The Paradox of Islamization Tombstone in

Leor Halevi
T H E PA R A D OX O F
I S L A M I Z AT I O N :
TOM B S TO N E
INSCRIPTIONS,
Q U R A N I C R E C I TAT I O N S ,
A N D T H E P RO B L E M O F
R E L I G I O US C H A N G E
Twenty or so years after the death of the prophet Muhammad in 632,

Abd al-Rahman ibn Khayr passed away. His name does not appear in the
annals of history and, but for an extraordinary record of his death, memory of his existence would have been altogether lost. His tombstone, once
hailed as the oldest known monument in the Islamic world, entreats the
reader to ask for Gods forgiveness:
In the name of God, the Merciful, the Compassionate, this grave belongs to
Abd al-Rahman ibn Khayr al-Hajri.
Forgive him, O God, and make him enter [Paradise] by your mercy, and let us
go with him.
Seek forgiveness for him whenever this inscription is read, and say Amen!
This preliminary study forms part of my forthcoming book on death rituals and the
making of Islamic society. I thank Roy Mottahedeh, Michael Cook, Wolfhart Heinrichs,
William Graham, and Lauren Clay for their critical comments and support in connection
with this project. I presented a version of this article at the November 2003 annual meeting
of the American Academy of Religion, in a panel dedicated to the theme, Popular Religion:
A Category and Its Problems. I thank David Frankfurter, the discussant, for his insightful
comments. I presented a second version of this paper at a March 2004 conference on Documentary Evidence and the History of the Early Islamic Mediterranean, and I am grateful to
Lennart Sundelin, one of the organizers of the conference, for his comments and a stimulating exchange of ideas. Finally, I would like to thank an anonymous reader at History of
Religions for his or her substantial critique, which forced me to dene problematic terms and
to rethink parts of my argument.
2004 by The University of Chicago. All rights reserved.

0018-2710/2004/4402-0002$10.00
History of Religions
121
This inscription was written in Jumada II of the year 31 (January or February

of the year 652 c.e.).1
Some skepticism has been expressed about the antiquity of this tombstone.
It has been suggestedmistakenlythat even if the date intended were
131 a.h., it would still be the earliest tombstone in Arabic.2 Actually,
there are a number of dated tombstones prior to 131 a.h. (749 c.e.), including a memorial embedded in the walls of a church in Cyprus, which
antedates Abd al-Rahmans tombstone by two years. In the name of
God, it reads, this is the grave of Urwa ibn Thabit, who died in the
month of Ramadan of the year 29 of the Hijra, or 650 by the Gregorian
calendar.3 There is, in fact, no good reason, to doubt the date of 31 a.h.,
which, after all, was written in stone.
The important objection to consider concerns not the date of Abd alRahmans tombstone but its categorization as an Islamic monument. If
the objective is to say something meaningful about the making of Islam,
we must wonder how justiable it is to label Islamic a tombstone that,
while referring to Allah explicitly and to the Hijri calendar implicitly, contains no reference to the prophet Muhammad and no allusion to Muslim
scripture. The tombstone records the death of a believer in Allah, yet it
1
Hussen Rached, Hassan al-Hawary, and Gaston Wiet, eds., Catalogue gnral du Muse
arabe du Caire: Stles funraires, 10 vols. (Cairo, 1932 42), vol. 1, no. 1: Bismillah alrahman al-rahim, hadha l-qabr li-Abd al-Rahman ibn Khayr al-Hajri [?]. Allahumma ghr
lahu wa-adkhil-hu [al-janna?] fi rahma minka, wa-iyy[a]na maahu. Istaghr lahu idha quria
hadha l-kit[a]b wa-qul amin. Wa-kutiba hadha l-kit[a]b Jum[a]da al-akhir[a] min sanat
spelled with ta maftuha] ihda wa-thal[a]thin. Henceforth I will refer to this work by the
abbreviation CMASF. The formula adkhilhu al-janna is rather common, so I have assumed
an ellipsis; literally, the text may be translated as let him enter into a state of mercy on your
part. All translations are my own, unless otherwise noted. On this tombstone, see Hassan alHawary, The Most Ancient Islamic Monument Known Dated a.h. 31 (a.d. 652), Journal
of the Royal Asiatic Society (1930): 32133. The ancient Kuc writing style and certain orthographic peculiarities (e.g., the spelling of sana with a ta maftuha), features that have
been discussed extensively by al-Hawary, indicate that we have at hand a remarkably early
inscription.
2
Yehuda D. Nevo, Zemira Cohen, and Dalia Heftman, Ancient Arabic Inscriptions from
the Negev (Jerusalem, 1993), 1:8, n. 20. Nevo mentions that other Egyptian tombstone texts
began only in the 170s a.h., a deduction he seems to have reached by consulting vol. 1 of
CMASF, which skips from the year 31 to the year 174 a.h. However, vol. 9 contains tombstones initially overlooked, dating from the years 71 and 111 a.h.
3
tienne Combe, Jean Sauvaget, and Gaston Wiet, Rpertoire chronologique d pigraphie arabe (Cairo, 1931), vol. 1, no. 5: Bismillah, hadha qabr Urwa ibn Thabit tuwufya
shahr Ramadan sanat tisa wa-ishrin lil-Hijra. Explicit reference to the Hijra is odd in such
an early inscription, and possibly an interpolation. Henceforth I will refer to this work by the
abbreviation RCEA. In the context of early epitaphs, it is also worth mentioning a report
pointing to a tombstone dated to 75 a.h. (694 c.e.), found among a cluster of Muslim tombs
under the Arc of Marcus Aurelius in Tripoli de Barbarie; see RCEA, no. 11. Below I will analyze tombstones from the years 71, 102, and 111 a.h.
122
Paradox of Islamization
otherwise lacks a distinctively Islamic identity. Its plea for divine forgiveness in the afterlife was commonplace in Jewish and Christian tombstones from late antiquity; and it is by no means clear that the inscribers
intention was to produce a uniquely Islamicrather than, more generally,
a monotheisticmemorial.4
Islamic markers, such as prayers for the prophet Muhammad and quotations from the Quran, emerged in the period between 690 and 720 c.e.,
and it was only by the 790s, as this article will show, that a formulaic pattern became established, including a standardized confession of faith.
These changes in the tombstone record reect a gradual process of Islamization, I will argue. To witness such a process unfolding is an exciting
matter for an historian, especially because of the scarcity of datable documents from the rst two centuries of the Islamic era.
This process of religious change, though widely rooted, became controversial in some circles. Upholders of traditions, pietistic ideologues
from the eighth century, began actively to oppose the popular practice of
inscribing tombstones. Despite the fact that the epitaphs contained pious
religious sentiments, traditionists decried the practice as a blameworthy
innovation that violated the customs of Medina, the city in Arabia where
God had revealed the new religion to the prophet Muhammad.
In this article, I will examine changes in the tombstone record from the
rst two centuries of Islam as well as the traditionist literature directed
against tombstones. An analysis of this record, which historians of early
Islam have largely overlooked,5 is in and of itself worthwhile, for several
4
Arguably, although the earliest dated references to the term would appear only several
decades after Abd al-Rahmans death, this epitaph commemorated the death of a Muslim.
Possibly, however, it belonged to someone who self-identied as a Muhajir (Emigrant) or
a Mumin (Believer) and was identied by others (polemically) as a Son of Hagar. There
is, in any case, no reason to assume a priori that every Muslims epitaph was intended, in any
deliberate way, as an Islamic memorial. Historians working on cognate elds have also
upheld this sort of distinction. Thus, for instance, Paul-Albert Fvrier, La mort chrtienne,
in Segni e riti nella chiesa altomedievale occidentale, 1117 aprile 1985, Settimane di Studio del Centro Italiano di Studi SullAlto Medioevo, no. 33 (Spoletto, 1987), 2: 881942,
aims to show how la mort du chrtien est devenue une mort chrtienne. Ross S.
Kraemer, Jewish Tuna and Christian Fish: Identifying Religious Afliation in Epigraphic
Sources, Harvard Theological Review 84, no. 2 (1991): 14162, makes the point that inscriptions attributed to Jews should not necessarily count as Jewish. Reference to the preIslamic plea for forgiveness is given below.
5
Tombstones have attracted almost exclusively the attention of epigraphists, not of historians. See, e.g., the bibliography in Encyclopaedia of Islam, New Edition, s.v. Kitabat.
