable groups within a multi-religious society, ⁴ although not yet of embryonic nations. It played a highly important role in the demographic evolution which began at that time. We have reconstructed this evolution ⁵ by putting together the figures from the censuses of two eras, the sixteenth and the nineteenth centuries, classified separately by Omer Lufti Barkan ⁶ and Kemal Karpat, ⁷ as well as by a few other contemporary historians. ⁸

Under the reign of the Ottomans, the percentage of Christians in the Fertile Crescent tripled. Having fallen to 7 per cent at the end of the Mameluke era (according to records of 1570-90), it rose to 20 per cent (1914), rising even to 33 per cent overall in 'Greater Syria' (Lebanon Syria, Palestine). It was a genuine turnabout in the tendency of the preceding millennium, which seemed to lead towards the imminent extinction of Christianity. The resurgence of the Christian population was certainly not continuous throughout the four centuries considered. It was concentrated instead in two periods in which there was a general population growth in the Southern and Eastern Mediterranean: the sixteenth century and the second half of the nineteenth century. Furthermore it is highly likely that it did not take place in Egypt, ⁹ where the percentage of Christians had fallen to about 8 per cent under the Mamelukes and was still the same at the beginning of the twentieth century (see Table 2.1).

The demographic reasons for the turnabout which appeared in the Fertile Crescent ¹⁰ lie in each of the four processes mentioned in the introduction. The two mechanisms which had allowed Islamization during the previous millennium, conversion and fusion, virtually ceased to operate, at least on a mass scale. The economic incentive of conversion to Islam disappeared and such conversions therefore became rare. In fact, due to the continual decline in the Christian population, who paid a per capita tax, taxation had been changed. From the beginning of the Ottoman era the jizya was no longer a very heavy tax. It was gradually reduced, until it was finally abolished through the tanzimat.

Individual conversions certainly continued, for sociological or marriage reasons, but there were no more conversions of whole groups. There are also known to have been conversions from Islam to Christianity in this period for the first time. The conversion of the Shebab Emirs of Mount Lebanon is proof of this. Originally Sunnites, although Emirs of the Druze, they adopted the religion of the Maronites in 1756. These were nevertheless isolated cases; what is most important to remember about the period of the millet is that each individual, whether Muslim or Christian, kept his own religion. Mixed marriages also became exceptional. Over time the Christian populations had gathered in territories where most of the population were of the same religion. The de facto separation of the communities, even within districts of the same town, made social contact between different religions difficult and rendered opportunities for mixed marriages rare. The consolidation of the communities through the establishment of the millet put a final seal on this separation. Although the Copts, the largest Christian community in Egypt, had a slightly different status from that of the millets of the Empire, from which Egypt in fact remained isolated, the two tendencies previously described also appeared in this country. The end of conversions and the rarity of mixed marriages brought an end to a very long, initial period of Islamization. In the Fertile Crescent the separation of the communities also encouraged the development of natural, independent demographic patterns. Differential birth and death rates led to very different population growth rates from one community to another.

Little is known about the variations of birth rates according to religion during the Ottoman era. On the basis of their greater family stability, mentioned earlier, we can only suppose that Christians must have had a higher birth rate than Muslims. At any rate this was perceivable on a local level in Mount Lebanon in the nineteenth century. ¹¹

