This study has a simple and almost anecdotal starting point. I am fascinated by the formula of the “Five Pillars of Islam,” of which I was unaware until relatively recently.1 Although I have run across it frequently since, both in Arabic sources (I knew instinctively that it was the translation of the Arabic arkān al-islām) as well as in Western ones, I never gave it much thought. Living in France, I noticed that only people unfamiliar with the Muslim world and/or the Islamic intellectual tradition attached any importance to this phrase. Is it because it presents Islam in a simple and pedagogical way, similar to a catechism formula? While studying the history of Muslim fasting, which is often presented as one of the Five Pillars, I realised that I knew neither the history of these Pillars nor their exact meaning.2
For this reason, I decided to shed some light on the history and signification of the Five Pillars. I began by examining numerous sources, although I knew that it would be impossible to do so systematically. One man alone could not undertake such a task, except for an examination of the Qurʾān and some of the hadith. I therefore surveyed various sources with which I had a certain familiarity, trying to consider different periods; one conclusion quickly became apparent. This formula is completely absent in the oldest sources – for example, in the works attributed to Mālik (d. 179/795), al-Shaybānī (d. 189/805) or al-Shāfiʿī (d. 204/820) – in any of its variants (arkān, daʿāʾim, etc.), even though its reference, namely, the idea that there are five key elements that define participation and belonging in Islam,3 is attested as early as the first half of the third/ninth century in hadith compilations. The formula itself, which is based on an analogy with a dwelling, does not appear in its definitive form until the fourth/tenth century. However, it is only in the fifth/eleventh century that an author explicitly describes and attributes great importance to it. It is also interesting to note that, despite variations in content, the formula is not unique to any specific religious current: the idea that Islam is defined by five elements (but not the allegory of the Pillars) circulates among the ahl al-ḥadīth, even before the constitution of fixed religious communities in approximately the third/ninth century.4 Thus, the concept was subsequently taken up by both Sunnis and their Shīʿī and Ibāḍī opponents.
While conducting my research, Jacqueline Chabbi’s book Les Trois Piliers de l’Islam was published.5 Naturally, I was quick to pick it up. As soon as I read the Introduction, I realised that it had nothing to do with the history of what is commonly known as the “Pillars of Islam” and therefore with the research I was carrying out. In fact, Chabbi, continuing her impressive reading of the Qurʾān (Les Trois Piliers de l’Islam being the third volume), attempts to show that the religious system of which the Qurʾān is the expression is based not on the historical formula of the Five Pillars (which is the subject of the present investigation), but on three distinct concepts (the covenant, gift and guidance) which, according to her, play a central role, hence the term ‘three pillars’. After isolating “the Islam of the Qurʾān,” thus contrasting it with the later Islam of the Five Pillars, Chabbi seeks to explain its nature more clearly.6 Her interpretation does not oppose mine, but complements it in a certain way: as will become clear in the following pages, while the allegory of the Five Pillars proposes a singular definition of the religion called Islam, it deviates somewhat from that which is outlined in the Qurʾān. The investigation proposed here concerns the Qurʾān only marginally; it focuses mainly on post-Qurʾānic literature, which Chabbi excludes from the scope of her study. Furthermore, I believe that her title is based on a certain misunderstanding: that the religion of the Qurʾān is very different from that of the corpus of hadiths and treatises composed from the third/ninth century onwards. This thesis says it all, and it is easy for a historian to demonstrate: simply make the comparison and the results will speak for themselves. But why insist on rereading the Qurʾān in the light of the allegory of the Pillars of islām, only to find an equivalent with different significations? The Islam of the Qurʾān and that of the hadith are two different forms, which share common features but differ in many other respects. Thus, if there is a break between the Qurʾān and later literature, it is not as radical as Chabbi suggests. Moreover, the differences between the two religious systems have less to do with what she calls the anthropological “substratum,” the importance of which she stresses in order to give a historical understanding to the Qurʾānic statements, since this substratum does not disappear with the emergence of an “imperial” islām. The social organisation based on kinship, with all that it entails, continues to prevail everywhere, as does the symbolic logic of the gift. What changes when we move from the Qurʾān to the hadith is, on the one hand, the very structure of the religious system (e.g. the passage from a “national” religion to a “universal” and personal religion) and, on the other, the relationship of this system to society.