Exceptionally, Fred Donner, Narratives of Islamic Origins: The Beginnings of Islamic Historical Writing (Princeton, N.J., 1998), pp. 8589, cites a number of early inscriptions in his
discussion of early Islamic piety. Robert Hoyland, The Content and Context of Early Arabic Inscriptions, Jerusalem Studies in Arabic and Islam 21 (1997): 77102, analyzes thematically an extensive corpus of inscriptions from the rst two centuries of Islam, yet reaches
a conclusion very different from mine about the possibility of using these sources as markers
of religious change. Yusuf Raib, Les pierres de souvenir: stles du Caire de la conqute arabe
123
reasons. First, tombstones hint at the changing role of the Quran in the
religion of early Islam. Quranic quotations are rst attested on tombstone inscriptions from the rst quarter of the eighth century; by the end
of that century, tombstones began to display fairly elaborate quotations,
suggesting a rise in the ritual uses of the Quran. They reveal a facet of
an emerging popular practice, that of reciting scripture in a ceremony of
intercession for the dead. This is a practice about which early Islamic literature has remarkably little to say. Given how limited our knowledge is
about the emergence of a Quranic liturgy in the formative period of Islam,
this documentary evidence is signicant.
Second, dated tombstone inscriptions provide an independent record
of religious trends, suggesting an alternative standard for periodization to
the chronology followed by historians of the early Islamic polity. In the
context of political narrative, it makes sense to adopt the year 750 c.e.,
when the Abbasids overthrew the Marwanid dynasty from power, as a
revolutionary turning point. But this date has no signicance in the tombstone record, which suggests that key changes in the practice of Islam
occurred at other times, irrespective of political developments, around the
years 691, 721, and 795 c.e.
Beyond these two contributions, an examination of the epitaphs in
conjunction with the relevant traditionist literature will offer something
unique: the opportunity to see, on the one hand, how popular practice came
to diverge from orthodox ideals and to understand, on the other hand,
how the traditionists who despised change reacted to an emerging religious practice. Our focus, in the end, will be on the productive tension
between popular piety and orthodoxy. This tension led not only to the development of the traditionist mentality that came to characterize Islam,
but also, and perhaps more importantly, to the formation of an alternative
religiosity.
Before proceeding, however, a caveat must be issued on the limitations
of the evidence to be considered. Dated tombstone inscriptions from the
Islamic period come primarily from the arid lands of Egypt. In recent
years epigraphists have begun collecting inscriptions from various other
sites in the eastern Mediterranean world and from the Arabian peninsula,
yet so far most of their discoveries date from a later period, beginning
circa 850 c.e. As a result, the tombstones to be analyzed reect primarily
changes in the practice of Islam in early Islamic Egypt. In this connection,
it is also worth stating the obvious: that this history is based on tombstones
la chute des Fatimides, Annales islamologiques 35 (2001): 32183, offers a valuable discussion of Egyptian epitaphs, covering various themes ranging from decorative motifs to
naming patterns.
124
already discovered. New discoveries may alter in ways minor or signicant the chronology to be developed below.
Several key terms I use require denition. By Islamization I do not
mean conversion to Islam, nor do I mean the transformation that Muslims
must undergo to adhere more strictly to preexisting, already-established
Islamic norms. Rather, I use the term to refer to the historical process at
work during the formative era of Islam, by which persons and objects
were made Islamic in character and became imbued with Islamic principles or forms.6 In discussing Islamization in reference to the emergence
of Islam, the historian must be careful not to distinguish arbitrarily between Islamic and non-Islamic forms. Attention should be drawn to forms,
such as Quranic quotations, that would have been readily recognizable
as Islamic by Muslims. In this spirit, I will refer here to Quranic citations as key markers of the process of Islamization; at the same time I
will show why, from a traditionist perspective, it would seem inappropriate to describe tombstones, even those including Quranic quotes, as
adhering to an Islamic form. This tension between a popular and an orthodox view of what should count as Islamic will enrich our understanding of this process, which resulted in the making of Islam.
By popular in this article I mean practices that were relatively widespread in society yet disapproved of by Islamic law. It could be argued
that tombstones and tombstone inscriptions, if prohibitively expensive
for the poor and relatively incomprehensible to the illiterate, were part of
elite rather than popular practice. Yet my use of the term popular
here is based not on such distinctions but on religious, dogmatic considerations. I use the term orthodox in opposition to popular to designate religiously sanctioned ritesrites approved by jurisprudents in view
of Muslim tradition and custom. Nevertheless, it is important to note that
certain orthodox rites were popular in the sense that they were widely
practiced. Technically speaking, then, the distinction I have in mind is
between orthopraxis and heteropraxis.7 Let us turn, rst, to the process of
Islamization as it unfolded in popular practice.
6
The Oxford English Dictionary denes the verb Islamize rather narrowly, yet provides
more expansive denitions, which I have here adapted, of Christianize and Judaize. It
may seem tautologous to speak of the process of Islamization in reference to Muslims;
such a formulation, if awkward, is necessary to convey the notion that men and women who
identied themselves as Muslims were involved in the making and shaping of Islam.
7
On various uses of the term popular in reference to religion, see Jacques Berlinerblau,
Max Webers Useful Ambiguities and the Problem of Dening Popular Religion, Journal
of the American Academy of Religion, 69, no. 3 (2001): 60526; J. Waardenburg, Ofcial
and Popular Religion as a Problem in Islamic Studies, in Ofcial and Popular Religion:
Analysis of a Theme for Religious Studies, ed. Pieter Vrijhof and Jacques Waardenburg (The
One Line Short
125
i. the process of islamization

The tombstones of Urwa ibn Thabit and of Abd al-Rahman ibn Khayr,
who died in 650 and 652, respectively, display no distinctively Islamic
formulas, as we have seen. Four decades after their deaths, an Egyptian
monument commemorating the death of Abbasa ibnat Jurayj in 691 c.e.
(71 a.h.) begins to show Islamic signs.
In the name of God, the Merciful, the Compassionate.
The greatest of misfortunes for the people of Islam (ahl al-Islam)
is their loss (musiba) of the Prophet Muhammad,
may God pray for him and grant him salvation.
This is the tomb of Abbasa, the daughter of
Jurayj, the son of Sanad, may the mercy
of God, his forgiveness (maghra) and good will be upon her.
She died on Monday, with fourteen days having passed of Dhu al-Qada,
of the year Seventy One (April 21 of the year 691 c.e.).
And she confesses (wa-hiya tashhadu) there is no God but Allah,
alone, having no partner, and that
Muhammad is his slave and messenger,
may God pray for him and grant him salvation.8
This inscription of the Muslim daughter of Jurayj, or Little George, a man

likely of Christian Coptic roots, is recognizably Islamic. At the closing
Hague, 1977), pp. 34086, suggests that scholars distinguish between normative, alternative, and popular Islam. Compare Frederick M. Denny, Islamic Ritual: Perspectives and
Theories, in Approaches to Islam in Religious Studies, ed. Richard Martin (Tucson, Ariz.,
1985), pp. 6377, who argues that the distinction between ofcial and popular Islam is
in some respects a red herring. In challenging the use of these terms Denny has a good
point. Unfortunately, he does not explain by what terminology we might distinguish ibadat
(orthodox religious rites) from bida (practices classied as dangerous innovations). On the
legal literature regarding ritual innovations, see Maribel Fierro, The treatises against innovations (kutub al-bida), Der Islam, 69 (1992): 204 46; and Boaz Shoshan, High Culture
and Popular Culture in Medieval Islam, Studia Islamica 73 (1991): 67107, at 8994.
8
CMASF, vol. 9, no. 3201. For a transcription and discussion, see Hassan al-Hawary, The
Second Oldest Islamic Monument Known Dated a.h. 71 (a.d. 691), Journal of the Royal
Asiatic Society (1932): 28993. There is some uncertainty about the reading of the name as
Jurayj ibn Sanad. Hoyland, Content and Context, p. 87, n. 65, suggests that the tombstones date should possibly be understood as 171 a.h. (788 c.e.). He bases this suggestion
in part on the fact that the phrase representing the death of Muhammad as the greatest calamity is not otherwise seen until 190 a.h. (806 c.e.). However, the tombstone of 190 a.h.,
which he cites by comparison (CMASF, vol. 1, no. 16), differs signicantly from the tombstone of Abbasa ibnat Jurayj, as it includes three Quranic citations, 3:16 (3:18), 22:7, and
9:33, whereas Abbasas includes none. In his article, al-Hawary argues for great resemblance in script between Abbasas epitaph and the famous monuments from Abd al-Maliks
reign.
126
we have an unequivocal confession of faith. The use of the imperfect

tense for this act in the phrase, wa-hiya tashhadu, indicates an interesting
possibility: that Abbasa continues confessing the Muslim creed after
death, in the present, by means of her written testament.
The lament over the death of Muhammad as a calamity (musiba) has a
distinctive elegaic quality, characteristic of Arab mourning. The most
striking element is the reference to the Muslim peoples as a family
(ahl ), perhaps understood as a branch of the ahl al-kitab, the people to
whom a sacred text had been revealed. Ahl al-Islam is not a generic designation in the way Muslims is. It also lacks the political overtones of
umma, or community. At work in the inscription is a different, perhaps
more intimate type of communal identity. While retaining a formal tie
to her father, who may well have paid for her tombstone, this Muslim
daughter also developed a sense of kinship with the greater family of
Muslims, who mourned at her graveside the death of the Prophet Muhammad as if it had happened yesterday, though two generations had
passed.