On the other hand we know that for long periods, although not continuously, Christians had a lower death rate than Muslims. This explains their extraordinary growth. Three main factors played a role in this: religious geography, behaviour patterns during epidemics, and the adoption of modern social customs. Religious geography was the first factor. Since the Conquest, Islam had taken root more deeply in the Arabian peninsula and Mesopotamia than in the areas bordering onto Byzantium in the West. The Crusades on one hand, and the Mongol invasions on the other had later accentuated the concentration of Christian communities in territories along the Mediterranean coast, in the mountains bordering the sea, and, to a lesser extent, in the westernmost cities of the steppe. The economic expansion of the Christian West, from the Renaissance onwards, increased the concentration of Eastern Christians in areas along the Mediterranean. A common religion, some Eastern branches of which had reestablished direct contact with Rome, ¹² soon made the Christians of the East the favoured partners in dialogue and trade of European merchants and diplomats. Furthermore in the sixteenth century the prosperity and stability brought by the domination of the oceanic trade routes spelled the disappearance of the great famines and led to a positive population explosion in the Mediterranean. For the simple reason that they were more closely linked to the Mediterranean world, and therefore had more contact with the West, Christians were the main beneficiaries of this boom. However, although famines became less frequent, the opening of the Mediterranean to the flow of goods and people from distant lands facilitated a few outbreaks of epidemics. There is ample evidence that, on these occasions, the behaviour of Christians limited the spread of the epidemics. While Muslims presented a united front against the illness, increasing contact with each other in their efforts to face it, Christians reacted instead by isolating themselves. They followed the example of the Western merchants of their own kind, falling back on their families and shutting up their houses. Consequently, almost all the statistics for deaths show that epidemics were actually less fatal among Christians than among Muslims. Nature also helped to mitigate the spread of epidemics in certain Christian communities, particularly the Maronite community. The mountain they inhabited provided a climatic barrier to many of the infections which struck in the plains and the cities at the time.

From the second half of the nineteenth century modern social customs became a third fact in favour of Christians. Control not only over epidemics but especially over the high death rate which was normal at that time is ultimately a consequence of the acquisition of knowledge and its diffusion among the population. Many demographic studies of reductions in the death rate in a wide variety of contexts show that the spread of school education is a highly important factor. It was in Christian communities that schools first became accessible to common people. ¹³ Access to knowledge, of an elementary kind for most, but nevertheless sufficient to change attitudes to hygiene and illness, undoubtedly became the most important reason for Christians' advantage in terms of a lower death rate. There was no such differentiation in the death rate of different religious groups in Egypt. The Copts had no special geographical characteristics - like the Muslims they were more rural than town-dwelling and the region where they were concentrated, Central Egypt, had no particular contacts with the West - nor were they any more advanced in the education of their children than the Muslims. From what we can suppose, they must have had the same death rate as Muslims. This explains why their percentage in relation to the whole population did not change.

Alongside the strong differential tendencies in death rates, migrations also played a role in strengthening or weakening the proportion of Christians in the population in different places. Local and regional migrations were relatively intensive during the Ottoman period, as the status of the millet gave the community a transnational quality, continually disassociating it from any particular territory. Whether from Istanbul, Aleppo, or Beirut, people who professed to belong to the same Church came under the same authority. The borders between provinces were open. Four great migratory movements within the Empire shaped the religious map of the Arab East one after the other. The first was the convergence towards Aleppo of Christians from Anatolia, Iraq, and Syria from the sixteenth century onwards. The expansion of the Maronite territory in the Lebanon Mountains, due to fighting among the Druze at the beginning of the eighteenth century was the second. This was followed by the settlement of Syrian or Lebanese Christians in Egypt from the first half of the nineteenth century. Lastly, there was the movement of Christians from Mount Lebanon first of all, then also from Damascus and central Syria, and soon from all the other provinces of the Empire towards the coast, especially to Beirut; this movement was noticeable from the beginning of the seventeenth century, and then increased until the beginning of the twentieth. These migrations were sometimes spurred by the desire to flee from religious tensions, ¹⁴ sometimes by the attraction of a dynamic region which had many contacts with Europe.

The first long-distance emigration of modern times in this part of the world also had a paradoxical effect on local demography. To escape the overpopulation of the mountainous area, made worse by the crisis in silk production, many Lebanese, mainly Christians, moved to the New World after the civil war of 1860. Three thousand people left every year between 1860 and 1899, and fifteen thousand between 1900 and 1913. ¹⁵ Many of them made their fortune. The links they kept with their country of origin soon took the form of investments, which overturned the economic situation of Mount Lebanon. The new prosperity enabled an increase in the population, and continued, vigorous population growth. This is how Mount Lebanon, a predominantly Christian area, became the most densely populated region in the Arab Orient east of the Nile.