As I shall endeavour to show, the idea of the Five Pillars was undoubtedly originally promoted to put an end to easy excommunications and the resulting dissensions and tensions. This allegory arose in a context marked by theological disputes, which inevitably led to political conflicts. In order to establish a more or less stable political order, but without dismissing theology as a determining factor, a minimum, common definition of Islam was proposed. It was only very late that the Five Pillars were used to present Islam to the uninitiated.
While recognising the relevance and temerity of Chabbi’s reading of the Qurʾān, it must nevertheless be stressed that both the Qurʾān and its exegetes totally ignore the ternary figure that she describes. That there are transformations, or even ‘ruptures’ in the history of a religion, cannot be denied. But should we focus exclusively on the differences and disregard the continuities? Certainly, there is reason to fear that those who insist on the links between Christianity and Islam want to make the latter a mere outgrowth of the former, or even “a misinterpretation” (Fr. Nietzsche), with the ever-present theme of the “Mohammed-Imposter” lurking in the background.7 From this point of view, it would be naïve to think that Islamic studies has no connection with the war of religions, which has come back into fashion in recent years, in forms that are not merely euphemistic.8 Also, to oppose, as Chabbi does, the Islam of the Qurʾān, as a ‘pure’ construction, identical to its substratum and in harmony with it, to ‘imperial’ Islam (i.e. from the Abbasids onwards) which would be necessarily hybrid and thus alien to its origins, can lead to a determinist vision that prevents us from understanding the passage from one form to another. The history of this religion, which emerged under the name of “Islam,” is a series of interpretations and reinterpretations involving as many appropriations of the literary corpus at each successive period.
I must mention another singular interpretation, but one that is far removed from academia.9 It is the one suggested by Titus Burckhardt, the talented Swiss author, albeit on very shaky foundations. According to him, the architecture of the madrasa (four rooms opening onto a central courtyard, making a total of five) is reminiscent of the Meccan temple. On this subject, he speaks of “one of the structural laws of Islam:” that “groups of four principles or elements can be found in all the plans of Islam, which generally relate to a fifth term, their centre or foundation.” Among the examples he lists is that of the Five Pillars, an expression that does not appear in his writings: “there are […] four ‘pillars’ of religion, prayer, fasting, taxation and pilgrimage, linked to a fifth, the profession of faith […].”10 Burckhardt’s vision is not a problem formulated in historical terms, but belongs rather to what is often presented as the ‘phenomenological approach’, the most widespread expression of which is to be found in the numerous writings of Mircéa Eliade.11 This approach is based on the idea of a homo religiosus, a permanent and transhistorical form. Burckhardt’s interpretation is based on late modern Sufism, with elements of contemporary Western esotericism. It should be noted, however, that his conception of the Five Pillars, with the profession of faith (shahāda) at the centre, is by no means essentially Sufi: several jurist-theologians who were not Sufis at all, such as Ibn Rushd (the philosopher’s grandfather),12 cite the profession of faith as the condition for the validity of ritual observance.
To carry out this investigation more comprehensively, it would have been necessary to consider a larger number of sources than I have done here. As it is not my intention to provide a complete or definitive analysis, this is only an introductory study which serves to present the problem and highlight the interest of such an investigation, indicating avenues of future research and suggesting a few tentative hypotheses. One day soon, other researchers will go further, shedding more light on the history of this formula and no doubt on Islam’s religious system as well.
In his lectures given during the 1985–1986 academic year, Mohammed Arkoun called for the creation of a “vocabulary of Semitic institutions,” based on the model of Emile Benvéniste (himself a graduate of the Collège de France).13 I hope that one day we will have a serious survey of the religious vocabulary of Islam.14 Such an undertaking will necessarily be a collective one, since sources of all kinds and origins spanning many centuries will have to be taken into account. It will also have to consider the relationship between this vocabulary and that used in other Semitic languages. There is often a tendency in the field of comparatism to simplify matters by limiting the field of study to relations with Hebrew or Syriac and neglecting other Semitic languages, such as Akkadian.