There may have been other commemorative stones much like Abbasas,
but these did not survive the vicissitudes of time. Abbasas tombstone
is a unique historical document. It expresses eloquentlyperhaps more
eloquently than any other datable record from the material culture of the
rst century of Islaman individuals sense of communal identity, from
a unique nongovernmental perspective and in an age for which we possess precious little evidence.
One may suggest, on the basis of the scant evidence available, that a
fundamental transformation occurred over the course of four decades of
the seventh century, between the deaths of Urwa in 650 and of Abbasa
in 691 c.e.9 This transformation culminated in the development of a selfconscious articulation of a Muslim communal identity, realized by a
rather complex operation. Abbasas tombstone seeks to induce in us a
mournful response to the death of the prophet Muhammad and to persuade the reader who considers him or herself part of the family of Islam
to plead for Gods forgiveness on behalf of a Muslim daughter.
Abbasas epitaph is in many ways exceptional. It differs signicantly
not only from earlier tombstones, as we have seen, but also from later
ones. Certain of its phrases would become formulaic: the basmala, an invocation of God, and the shahada, a testimony of the Muslim creed. The
9
Compare Patricia Crone and Michael Cook, Hagarism: The Making of the Islamic World
(Cambridge, 1977). This revisionist history recognizes the rst redoubtable signs of Islamic
self-expression in the numismatic and architectural evidence stemming from the reign of the
caliph Abd al-Malik, during which Abbasa died.
127
lament over the loss of Muhammad would also appear with some frequency in subsequent tombstones.10 Yet Abbasas tombstone lacks one
important element characteristic of eighth-century commemorative monuments: a quotation from the Quran.
Quranic citations are a key marker of the process of Islamization. Quotations of this kind became commonplace in epitaphs of the late eighth
century and thereafter. It is striking, however, that not a single one of the
handful of dated tombstone inscriptions from the rst century of Islam
boasts a quotation from the Quran. The earliest citation known to me belongs to Fatima ibnat al-Hasans tombstone, who died in 721 c.e. (102
a.h.) and afrmed the sovereignty (mulk) and omnipotence of God (huwa
ala kull shay qadir) precisely according to the formula in Quran 67:1.11
Another one of the earliest citations afrmed the oneness of God, in a
tombstone of 729 c.e. (111 a.h.), precisely according to the formula in
Sura 112.12
It is worth pausing a moment to consider this evidence in light of the
scholarly debate on the canonization of the Quran. According to one
expert on inscriptions of the Negev, the rst instances of Quranic paraphrases arrived on the scene in the late eighth century. The language of
the Umayyad inscriptions is patently non-Quranic, he afrmed, and endeavored on this basis to corroborate John Wansbroughs arguments about
the canonization of the Quran in Abbasid times, that is, around the turn
of the ninth century.13 Unfortunately, these bold statements, applicable
though they might be to the Negev inscriptions, fail to take into consideration the material cited here.
Wansbrough speculated that Quranic logiathat is, discrete statements
disjoint from narrativeoriginated in a multiplicity of communities and
regions. They served a polemical and a liturgical function, in his view,
before they were somehow gathered and canonized in a text reecting
10
See Gaston Wiet, Stles Couques dgypte et du Soudan, Journal Asiatique 240
(1952): 27397, at 27986. Solange Ory, Aspects religieux des textes pigraphiques du
dbut de lIslam, Revue du monde musulman et de la Mditerrane 58 (1990): 3039, examines the appearance of the basmala and the shahada as signs of Islamization in lands conquered by the Arabs.
11
Abd Rahman M. Abd al-Tawab, Stles islamiques de la ncropole d Assouan, 3 vols.
(Paris, 197786), vol. 1, no. 1.
12
CMASF, vol. 9, no. 3202.
13
Nevo et al. (n. 2 above), pp. 89. For a recent article criticizing the revisionist chronology, see Harald Motzki, The Collection of the Quran: A Reconsideration of Western Views
in Light of Recent Methodological Developments, Der Islam 78 (2001): 134; also see
Wadad al-Qadi, The Impact of the Quran on Arabic literature during the Late Umayyad Period: The Case of Abd al-Hamids Epistolography, in Approaches to the Quran, ed. Gerald
R. Hawting and Abdul-Kader Shareef. Compare Patricia Crone, Two Legal Problems Bearing
on the Early History of the Quran, Jerusalem Studies in Arabic and Islam 18 (1994): 137.
128
polygenesis rather than diffusion through historical contact. Muslim scripture, he argued provocatively, betrays a gradual process of canonization
that must have taken several generations. Specically, he proposed that
Quranic canonization occurred because of interconfessional polemics that
extended to the end of the eighth century and the beginning of the ninth
century.14
Precise Quranic quotations on eighth-century tombstones call into question Wansbroughs chronology as well as his theory about polygenesis.
The tombstones of 721 and 729 c.e. were both discovered in the cemetery of Aswan, a city on the Nile River in southern Egypt. Since they
include exact citations from the Quran as we know it, we must suspect
that a copy of this text had reached southern Egypt by the third decade of
the eighth century. Indeed, geographically widespread quotations from the
Quran on eighth-century tombstones are indicative of a process of scriptural diffusion, not of polygenesis. These tombstones, I should clarify, do
not reveal anything about the collection of the standard text in the socalled Uthmanic codex. Rather, they serve to illustrate when this authoritative text, whose prehistory is a matter of speculation, became established
in various regions throughout the Islamic world.
In this context, it is worth examining not only exact Quranic quotes in
epitaphs but also paraphrases or variants.15 Occasionally inscriptions from
the end of the rst century of Islam contain material that sounds Quranic
but does not conform to the standard text. Thus, for instance, the grafto
of Abd al-Malik ibn Ubayd, which dates from 92 a.h. (710 c.e.), reads;
Pardon his [cumulative] sin, the preceding and the forthcoming parts of
it (ighr lahu dhanbahu ma taqaddama minhu wa-ma taakhkhara).16
14
John Wansbrough, Quranic Studies: Sources and Methods of Scriptural Interpretation
(Oxford, 1977), pp. x, 3352, 78, and The Sectarian Milieu: Content and Composition of
Islamic Salvation History (Oxford, 1978), pp. 14, 45, 48. One must be cautious in relating
Quranic inscriptions to Wansbroughs theories about the evolution of Muslim scripture. The
existence of isolated Quranic quotes from the early eighth century calls into question, but
does not invalidate, his suggestion that canonization of a deliberate edition, the style and
structure of which represents the application of considerable literary technique, took place
only around the turn of the ninth century.
15
Wansbrough, Quranic Studies, p. 21, suggested that variant traditions present in the
Quran indicate the existence of independent, possibly regional, traditions incorporated
more or less intact into the canonical compilation, itself the product of expansion and strife
within the Muslim community. According to this model, one would expect the tombstone
record to illustrate how different regions subscribed to different, eventually canonized traditions. This is not what the record shows, however. It may reect the existence of regional
variants, as in the case of Quran 33:56 at Asham (discussed below), but these variants were
not incorporated into the Uthmanic text.
16
Antonin Jaussen, Mission archologique en Arabie: marsmai 1907 (Paris, 1922),
3:100102. On Quranic grafti, see the preliminary study by Frdric Imbert, Le Coran
dans les grafti des deux premiers sicles de lHgire, Arabica 47 (2000): 38190.
129
This formula corresponds to Quran 48:2 (li-yaghra laka llahu ma taqaddama min dhanbika wa-ma taakhkhara), but not exactly. Unfortunately, in this as in most such cases, it is impossible to determine whether
Abd al-Malik was paraphrasing the Uthmanic text or quoting precisely
from a non-Uthmanic version.17
Early Islamic epitaphs from the environs of Asham, an Arabian village
in Tihama on the trade and pilgrimage route between the Hijaz and the
Yemen, insert the Prophets name, Muhammad, into what is otherwise
a precise quotation of Quran 33:56. Signicantly, a later quotation of this
verse from the same region, deriving from the late eighth or early ninth
century, omits the Prophets name, as if in order to conform to the Uthmanic version.18 By concentrating on such minor changes we gain some
insight into an important process: the dissemination of the canonical version of the Quran to multiple locations ranging from Aswan to Asham.
This process, the epitaphs indicate, began no later than the third decade of
the eighth century and continued in force for several decades afterward.
Quranic epitaphs serve also to illustrate a neglected issue in Quranic
studies: the changing role of Muslim scripture in everyday rituals. They
illustrate this role in a unique way, in fact, because these epitaphs, unlike
coins with quotations from the Quran, belong to a special eld of ritual
activity, the cemetery, and because they contain prayers for the dead. On
the basis of the tombstone record, I would tentatively date the origins of
this liturgical function to the third decade of the eighth century, though
with one reservation: the tombstones from the 720s contain only snippets
from the Quran, which in and of themselves do not necessarily indicate
ritual activity. It is only by the late eighth century, when Quranic quotations proliferate on tombstones, that we have evidence for the liturgical
(or quasi-liturgical) uses of the Quran at the cemetery.19
17
For a straightforward discussion of the issue, see Michael Cook, The Koran: A Very
Short Introduction (Oxford, 2000), pp. 12022.