3. The Increase in Islamization in the Era of the Nation-States

There are various estimations, often contradictory, of the current size of the Christian communities, provided both by the authorities of the Eastern or Western Churches and by a number of scholars. However, in both cases, the estimations are not always based on the best modern source of demographic information, the population census. On the contrary, when they inform us about the size of the different religious communities they often dispute the validity of the censuses. I therefore hope that the reader will forgive the following digression on the question of statistics, which I feel is necessary before going back to the four processes which influence the religious composition of society.

Among the seven Middle Eastern countries which currently have an Arab Christian community - Egypt, Lebanon, Syria, Iraq, Jordan, Israel, and Palestine -Egypt is the only one to meet the two conditions which enable its Christian population to be examined over the course of time. One of these conditions is the recording of religious affiliation as a variable, individual characteristic in all the population censuses since 1897; the other is the geographical consistency of the territory covered by these censuses. The latest, carried out in 1986, showed 2,829,349 Christians, 5.9 per cent of the whole population of the country. Updated to 1995, this percentage would give a figure of 3,300,000 people. The Coptic authorities, some politicians, and a number of Coptic or Muslim intellectuals consider this figure to be a gross underestimation of a community which may actually number between 6,000,000 and 12,000,000 people, in other words between 10 per cent and 20 per cent of Egyptians. The statistic is thus alleged to be false, both directly - the results of the census are said to have been manipulated at the top before being published - and indirectly, as it is claimed that some Christians tend to declare themselves Muslims, due to pressure from their environment. Contrary to these allegations however, the statistic seems to be reliable, with the margin of error that all statistics carry. This argument, developed in another essay, ¹⁶ is based on the observation that the Egyptian figures are consistent in two respects: both over the whole period of time and at each individual moment in time. The graph formed by the nine Egyptian censuses over a period of ninety years is quite regular, both as regards the whole country, where the percentage of Christians falls in a gradually sloping straight line from 1927 (see Table 2.2 and Figure 2.1), and in the different provinces. If there had been any manipulation, this pattern would not have been constant. On the contrary there would have been variations in the numbers and perhaps also in the consistency of the graph, according to the changing attitude of political forces, of regimes, and of society towards the Christian part of the population. Furthermore, censuses are not the only source of religious statistics in Egypt. The registration of births, marriages, and deaths provides another source. The two operations are carried out by completely separate branches of public administration using quite different methods. They nevertheless provide results which are completely consistent with each other. For example, the birth rate of 30 per thousand, obtained by matching the 85,000 Christian births recorded every year throughout the 1980s to the 2,800,000 Christians counted in the census of 1986, would only be 11 per thousand if the number of Christians were 8,000,000. ¹⁷ With such a low birth rate their percentage of the whole population would have undergone a very sharp fall, instead of showing just a slight drop, which clearly contradicts the numbers claimed. So we might as well admit the plausibility of the published statistics.

An observation of the countries of the Fertile Crescent in the twentieth century is not as straightforward as in Egypt. The first reason is the interruption in the series of statistics, caused by the break-up of the Ottoman Empire and the reconstruction of the Arab East into national States. As the new borders were not traced over those of the old provinces, the figures from the Ottoman era have to be reorganized in order to match the last population count given by the Imperial administration in 1914 with the first ones provided by each of the new States in different years. This calculation, the results of which are shown in Table 2.2, introduces a certain amount of uncertainty about the growth rate of the Christian populations up to the date of the first national censuses. The second difficulty in ascertaining the size of religious communities stems from the fact that the Ottoman practice of recording religious affiliation in the censuses was not continued to the same degree by all the States. Only Jordan continued it in all its censuses, though the most recent one available is from 1979. ¹⁸ Iraq and Syria discontinued the practice when the secular Ba'th Party came to power in the 1960s. Israel records the religious affiliation of Israeli citizens at every census, though it has not published any statistics on the religion of the Palestinians in the occupied territories since 1967. Lastly, Lebanon has simply stopped taking censuses of its population and therefore of its religious communities, since the census of 1932, which was used to apportion political power among the religious groups. As changes in the religious composition of the population threatened to fuel contestations of the National Pact, arguments over demographic figures were simply excluded from the political debate. As statistics on religion have been discontinued for various lengths of time from one country to another, an estimation of the current situation (1995) can only be made by intelligent suppositions of likely tendencies (see Table 2.3). These suppositions carry a margin of uncertainty, which increases in proportion to the length of the period considered.