A closer look at this minor question reveals that, compared with Christianity or even Judaism, Islam is a lesser-known religion because it is less studied, particularly in France, which, along with Great Britain, was the main ‘Muslim’ power in Western Europe for over a century.15 This is not the place for an overview of French Islamic studies, but the work of scholars like Louis Massignon and Henry Corbin, as important as it was, built up a mutilated image of Islam. They wanted to preserve from this foreign religion only that in which they believed they recognised themselves and which, in their view, was the only thing worth saving.16 We know that for Corbin, more radical than Massignon, his elder and mentor, fiqh – the main source used in this study – was “the canker of Islam.”17 Since then, another – and much worse! – sort of mutilation has been imposed by essayists of all kinds, reducing Islam to the contemporary avatars of Wahhābism.18
The aim of this study is to break with this distorted vision of the religious system of Islam, which is reduced to certain of its elements, either because they are considered the only ones worthy of study and hence the only ones worthy of salvation, or because they are treated as exhibits in the ‘trial’ brought against this religion.
Despite the fact that I was born in Algeria into a very devout Muslim family.
This research stems from my teaching at the Ve Section of the Ecole des Hautes Etudes, Sorbonne, during the first semester of 2015–2016. In particular, I looked at the Qurʾān, examining verses that might have certain analogies with the Pillars of Islam. I would, of course, like to thank my listeners at the time for their patience and their stimulating contributions.
In the following pages, I will sometimes write ‘Islam’ and sometimes ‘islām’. In the first case, I am referring to the religion as it is commonly designated. In the second case, it is a more technical usage: it refers to the manifest and public religion that theologians contrast with the inner religion, which they call īmān, a term that is usually translated as ‘faith’. The seat of ‘inner’ religion is the chest (ṣadr) or heart (qalb), while ‘outer’ religion manifests itself through the physical organs and limbs (mouth, hands and feet, eyes, ears, genitals, belly) and is referred to by the plural jawāriḥ. See Daniel Gimaret, La doctrine d’al-Ashʿarī, Cerf, 1990, p. 469 ff.
It is essential to point out that it was only from the third/ninth century onwards that the political and religious currents that have left their definitive mark on Islam to the present day began to take shape from a doctrinal point of view. In this, I follow in particular the work of Henri Laoust.
Paris, Seuil, 2016. Several years later, Abdennour Bidar, Inspector of Philosophy at the French Ministry of Education, published Les Cinq Piliers de l’Islam et leur sens initiatique, Paris, Albin Michel, 2023. Clearly, this book, whose true content is indicated by the final part of the title, is nowhere concerned with historical inquiry into the subject. Rather, it is an interpretation inspired by Sufism.
I wrote about this topic for Archives de Sciences sociales des Religions, “Bulletin bibliographique,” LXII, n° 180, 2017, 309a–312a and in “L’islam de Mahomet : une hypothèse sur les débuts de l’islam,” in Mehdi Azaiez (ed.), Le Coran: de la tribu à l’empire. Autour de l’œuvre de Jacqueline Chabbi, Louvain: Presses Universitaires de Louvain, 2023, 139–151. For an idea of the complexity of the problem when limited to the Qurʾān, see chapter 3 below.
According to G.S. Reynolds (The Qurʾān and its Biblical Subtext, New York: Routledge, 2010), one must search the Qurʾān for the subtext, i.e. make explicit the implicit reference to a particular Christian writing. As one enthusiastic commentator puts it: “the biblical milieu is an essential key to penetrating and understanding the Qurʾanic universe” (Emmanuel Pisani in MIDEO 36/2021, 372–3). Is such an approach not an arbitrary extrapolation? Moreover, and this is undoubtedly more problematic, does it not further discredit the way in which Muslims understand and have understood the Qurʾānic statements, dismissing it as error, or even nonsense? By making the Christian scriptures “the key” to understanding the Qurʾān, this same approach makes islām – which is based to a very large extent on Muslim exegesis – a religion founded on “errors” and “ramblings.” The old anti-Muslim polemic is thus being revived in the twenty-first century by conflating two distinct problems: that of certain textual sources of the Qurʾān, and that of the relationship of the Qurʾān to the religious system of islām as determined in part by Qurʾānic exegesis. What is important for understanding Islam as a religion is not the literary sources that the historian can point to, with the risk of introducing a false original/copy dichotomy, but to elucidate the links between the ways that Muslims understood and interpreted Qurʾānic statements and the institutions of Islam. We can look for ‘borrowings’ from ancient forms of Christianity in the Qurʾān and other Islamic scriptures, but the main concern is to understand what Muslims do with them. Moreover, while the Bible has lost its centrality in studies of the Ancient Near East (it was not so long ago that the period was studied through the lens of the Bible!), there have been attempts to give it a central role in Islamic studies.