18
Hasan al-Faqih, Mawaqi athariyya Tihama, vol. 1, Mikhlaf Asham (Riyadh, 1992),
pp. 353, 388; cf. p. 361; editors discussion on pp. 33133. The collection includes one dated
tombstone from 157 a.h. (774 c.e.), which should serve as a control (see p. 360). Al-Faqih
dates the tombstones either to the rst or to the second century of the Islamic era, but this division does not seem justied. A few of the tombstones he ascribes to the rst century may
well belong to the middle of the second, and those he ascribes to the second may well belong
to the beginning of the third. This variant of Quran 33:56 does not correspond to any established variant reading (qiraa) recorded by Ibn Mujahid (d. 935 c.e.), Kitab al-saba lqiraat (Cairo, 1972), pp. 51824. Nevertheless, variant readings of Quran 33:56 did exist;
see Arthur Jeffery, Materials for the History of the Text of the Quran (Leiden, 1937), pp. 76,
262, 298. Compare the citations of Quran 33:56 in CMASF (n. 1 above), vol. 1, nos. 14
and 24.
19
Cook, The Koran, pp. 79 and 142, distinguishes between liturgical and what he calls
semi-liturgical uses of the Quran, depending on whether or not the recitations took place
130
By the late eighth century, the Quran would play a key role in acts of
devotion toward the dead. To appreciate this role and to gain further insight into the process of Islamization that unfolded over the course of the
second century of the Islamic era, let us turn to a typical epitaph from the
end of the eighth century. Rabia ibn Maslamas tombstone from the year
795 c.e. (179 a.h.) resembles many ninth-century epitaphs, in that it quotes
two popular Quranic verses. Yet it differs signicantly from the tombstone of Abbasa, who had died about a century before Rabia.
In the name of God, the Merciful, the Compassionate,
This is what Rabia ibn Maslama ibn Hunata al-Sadafi testied.
He confessed (yashhadu) that there is no God but Allah
Alone, who has no partner, and that Muhammad is his slave and messenger.
He sent him with guidance and right judgment in order to triumph over all of
religion (li-yuzhirahu ala l-din kullihi ),
[And he confessed] that the Garden is real and the Fire is real, and that the
Hour [of Judgment] is coming, no doubt about it, and that God shall
resurrect those who are in the graves [cf. Quran 22.7].
He believed in Gods omnipotence (al-qadar)all of it, the good and the evil,
according to it he lived and according to it he died, and according to it he
will be resurrected alive, God willing.
O God, insinuate to him his proof (laqqinhu hujjatahu), and ll with light his
grave (nawwir alayhi qabrahu),
And may the abode of your pleasure be recognized between him and his
Prophet.
[May] the mercies (rahamat) of God [be] upon him, and his forgiveness
(maghra) and satisfaction, he passed away on Sunday, with six nights left
to Muharram [i.e., the twenty-fourth] of the Year 179 [April 19, 795 c.e.].20
I have chosen this particular inscription because, though conventional or

ordinary, it is altogether an apt representative of the emerging pattern. It
includes quotations from two of the most favored verses in ninth-century
in the ritual of prayer. I will refer below to the liturgical function of the Quran at the
cemetery, yet the reader should bear in mind that, from a certain perspective, this use of the
Quran would appear quasi-liturgical in that it did not form part of the prescribed form of
service at the cemetery. On the potential liturgical function of Meccan verses before canonization of the Quran and on the subsequent loss of this function, see Angelika Neuwirth, Studien zur Komposition der mekkanischen Suren (Berlin, 1981), pp. 263, 316, and passim, and
Du texte de rcitation au canon en passant par la liturgie, Arabica 47 (2000): 194229. The
focus here will not be on the liturgical role Quranic logia might have played prior to canonization, but rather afterward. In the eighth century, when Muslim scripture had already reached
standard form, it began to serve a novel ritual function in various cemeteries throughout the
Islamic world. This development is our concern.
20
CMASF, vol. 1, no. 3.
131
Egyptian epitaphs, 9:33 and 22:7.21 The relevance of verse 22:7 is immediately evident: unsurprisingly, it conrms Rabia ibn Maslamas faith in
the resurrection of the body at the last hour. The signicance of verse 9:33
is more difcult to unravel. In the view of early Islamic exegetes, the verse
celebrated the inexorable triumph of Islam over other religious communities (milal); the historical rivalry between Islam and Christianity played
a prominent role in apocalyptic interpretations of it.22
Why was this verse of relevance in Egypt in the late eighth century and
in the ninth century, given that the land had been conquered in the seventh century? Perhaps its celebration of the triumph of Islam over other
religions coincided with the period of mass conversions to the religion in
power. Alternatively, given that many of these tombstones boast an Arab
lineage for the deceasedgoing back to the generation of the father and
the grandfather, sometimes even to the great-grandfatherperhaps the
tombstones celebrated instead the historical triumph of Muslims of Arab
origin in a land that had once been ruled by Christians and where a high
proportion of the population (surely still the majority) had not yet converted to Islam.23
In the tenth century verses 9:33 and 22:7 do not fall out of favor, but
they do decline drastically in popularity as a new trend arises in the record,
with verses 41:30, 67:12, and Sura 112 become most popular.24 These
21
Thus, out of a total of eighty-three Quranic citations from the tombstone record of
Aswan in the years 721870, corresponding to vol. 1 of Abd al-Tawab (n. 11 above), there
are forty citations to Quran 9:33 and 30 to Quran 22:7. In CMASF, vol. 1, devoted to the
years 790854, the three verses most frequently cited are Quran 9:33, 22:7, and 3:16 (3:18).
Citations to Quran 9:33 and 22:7 often appear on the same tombstone.
22
See al-Tabari, Jami al-bayan an tawil ay al-Quran, 2d ed. (Cairo, 195468), at 9:33
for this and other slightly divergent interpretations; Hud ibn Muhakkam, Tafsir Kitab Allah
al-Aziz (Beirut, 1990), 2:12728.
23
Tombstones may serve as indicators of the curve of conversion, especially in Egypt, for
its arid land preserved an extensive record that may properly be used for statistical analysis,
though only beginning in the late eighth century. The haphazard record from the seventh and
early eighth centuries can suggest trends but does not provide data of any use for cliometric
research. On the conversion of Egypt to Islam, see Ira Lapidus, The Conversion of Egypt to
Islam, Israel Oriental Studies 2 (1972): 24862; and Richard Bulliet, Conversion to Islam in
the Medieval Period: An Essay in Quantitative History (Cambridge, 1979), chap. 8. See also
Jonathan M. Bloom, The Mosque of the Qarafa in Cairo, Muqarnas 4 (1987): 720, at 12.
24
The most dramatic comparison can be drawn between the course of Quran 9:33 and
Quran 112 at Aswan. In the middle of the ninth century (ca. 82070 c.e.), Quran 9:33 was
the most popular of two extremely popular verses, earning almost 50 percent of all citations.
By the rst half of the tenth century (ca. 91050 c.e.), Quran 9:33 had declined in popularity
to under 5 percent of citations. In the same time frames, Quran 112 jumped from zero citations to being the most popular verse, but won with only 20 percent of the votes. A similar
contrast can be drawn between Quran 22:7 and Quran 41:30. (These observations are based
on a comparison of citations from vol. 1 and vol. 3 of Abd al-Tawab. CMASF, vol. 4, shows
that Quran 3:16 (=3:18), 9:33, and 22:7, remained popular in the years 885927 c.e., yet
began to face serious competition from Quran 67:12 and Quran 112.
132
verses invoke the oneness of God, his omnipotence, and the consolation
of the angels to those who have been promised an otherworldly garden.
For Christian converts to Islam as well as for Muslims living in lands
populated by Christians, Sura 112 had special importance, since it specied that God had begotten no one. Yet this verse, though polemical,
neither celebrated the historical triumph of Islam, as Quran 9:33, nor did
it draw an explicit distinction between Muslims and non-Muslims. The
declining popularity of verse 9:33 may well reect an era when most of
Egypt had already converted to Islam.
The key point, however, is that beginning in the late eighth century a
limited number of Quranic verses began to appear regularly in tombstone
inscriptions. This shift indicates a certain degree of formalization in the
religion of Islam, the emergence of standard patterns that did not exist but
a half century earlier. It also reects an important change in the orientation of Islam in the mainstream. Finally, in the late eighth century, Muslims turned in large numbers to commemorate their dead with Quranic
inscriptions, ostensibly believing in the efcacy of Muslim scripture in
prayers for the deceased. To appreciate this change in the relation of
tombstones to the ritual of interceding for the dead, let us consider the
transformation that occurred between 650 and 800 c.e.