Despite these reserves it is nevertheless certain that the exceptional growth of the Christian population east of the Suez ended with the close of the Ottoman Empire (see Figure 2.2). The percentage of Christians in the whole population reached its highest in 1914: 26.4 per cent in the whole of the Near East (Israel, Palestine, Jordan, Lebanon, and Syria); 58.6 per cent in Lebanon (within the borders of 1920), 11.3 per cent in Palestine, and 10.1 per cent in Syria. However, it only reached 8.1 per cent in Egypt and 2.2 per cent in Iraq. Today Christians only make up 9.2 per cent of the population of the Near East. In Lebanon, where they have undoubtedly lost their position as the majority, they number little more than 40 per cent, ¹⁹ in Syria they are about 6.4 per cent, in the Palestinian-occupied or autonomous territories the figure is 3.8 per cent, and in Israel 2.1 per cent. In Egypt they constitute 5.9 per cent of the population, and in Iraq presumably 2.9 per cent. Thus eighty years seem to have wiped out four centuries of a population resurgence, though perhaps not a continuous one, and brought the Christian population back to its lowest level in history, the figure recorded at the end of the Mameluke era. The nation States provided the political context for the resurgence of Islamization. The breakup of the communities into one, single new national identity, as well as the equality of citizens, now linked directly to the State without the interface of religious bodies, affected all the four processes which have been examined for the previous periods.

Conversions to Islam due to political or economic pressure were probably not any more frequent than at the time of the Ottomans. An increase was certainly recorded, but more in response to sociological pressure, at moments of conflict between religious groups. Examples of these are the violent attacks of the Muslim Kurds against the Nestorian Assyrians in Iraq in 1933, the civil war in Lebanon between 1975 and 1990, and incidents in the last few years in the suburbs of Cairo and in some parts of Central Egypt, where relations between Muslims and Copts have deteriorated. Although there are no statistics, conversions of this kind very likely remain rare, individual cases.

On the other hand, substitution by violent means, involving whole population groups, has reappeared. In Iraq it drastically reduced the Christian population of the hinterland of Mosul: the Kurds drove almost all the Nestorians to flee abroad, and a large part of the Chaldeans to flee to Baghdad. During the civil war in Lebanon, territories were physically divided up according to religion. We know that there was reciprocal substitution of population groups by violent means. Many Christians left regions of the country which had been taken over by Muslim militia. However, just as many Muslims ²⁰ were forced to leave regions controlled by Christian militia. A hundred and twenty years of coexistence, which modern economic progress had gradually introduced into almost all parts of the country, were wiped out in fifteen years. Paradoxically however, the Beirut area attracted a flow of people from all directions, not least the Shi'ites who came up from the South. Although the largest Lebanese community numerically, they had been under-represented in the capital until then. Thus for the first time in history all the communities of the country came to live alongside each other, though not to intermingle, within the confined space of its capital city. In Lebanon the mechanism of substitution had an effect on local demography in certain areas, but it hardly changed the overall percentage of Christians within the whole population. However, substitution by violence, which changed the religious composition of the population, did not benefit either the Christians' demographic situation or the Muslims', as it consisted of a flow of Jewish people into Palestine, then into Israel, followed by the dispersion of a population with a Muslim majority and a Christian minority around the region and soon around the world.

International migrations which did not involve the substitution of population groups and were sometimes caused by political or economic circumstances, also changed the size of religious groups. Three such important kinds of movement took place in the twentieth century. The first of these was the convergence on Lebanon of populations from the whole Middle Eastern region. Many of these were Christian, and were granted Lebanese nationality. They included the Armenians fleeing from the Turkish massacres at the beginning of the century, the Catholics and Orthodox of Aleppo and Damascus, who were escaping the political and economic insecurity of the first two decades of Syrian independence, and the Syro-Lebanese, who left Egypt, the land where their ancestors had settled, but whose regime and political climate alienated them shortly after the Free Officers came to power. Others, however, were Muslims: Kurdish refugees from Iraq and Turkey, some of whom became naturalized Lebanese, while only a very few of the largest group of immigrants, the Palestinians, mostly Muslim, became Lebanese citizens. Thus the flow of immigration increased both the Christian part of the population of Lebanese nationality and the Muslim part of the population resident in Lebanon.