This observation dates back to 2017.
W.M. Watt has made singular use of the expression “pillars of Islam.” In his Mahomet à La Mecque (Payot, 1958, 1977), when presenting his sources, he says of the great biographical dictionary of Ibn Saʿd (d. 230/845) that it contains the “biographies [of] the principal companions and the last ‘pillars of Islam’” (p. 15, Introduction 2). In this usage, the expression refers to the ancient authorities of Islam. As it happens, Ibn al-Ṣalāḥ (d. 642/1244) cites a comparable usage, but of the term rukn alone in his Muqaddima. Ahmad b. ʿUmayr, a Damascene traditionist, is introduced thus: kāna min arkān al-ḥadīth, “he was one of the pillars of hadith” (Muqaddima, ed. ʿ
Titus Burckhardt, L’art de l’islam. Langage et signification, Paris, Sindbad, 1985, p. 236. He refers to the speculations of Frithjof Schuon. On the current represented by Schuon, but also by R. Guénon, see M. Sedgwick, Contre le monde moderne, Paris: Devry, 2004 and on Burckhardt, see P. Ringgenberg, Les théories de l’art dans la pensée traditionnelle, Paris: L’Harmattan, 2011, pp. 559–629. I would like to thank Jean-Pierre Brach (Director of Studies at the EPHE), who kindly provided me with these references.
For a critique of this approach, see David G. Robertson, Gnosticism and the History of Religions, London, Bloomsbury, 2021. The main representatives of this movement in France include Henry Corbin and Gilbert Durand.
See note 24 below.
See Le Vocabulaire des institutions indo-européennes, 2 vols, Paris, Editions de Minuit, “Le sens commun” series, 1969.
This little book is intended as a contribution to the vast undertaking to come. As far as I am aware, even if widely-used working tools such as the Encyclopaedia of Islam or the Encyclopaedia of the Qurʾān contain some information on this subject, their aim is generally different.
Islam’s ubiquity in international news for more than a quarter of a century has not helped matters, since its study is strongly influenced by political and even military considerations. I cannot go into detail on this important point, which would require an accounting for the history of the whole region in its confrontation with Euro-Christian imperialism since at least the sixteenth century.
A quotation will give the measure of what I am saying here. Paul Nwyia, who was briefly director of studies at the Religious Sciences Section of the EPHE, and who undeniably followed in the tradition of Massignon, writes: “In the history of the Muslim world, there is only one great adventure which has universal value and which raises Islam to the level of an authentic search for the Absolute [my emphasis]: the adventure of the mystics commonly called Sufis” (Exégèse coranique et langage mystique, Beirut, 1970, p. 1). Sufism is undoubtedly an exciting “adventure,” but can Islam be reduced to this aspect alone if we are to strive to understand it and grasp its historical significance? We might add that it is of little importance to the historian of ideas or religions that those whose conceptions he is endeavouring to reconstruct are seeking the ‘Absolute’ (another name for the Christian god?). From this point of view, his work does not aim to paint a portrait of heroes. Knowing a religion, or more generally a human culture, implies abandoning the attitude of the collector.
Oral communication by Jean-Paul Charnay during a debate at the Institut du Monde Arabe, Paris, 2007.
Hence the resumption of a verbal “war” against Islam, waged by all sorts of individuals (editorialists, writers, etc.) who belong to reactionary circles and are rehabilitating colonialist themes.