All tombstone inscriptions prompt the literate passerby to readprobably out loud.25 An illiterate audience wielding active competence in
Arabic, or commanding the language for liturgical purposes, could easily
participate in the oral performance at key moments, when prompted by
the reader. (This is an important qualication because it suggests how the
illiterate and semiliterate could participate in a ritual that required liter-
25
There is no reason to expect that qaraa designated silent reading except where specied otherwise. See, e.g., Abu Ghanim, Al-Mudawwana al-kubra (Beirut, 1974), 1:210, quoting Abu Ubayda (ibn al-Jarrah, d. 639 c.e.), who recommended that one read the Quran in
the funerary prayer but inwardly (baynaka wa-bayna nafsika). Similarly, according to alTahawi (d. 933 c.e.), Ibn Shihab al-Zuhri (d. 742 c.e.) reported on the authority of Abu
Umama ibn Sahl ibn Hunayf (d. 718 c.e.): One of the Prophets Companions related to him
that the sunna in the funeral prayer was for the Imam to praise God, then to read the Fatiha
privately to himself (sirran f i nafsihi ). See Ibn Rushd, Bidayat al-mujtahid wa nihayat almuqtasid (Beirut, 1996), 3:3536. Compare Abd al-Razzaq, Al-Musannaf, 2d ed., vol. 3
(Beirut, 1983), KJ, pp. 48990, no. 6428. The norm at funerary prayers, despite these minority reports, was to recite the Quran audibly. Recent scholarship on the Quran has emphasized the importance of liturgical recitation. In this regard, see Neuwirths work cited above;
William A. Graham, The Earliest Meaning of Quran, Die Welt des Islams 2324 (1984):
36177, Quran as Spoken Word: An Islamic Contribution to the Understanding of Scripture, in Approaches to Islam in Religious Studies, ed. Richard Martin (Tucson, Ariz., 1985),
pp. 23 40, and Beyond the Written Word: Oral Aspects of Scripture in the History of Religion (Cambridge, 1987), pp. 79115; Asma Asfaruddin, The Excellencies of the Quran:
Textual Sacrality and the Organization of Early Islamic Society, Journal of the American
Oriental Society 122 (2002): 124.
133
acy.)26 In any case, without a reader, it is difcult to imagine much of a

ceremony. With a reader in mind, however, we can easily hear him or her
identifying the grave of Abd al-Rahman ibn Khayr, who died in the year
31 a.h. (652 c.e.), and uttering: Seek forgiveness for him whenever this
inscription is read, and say Amen! with the congregation echoing
these words and pleading on behalf of a particular dead man. In this case,
the tombstone provides explicit ritual guidance: it beckons or beseeches
the reader not merely to read the epitaph passively but actively to intercede with God in favor of the deceased.
It is notable that already in the 650s writing and reading occupied a
role in Muslim public life. The tombstone of 652 c.e., which beckons us
to read, undeniably calls for a religious performance of sorts, the ultimate objective being to persuade God to forgive Abd al-Rahman (Slave
of the Merciful) and grant him entrance into the Garden. This ceremony
of seeking Gods forgiveness for the sins of an individual was one of the
central aspects of Muslim piety during the rst century of Islam, though
it was not original to Islam.27
This ritual of intercession began to change in a signicant way in the
720s, as the tombstones from this period triggered a brief recitation derived from the Quran. However, due to the brevity and lack of uniformity of their Quranic citations, a ritual of Quranic recitation cannot be
readily inferred from these tombstones. Only by the late eighth century,
when Quranic quotations proliferated on tombstones such as the one from
795 c.e., does it become fairly clear that a quasi-liturgical service involving Quranic readings was taking place in the cemetery, during visits to
particular graves. These Quranic inscriptions served, in a sense, to mediate between the living and the dead.
Outside evidence serves to prove the suggestion that, in the late eighth
century, a Quranic liturgy developed at the cemetery. Opposition to the
reading of the Quran in the funerary prayer by Malik (ca. 71295 c.e.),
Abu Hanifa (d. 767 c.e.), and al-Awzai (d. 773 c.e.), indicates that the
26
On the participation of the illiterate in a culture of literacy, see in a different context the
remarks of Roger Chartier, Culture as Appropriation: Popular Cultural Uses in Early Modern
France, in Understanding Popular Culture: Europe from the Middle Ages to the Nineteenth
Century, ed. Steven Kaplan (Berlin, 1984), pp. 22953, at pp. 241 ff.; and David Frankfurter,
The Magic of Writing and the Writing of Magic: The Power of the Word in Egyptian and
Greek Traditions, Helios 21, no. 2 (1994): 189221, at 196 ff.
27
See Donner (n. 5 above), p. 86, citing an inscription dated to 64 a.h./684 c.e.: Lord of
Gabriel and Michael and Israfil, grant forgiveness to Thabit ibn Yazid al-Ashari for his past
and future sins. Also see Hoyland (n. 5 above), p. 89, citing a Greek formula used in Jordan
during the sixth and seventh centuries: Kyrie synchrson tas [h]amartias autou (O Lord
forgive his sins). Nevo et al. (n. 2 above), pp. 89, has discussed later versions of this formula. He claims it is commonplace, monotheistic fare, rather than a Quranic paraphrase but
does not cite Jewish or Christian inscriptions containing the topos of early/late sins.
134
Quran was, in fact, recited by the late eighth century during the funerary
ceremony. Their criticism of Quranic readings as an innovation deviating from the right practice (amal ) of Medina underscores the trend on
display in the tombstone record.28 This funeral prayer, which was issued
once and immediately before burial, must not be confused with the intercessory prayer, which was rooted in the epitaphs and delivered repeatedly
after burial. Both concerned the well-being of the deceased, however, and
by the late eighth century both were subject to popular pressure for an
enhanced role for the Quran.
As the practice of uttering Quranic prayers for the dead gained popularity in various cities of the Islamic world outside of Medina, certain
jurisprudents decided realistically to adjust to the times. Al-Shai (767
820 c.e.) in particular recommended reading at funerals the opening verses
of the Quran, though he was aware of the fact that some people opposed
these recitations. As a result, he advised that opponents be told, you have
diverged from the sunna! (or revered practice). Other pietists, siding
with al-Shai, produced oral traditions on the authority of Muhammad
to support Quranic recitations, which they perceived if not as the venerable practice of the Muslim ancestors, then as the contemporary practice (amal ) in perfection of the sunna (tamam al-sunna). Ibn Hanbal
28
An abbreviation used in the footnotes below, KJ, stands for Kitab al-Janaiz, the Book
on Funerary Practice. The number following KJ often refers, not to a given page, but to a
specic oral tradition. Malik, Al-Muwatta in the recension by Yahya ibn Yahya al-Laythi alAndalusi, vol. 1 (Beirut, 1996), KJ, no. 611, related from Na that Abdallah b. Umar was
not accustomed to reciting [the Quran] during the prayer over the bier (kana la yaqrau
l-salat ala l-janaza); see also Al-Muwatta (li-Malik b Anas) in the recension by Suwayd
b. Said al-Hadathani (Beirut, 1994), KJ, p. 315. Also on Maliks authority, Sahnun, AlMudawwana al-kubra li-imam Malik, vol. 1 (Cairo, 1905; reprint, Beirut, n.d.), KJ, p. 174,
declares there is no reciting [of the Quran] over the bier. Abu Hanifas position, in agreement with Maliks, is reported by al-Shaybani, Muwatta al-Imam Malik, 2d ed. (Beirut,
1979), KJ, p. 111, no. 311. Ibn Abi Zayd, K. al-Nawadir wa al-ziyadat ala al-Mudawwana,
vol. 1 (Beirut, 1999), KJ, pp. 542 and 591, with Ibn Habib quoting from Malik: There is no
reading over the bier in the practice of our town (mimma yumalu bihi bi-baladina). Ibn
Abd al-Barr, Al-Kaf i f i qh ahl al-madinah al-Maliki (Beirut, 1987), KJ, p. 84, declares
there is no reading of the funerary prayer according to Malik, yet reports that some of the
magnates of Medina (kubara ahl al-Madina) did recite the introduction to the Quran
after the rst takbira. Al-Bukhari, Sahih, ed. and trans. M. Khan, vol. 2 (Riyadh, 1997), KJ,
chap. 56 (sunnat al-salat ala l-janaiz), claries that the funerary prayer has neither bowing
(ruku) nor prostration (sujud ), and there is no utterance during it (wa-la yutakallamu
f iha); cf. chap. 65 (qiraat fatihat al-kitab ala l-janaza). Ibn Rushd, Bidayat al-Mujtahid,
3:3233, mentions two reasons for disagreement amongst scholars about Quranic recitations at the funeral prayer: rst, the contradiction between the practice (amal ) (i.e., at Medina) and the tradition (athar) (relating to the deeds of Muhammad and his Companions);
second, the question about whether or not funerary prayer belongs to the category of
prayer, in which case recitation is acceptable. For presentations on the divergence among
jurisprudents, see al-Tusi, Al-Khilaf (Qum, 198797), 1:723; and Ibn Hazm, Al-Muhalla,
vol. 5 (Cairo, 192834), KJ, pp. 12931, no. 574. The entire discussion concerns the funerary prayer, not intercessions delivered on Fridays, which were rather more controversial.