The second type of movement was the emigration towards more distant lands: the Americas, Europe, Africa, and Oceania. This emigration came above all from Lebanon, but also from Palestine, Syria, and Iraq for the whole of the century, while in Egypt it was not until the 1960s that any sizeable emigration occurred. These long-distance and permanent migrations were undoubtedly more common among Christians than Muslims. However, because they settled in secular countries, whose statistics do not take religious affiliation into account, it is impossible to estimate their religious composition.

The third and final type of movement is the regional migrations. These were probably temporary and sometimes precarious: the rush to the countries of the Gulf and to Libya caused by the boom in the oil trade between 1973 and 1985. With the exception of Iraq, which received these migrations, such movements came from all the Arab countries considered here. However, as they arrived in countries which gave little importance to statistical surveys, these immigrants cannot be counted according to their religious affiliations, as in the previous case. It can simply be supposed that the great majority of them were Muslims.

Except for Palestine, it has been the silent and steady process of differential population growth which has changed the religious composition of the area most deeply. As education has been more widespread and longstanding among them, Christians have kept the advantage they gained over the Muslims in the Ottoman era in the universal process of population transition. Thus, having been the first to reduce their death rate, which had increased their numbers in relation to the rest of the population, they were also the first to limit their birth rate, with the result being an erosion of their size in relation to the rest. At the same time the instance of divorce in Muslim families decreased and pushed up their birth rate. Many figures show that the two curves of the Muslim and the Christian birth rates actually crossed each other in the interwar period, the former in an upward direction, and the latter in a downward one. In Egypt the average number of children was 6.6 for Muslim women, and 5.1 for Christian women between 1944 and 1948. Between 1970 and 1974 it was 5.7 for Muslim women, and 4.3 for Christian women. In Palestine the difference increased more rapidly and more frequently: between 1925 and 1929 the average birth rate was 8.7 children for every Muslim woman, and 5.8 for every Christian woman; between 1940 and 1944 these figures had changed to 7.4 and 4.5 respectively. In Lebanon the birth rate at the end of the 1960s was 5.1 children for every Christian woman, 6.9 for every Sunnite woman, and 8.5 for every Shi'ite woman. It is only recently, since the 1970s in Lebanon, and since 1985 in Syria and Egypt, that the Muslim birth rate has also decreased. This trend can now be seen everywhere, except in Palestine, where the birth rate in the occupied or autonomous territories, whose population is 98 per cent Muslim, is now at its peak (around 8 children per woman). ²¹ In Lebanon there was no longer any variation in the average birth rate from one region to another between 1983 and 1987: in Beirut, a mixed city, it was 2.3, in Mount Lebanon, a mainly Christian area, it was 3.0, and in southern Lebanon, mainly Muslim, it was 3.6. In Egypt the differences in the birth rates of different religious groups became minimal. However, the difference in the birth potential, the result of decades of differential birth rates, started to have an effect, and the percentage of the Christian population continued to fall.

Finally, we come to the fourth process of Islamization, the fusion of population groups through mixed marriages, which modern economic and political conditions could make more common again. There are no statistics for mixed marriages, ²² but they are believed to be frequent, and there is a certain amount of evidence which supports this. In Egypt, for example, marriages between Christian men and Muslim women are thought to account for the 15,000 conversions to Islam said to take place every year on average. ²³ While the Empire's organizational framework was based on difference and separation, the nation State promoted homogeneity and non-distinction. Many of the places where social contact takes place have lost their religious character. This has largely happened in schools, universities, in companies, and in public services, as well as on daily journeys within the cities; in short in many of the places where people are likely to meet their future spouse, now that families have less say in arranging marriages. This last process of Islamization could be the next religious equalizer after natural demographic processes. After all, it seems to have been the most active one during the earliest years of Islam.