135
(d. 855 c.e.) rst opposed such recitations as an innovation (bida), but
then, after hearing that a reliable authority from Syria, Mubashshir alHalabi (d. 816), read the Quran at funerals, he reversed his stance and
recommended the new practice.29
This legal discourse on the permissibility of reciting the Quran during
the prayer over the bier provides, alongside the tombstone record, a fascinating example of a fundamental transformation that took place in the
liturgy of Islam in the eighth century. We observe an innovative, popular
use of the Quran gradually taking hold over the course of the eighth century. Eventually, in the various ceremonies honoring the dead, held on such
occasions as the rst night or the fortieth day after the burial, and generally on Fridays, recitations of Surat Ya Sin (Quran 36), of Surat alBaqara (Quran 2), and of Surat al-Mulk (Quran 67) would become
popular. In some circles, the beginning of al-Baqara would be read at the
front of the grave, the end at the back.30
The origin of such liturgical uses of the Quran in prayers for the dead
lies in the eighth century, as the tombstone record suggests and the legal
discourse corroborates. The absence of Quranic references in the seventhcentury tombstones does not prove that Muslims of that age did not recite
from the Quran in front of the grave. (We must be careful not to confuse
the absence of evidence with the evidence of absence.) Yet clearly the
seventh-century Muslims commissioning the inscriptions did not nd it
important to encourage the recitation of particular Quranic passages. By
contrast, eighth-century inscriptions served to link a personage in the grave
to specic verses from the Quran, through the mediation of a reader.
To a certain extent, prayers for the dead were controlledor at least
directedby the inscriptions. Seventh-century Muslims did not conceive
of exercising such control over the prayers delivered on behalf of the
dead. In that age, it mattered little to the dying whether the living sought
forgiveness in one fashion or another, so long as it was done with good
intention. The ceremony of reading at the grave might well have been
29
Al-Shai, Mawsuat al-imam al-Shai: al-kitab al-umm, vol. 2 (Beirut, 1996), KJ,
p. 381; and al-Tirmidhi, Al-Jami al-kabir, vol. 2 (Beirut, 1996), KJ, chap. 39, no. 1026. The
latter reports that several of the ahl al-ilm, including al-Thawri and some of the Kufans, were
opposed to this reading. Al-Ghazali, Ihya ulum al-din (Beirut: Dar al-Kutub al-Ilmiyya,
n.d.), bk. 40, chap. 6, reports a story that Ibn Hanbal considered recitation of the Quran as an
innovation, yet concluded there is no harm in the practice. See also Ishaq al-Nisaburi, KJ,
Masail al-Imam Ahmad ibn Hanbal (Beirut, 1974), no. 946, cf. no. 931; and Abu Bakr alKhallal, Al-Amr bi l-maruf wa l-nahy an al-munkar (Cairo, 1975), pp. 18791.
30
Al-Khallal, p. 188; Ibn al-Jawzi, K. al-Qubur, in K. al-Mawduat min al-ahadith almarfuat (Riyadh, 1997), chap. 11, no. 1784; al-Suyuti, Sharh al-sudur bi sharh hal al-mawta
wa al-qubur (Damascus and Beirut, 1989), pp. 152 and 197. See also Edward William Lane,
An Account of the Manners and Customs of the Modern Egyptians (London, 1846), 2:294,
301, 307, 308, and passim; and Shaun Marmon, Eunuchs and Sacred Boundaries in Islamic
Society (New York, 1995), pp. 2324.
136
complex in social and anthropological terms, and deeply meaningful. But

it was simple in liturgical terms. By contrast, the late eighth-century inscriptions reveal a religion with a deliberate sacramental character, as by
then Muslims had grown intent on persuading other Muslims to read certain incantations for their benet in the afterlife.
Having discussed the social and religious function of Quranic readings
in the late eighth century, let us turn our attention now to extra-Quranic
formulae. Rabia ibn Maslamas tombstone confesses that the Garden and
the Fire are real.31 Such a pronouncement, underscoring Rabias faith in
Gods justice and in the resurrection, is not only appropriate for a tombstone, it is also to be expected.
The extra-Quranic formula on Rabias tombstone that I nd interesting is the prayer for God to ll his grave with light and to insinuate his
hujja to him.32 Above I translated hujja as proof, but the term means the
evidence or testimony by which a person defends him or herself when
called into question. In this context, hujja refers in particular to the proof
of personal faith delivered to the angels Munkar and Nakir during the inquisition in the grave.
The appeal for God to ll the grave with light is predicated upon a belief system upholding not merely the continuity of human existence after
death but more specically the existence of an afterlife in the grave. God
had the power to make the grave more expansive or narrower, lighter or
darker, in a eld where the normal laws of physics were altogether suspended. His action, whether merciful or punishing, would depend entirely
on the religion of the subject and on his or her record of good and bad
deeds, according to the brand of orthodoxy that upheld the dogma of personal responsibility.33
31
Compare Wasiyat Abi Hanifa, art. 20, p. 129, and Fiqh Akbar II, art. 21, p. 195, both in
Arent J. Wensinck, The Muslim Creed (Cambridge, 1932).
32
A slight variant of this formula was also used in Jordan. See Frdric Imbert, La
ncropole de Qastal al-Balqa en Jordanie, Archologie islamique 3 (1992): 1759, including a number of tombstones dated to the eighth and ninth centuries. Ibn Abi Shayba recorded
that Mujahid used to say: In the name of God and in his path, O God make his grave spacious, and bring to him light in it (nawwir lahu f ihi ), and make him adhere to his Prophet.
Ibn Shahin related a hadith that the Prophet Muhammad used to say, with respect to the punishment in the grave by the angels: Learn your proof, for you will be asked (taallamu hujjatakum fa-innakum masulun). Both quotes are in al-Suyuti, pp. 153 and 195. The specic
formula laqqinhu hujjatahu occurs rarely. It is attested in the funerary prayer recorded in
Shiite lawbooks and directly associated with the punishment in the grave by Munkar and
Nakir: al-Tusi, Tahdhib al-ahkam fi sharh al-muqnia, 2d ed. (Al-Najaf, 1959), 3:330; Ibn Babawayh al-Qummi, Man la yahduruhu al-faqih, 2d ed. (Beirut, 1992), vol. 1, chap. 25, p. 205,
no. 47/500, and p. 206, no. 48/501; and in the Malikite work of Ibn Abi Zayd, KJ, p. 643. The
formula on bringing light to the grave is commonplace and probably has its roots in the Roman and Christian prayers for lux perpetua in the afterlife.
33
On the afterlife in the grave, see Ragnar Eklund, Life between Death and Resurrection
according to Islam (Uppsala, 1941); and Leor Halevi, Muhammads Grave: Death, Ritual
and Society in the Early Islamic World (Ph.D. thesis, Harvard University, 2002), chap. 6.
One Line Long
137
Islam displays a new face on the tombstone of 795 c.e. The merciful
action of God will descend upon Rabia ibn Maslama, it is hoped, on the
basis of invocations delivered by the visitors to his grave on his behalf,
asking for Gods forgiveness (maghra) for his sins. This social fact I
must underscore. Such a tombstone inscription does not simply commemorate the dead, indicating the site where lies the deceased and
thereby allowing the mournful to gather, deprived of the beloved and
removed from God. Rather, it actives a web of complex, imagined connections between God, a particular dead person, and the visitors uttering
Quranic prayers for his or her benet not on the Day of Judgment, but,
with greater immediacy, in the period between death and the resurrection. It was in this period that Muslims would be forced to abide in the
darkness of the grave, which Rabia ibn Maslama feared, and to suffer
grave tortures.34
Only in rare instances do tombstones display such poor taste as to mention the punishment in the grave (adhab al-qabr) explicitly. This was
the punishment, of a purgatorial or a retributive nature, that many a Muslim was obliged to endure in the afterlife between death and the resurrection. It is referred to rst in a tombstone from the year 180 a.h. (796 c.e.)
and occasionally thereafter.35 Muslims began to develop an original eschatology related to this form of punishment in the rst quarter of the eighth
century.36 It is over the course of the eighth century that this belief system gained doctrinal status and a certain degree of institutionalization.37
References in the late eighth-century epitaphs to this belief in the unpleasant nature of the sojourn in the grave, when coupled with prayers
seeking Gods merciful deliverance, suggest a belief in the magical efcacy of the act of reading tombstone inscriptions. The dead, according
to oral traditions recorded in Ibn Abi al-Dunyas Book of the Graves,
34
Note the qualitative change between Abbasas and Rabias maghra; the latter, but not
the former, seems intimately tied to a belief in the continuation of life in the grave. An interesting variant on the plea for forgiveness ends with the qualication hayyan wa-mayyitan,
(while living and while dead). Nevo et al., p. 8, consistently transliterates the latter term as
mayyt-an and mistranslates the phrase, as long as he lives and until his death. Such an
odd translation is justied, in his view, because the concept of forgiveness after death makes
sense only if we assume that Muslims from the Negev had developed a belief in the Resurrection or in the Last Judgment by the Umayyad period! In fact, the last part of the formula
probably refers to the Muslim belief in the need for forgiveness during al-barzakh, the period between death and the resurrection corresponding to the punishment in the grave.
35
See CMASF (n. 2 above), vol. 1, no. 4; and Madeleine Schneider, Stles funraires
musulmanes des les Dahlak (mer Rouge), 2 vols. (Cairo, 1983), 1:59, 60, 69.
36
For one of the earliest references to the punishment of the grave, see Jarir ibn Atiyyah,
Sharh Diwan al-Jarir (Cairo, 1934), p. 279, line 7.
37
See, e.g., the invocation (dua) for the deceased recommended by al-Shai, KJ, no.
3134. It includes a prayer asking God to shelter [his slave] from the torment of the grave.
It mentions also the darkness of the grave and its constricting pressure.
138
needed the invocation of the living (dua al-ahya), which in the darkness of the tomb would be transformed into something resembling light.38
The tombstones from the 790s mark the endpoint of this inquiry into
the reorientation of early Islam. We began with the monuments from the
650s, which commemorated the deaths of Muslims, but without displaying distinctively Islamic signs. Then we turned to the fascinating tombstones from 691 and 721 c.e., observing the emergence of a new sense of
communal identity and remarking upon the rst uses of the Quran to
Islamicize death. Finally, we reached Rabia ibn Maslamas formulaic
tombstone, with its standard references to popular Quranic verses and to
Muslim beliefs in the afterlife.
Yet we must not think of Rabia ibn Maslamas tombstone as the nal
destination of this process of Islamization, for two reasons. First, tombstone inscriptions did not remain xed in this form down the ages. Despite
the high incidence of recurrence of certain Quranic quotations, tombstone style continued to change within a particular locality and to vary
from region to region.39 Second, we cannot take for granted the Islamic
character of Rabia ibn Maslamas tombstone since epitaphs in general
were not universally admired as repositories of an Islamic form.
Critics of this institution will indeed force us to consider a different Islamic path. So far, by analyzing changes in the tombstone record, I have
presented a picture of Islam in motion. But how did the gatekeepers of
the Muslim tradition, the pietists who were generally opposed to religious
change, respond to the rise of an Islamic tombstone culture?
ii. traditionist opposition to tombstones
Not all Muslims who lived in the century between the deaths of Abbasa
and Rabia would have approved of the process of Islamization that
tombstones illustrated. Many traditionists would have in fact disavowed
the historical view that it was an Islamic process, for, as we shall see,
38
Ibn Abi al-Dunya, Kitab al-Mawt (The Book of Death) & Kitab al-Qubur (The Book of
Graves), reconstructed with an introduction by Leah Kinberg (Haifa, 1983), nos. 811, pp.
7374. This reference to the invocation of the living does not directly connect the ritual to
tombstones, as I have. Compare William Graham, Islam in the Mirror of Ritual, in Islams
Understanding of Itself, ed. R. Hovannisian and S. Vryonis (Malibu, Calif., 1983), pp. 5371,
describes orthoprax Islam as a ritualistic religion with a reformational cast that entails
a rejection of sacramental practices. This characterization does not apply to certain popular
Islamic rituals that were imbued with magical efcacy, such as intercessions for the dead.
These popular rituals developed despite the anti-sacramental emphasis of orthodox Islam.
39
At the end of the Kitab al-Qubur, nos. 113 ff., pp. 1013, Ibn Abi al-Dunya includes a
number of unusual specimens. Occasionally epitaphs include poems instead of Quranic quotations. For a sense of regional variety in later tombstones, see Louis Bazin, Persistances
prislamiques et innovations dans les stles funraires ottomanes, in Les Ottomans et la
mort, ed. Gilles Veinstein (Leiden, 1996), pp. 1938; and I. D. Mortensen, Women after
Death: Aspects of a Study on Iranian Nomadic Cemeteries, in Women in Islamic Societies:
Social Attitudes and Historical Perspectives, ed. Bo Utas (London, 1983), pp. 26 47.
139
they dismissed all tombstone inscriptions as an anti-Islamic innovation

and related oral traditions in the name of the Prophet to support their case.
Since countless Muslims chose to commemorate their deaths with
Quranic inscriptions, either disregarding or ignoring traditionist protestations, historians may be tempted to dismiss oral traditions against
tombstones as ineffective if not irrelevant. Yet the examination of these
oral traditions is worthwhile because it offers one of the few instances
where the oral tradition can be linked in a straightforward manner to a
record of material culture. In this way, we will observe how the traditionist mentality developednot so much in order to preserve the collective
memory of Medinas golden age but in reaction to popular practices that
emerged outside of Medina after Muhammads death.
Oral traditions specifying that the Prophet had prohibited plastering
and building over the grave sometimes include a trailer indicating that he
had also prohibited writing upon the grave (naha . . . an yuktaba alayhi).
The person associated with this addendum (ziyada) was Sulayman ibn
Musa (d. 119 a.h./737 c.e.), an obscure collector of oral traditions from
Syria. Reportedly, he delivered the enhanced tradition on the authority of
Jabir (ibn Abdallah, d. 78 a.h./697 c.e.), one of the Prophets Helpers
from Medina, and transmitted it to Ibn Jurayj (Abd al-Malik ibn Abd alAziz, d. 150 a.h./767 c.e.). A famous traditionist from Mecca, Ibn Jurayj
in turn transmitted the tradition to Hafs (ibn Ghiyath ibn Talq, d. 194 a.h./
809 c.e.) of Kufa and, in a slightly different version, to Abu Muawiya
(Muhammad ibn Khazim, d. 195 a.h./810 c.e.), also of Kufa. In variant
traditions that integrate smoothly the prohibition against inscribing tombstones into a general pronouncement against building on graves, Sulayman
ibn Musa is not invoked. He is then replaced in the chain of transmission
by Abu al-Zubayr (Muhammad ibn Muslim, d. 126 a.h./744 c.e.) of
Mecca.40
40
Al-Nasai, Sunan, vol. 4 (Beirut, 1995), KJ, no. 2027; Abu Daud, Sunan (Beirut, 1998),
KJ, nos. 322526; Ibn Maja, Sahih Sunan, 3d ed., vol. 1 (Riyadh, 1988), KJ, chap. 43, no. 1269
(on Jabirs authority); al-Bayhaqi, Al-Sunan al-kubra (Beirut, 1994), vol. 4, chap. 71, p. 6;
al-Tirmidhi (n. 29 above), KJ, no. 1052, replaces Sulayman b. Musa with Abu al-Zubayr; alHakim al-Naysaburi, Al-Mustadrak ala l-sahihayn, vol. 1 (Beirut, 1990), KJ, p. 525, nos.
1369 and 1370, does the same yet records different authorities preserving the tradition from
Ibn Jurayj. Typically, the subject under consideration is subsumed under oral traditions prohibiting the plastering of graves (tajsis al-qubur). However, Muslim ibn al-Hajjaj, Sahih
(Beirut, 1995), KJ, chap. 32 (al-nahy an tajsis al-qabr wa l-bina alayhi ), and Ibn Hanbal,
Al-Musnad, 2d ed. (Beirut, 1994), nos. 14571 and 2261718, among others, mention plastering
while passing over in silence the question of writing. Atypically, Ibn Abi Shayba, Al-Kitab
al-musannaf f i al-ahadith wa al-athar, vol. 3 (Bombay, 1979), KJ, pp. 33435, includes a
heading on the writing on tombstones for identication purposes ( l-qabr yuktabu wayuallamu alayhi ). Here Ibn Abi Shayba includes a tradition that Ibn Jurayj ascribes to Abu
al-Zubayr, mentioning only the prohibition against building on graves, followed by Sulayman ibn Musas addendum concerning tombstone inscriptions.
140
Sulayman ibn Musas tradition conveys the image of a single, thin

thread traveling far and wide over the course of the eighth century, from
Damascus to Mecca and onward to Kufa. The smoother version ascribed
to Abu al-Zubayr leaves aside the Syrian informant whose link to the
Medinan Jabir ibn Abdallah, the Prophets companion, was questionable.
It emphasizes instead circulation within the Hijaz, in Arabia, culminating
in transmission to Mesopotamia. At some point, Ibn Jurayj, the common
link in both traditions, may have substituted Abu al-Zubayr for Sulayman ibn Musa in order to make the chain of transmission appear more reliable. However, such doctoring of Sulayman ibn Musas tradition may
have taken place after transmission by Ibn Jurayj.41
One of the principal aims of modern criticism of Muslim tradition has
been to identify the individuals responsible for placing in circulation traditions of dubious authenticity. In this academic exercise, Sulayman ibn
Musa would likely be crowned as the master forger and Ibn Jurayj as the
complicit transmitter of a questionable tradition. In seeking to account
for the origins of traditionist opposition to tombstone inscriptions, modern
critics would also consider the possibility of Jewish or Christian inuence.
In the case under consideration one may point to a Jewish precedent
but not, so far as I have been able to determine, to a Christian one. Both
Genesis Rabbah and the Jerusalem Talmud report that Rabban Simeon
ben Gamaliel declared, One does not erect ornate tombstones (nefashot)
for the righteous, as their oral teachings ought to comprise their memorial
(divreyhen hen zikhronan). It turns out that wealthy Jews began decorating tombstones with inscriptions only under Greek and Roman inuence,
and this fashion of commemorating the dead was not much appreciated
by austere-minded rabbis. (The Babylonian Talmud even admonishesI
should have warned the reader earlierthat reading inscriptions on graves
affects ones learning adversely, as does staring at the countenance of the
dead.)42 Did Sulayman ibn Musa, perhaps a Jewish convert to Islam
41
Al-Mizzi, Tuhfat al-ashraf bi marifat al-atraf (Beirut, 1999), 2:23334; 12:344 45,
comments that Sulayman could not have heard a tradition from Jabir; the hadith therefore
counts as mursal (incompletely transmitted). See also Ibn Hajar al-Asqalani, Tahdhib altahdhib (Beirut, 1996), 2:11112. With Abu al-Zubayr in the chain of transmission, the hadith would count as musnad (a tradition with uninterrupted ascription going all the way back
to the prophet Muhammad). In versions that substitute Abu al-Zubayr instead of Sulayman,
we see a process of tadlis, dened by Juynboll as tampering with isnads to make them
appear more reliable. Ibn Jurayj, the common link in this set of traditions, was infamous
for resorting to this practice. See G. H. A. Juynboll, Muslim Tradition: Studies in Chronology, Provenance, and Authorship of Early Hadith (Cambridge, 1983), p. 180.
42
Talmud Yerushalmi (TJ), Shekalim, 2: 6/4, 47a. Both the Encyclopaedia Judaica, s.v.
Tombs and Tombstones and Epitaphs, and The Jewish Encyclopedia, s.v. Tombstones,
refer to TJ, Shek. 2: 7, 47a, but I found quote in question in Synopse zum Talmud Yerushalmi,
ed. Peter Schfer and Hans-Jrgen Becker (Tbingen, 2001) at the location specied above.
141
originally named Solomon ben Moses, garb in Muslim form a rabbinical

dictum?
This suggestionthat Sulayman ibn Musa of Damascus ascribed an
opinion that he derived from the Jews to the prophet Muhammadmay
seem offensive to anyone who would like to believe the oral tradition
was revealed in Mecca and Medina before 632 c.e. and by and large preserved integrally and authoritatively for two centuries until nally it
reached canonical, written form.43 But there are also reasons for secular
historians to feel highly dissatised with a conclusion of this order. Why
should it matter to us that a derivative opinion against writing on tombstones originated with a certain Sulayman ibn Musa?
To trace atomistically a specic opinion against tombstone inscriptions
to a particular individual would betray a less than adequate understanding of how ideas developed and why they mattered. The simple truth is
that new ideas arose and old ones resurfaced for good reasons. An inexorable process, having relatively little to do with the individual genius of
gures such as Sulayman ibn Musa, was unfolding.
The rabbinical bias against tombstone inscriptions may be regarded at
most as a proximate cause of the early Muslim bias. It is possible that
Jewish converts to Islam, or Muslims engaging in dialogue with Jews, colored Muslim discourse. Yet, as we shall see, the ultimate causes for the
Muslim bias were different: current use of tombstones, knowledge that
tombstone inscriptions were not an indigenous Medinan-Muslim practice, and general opposition to pompous structures commemorating the
dead. These three causes sufciently account for the Muslim prejudice,
which might have arisen in some form or other without Jewish inuence.
Opposition to tombstone inscriptions was in fact geographically widespread in the eighth century. Umar (Ibn Abd al-Aziz, d. 720 c.e.), the
caliph who has earned a reputation for advancing posthumously the pious
practice of the good old days,44 not only forbade that graves be built
upon with brick (bi-ajurr),45 he also abhorred the grave be given a distinguishing mark (kariha an yuallama l-qabr).46 One of the seven
lawyers of Medina, al-Qasim (ibn Muhammad ibn Abi Bakr, d. 106
a.h./724 c.e.), a man who might also have campaigned against writing
See also Bereschit Rabba mit kritischem Apparat und Kommentar, ed. Julius Theodor (Berlin, 191236), 82:10; Babylonian Talmud (Wilna, 188086), Horoyoth 13b. Fortunately, say
the rabbis, drinking water left over from kneading bread restores ones learning.
43
Muslims with critical expertise in hadith do recognize the problem of forgeries. See Muhammad Zubayr Siddiqi, Hadith Literature: Its Origin, Development and Features, rev. ed.
(Cambridge, 1993), pp. 3136.
44
Joseph Schacht, The Origins of Muhammadan Jurisprudence (Oxford, 1950), p. 192.
45
Ibn Qudama, KJ, Al-Mughni (Cairo, 1987), 3:439.
46
Ibn Abi Shayba, KJ, pp. 33435.
142
on another front,47 decreed on his will: O, my son, do not write on my

grave (la taktub ala qabri).48 A Khurasani who settled in Kufa and
traveled to collect oral traditions, Zayd ibn al-Hubab (ibn al-Rayyan, d.
203 a.h./818 c.e.), related on the authority of a Basran who settled in
Mecca, Mubarak (ibn Hassan al-Sulami), that al-Hasan (d. 110 a.h./728
c.e.) of Basra found repugnant the placing of a tablet over the grave
(lawh ala l-qabr). They used to detest that a man mark his grave
(yakrahuna an yuallima l-rajulu qabrahu), another hadith relates, on the
authority of Salim ibn Hayyan of Basra, Hammad (ibn Abi Sulayman, d.
120 a.h./738 c.e.) of Kufa, and Ibrahim (al-Nakhai, d. 96 a.h./715 c.e.)
of Kufa.49 Yahya ibn al-Ala (al-Bajali), an informant who acquired a
reputation for mendaciousness, reported that Rashid ibn Sad (al-Himsi,
d. 726 c.e.) told his fellow traditionist from Homs/Emesa in central Syria,
al-Ahwas ibn Hakim (ibn Umayr al-Hamdani al-Himsi), that Muhammad
had prohibited plastering graves (taqsis al-qubur), crowning them with
an erect structure (taklil ), and writing on them.50
Considered broadly, the traditions above suggest that opposition to
tombstone markers took place wherever traditionists lived. This reaction
occurred sometime in the seventh or eighth centuries. Certainly it originated at the latest by the last quarter of the eighth century, since the oral
traditions in question were recorded by Abd al-Razzaq (744827 c.e.)
and Ibn Abi Shayba (775849 c.e.), among others. Maybe the traditions
originated earlier, however. It is impossible to determine whether or not
particular authorities from the turn of the eighth centurygures such as
al-Qasim of Medina, Abu al-Zubayr of Mecca, Ibrahim of Kufa, al-Hasan
of Basra, Rashid from Homs, and Sulayman of Damascusactually delivered the opinions ascribed to them. Nevertheless, the existence of Islamic tombstones from 691, 694, 721, and 729 c.e. makes it reasonable
to hold that certain religious authorities from this time period opposed the
developing practice.51 Given the difculties of dating the origins of any
controversy played out in the oral traditions, this is remarkable evidence.
47
Michael Cook, The Opponents of the Writing of Tradition in Early Islam, Arabica 44,
no. 4 (1997): iiii and 437530; pp. 462, 472, and 503 in particular.
48
Ibn Abi Shayba, KJ, pp. 33435.
49
Both in ibid., pp. 33435.
50
Abd al-Razzaq (n. 25 above), KJ, no. 6497. Note the alternative wording of this traditionnamely, the use of taqsis instead of the standard tajsis and of taklil instead of the standard bina.
51
By the same logic, the Muslim tombstones from 650 and 652 c.e. make it not reasonable to believe that opposition to tombstone inscriptions might have begun concomitantly,
roughly a quarter of a century after the death of the prophet Muhammad. Hadith criticism
would not lend support to this belief, however, given that the isnad of Sulayman ibn Musas
tradition is mursal; no link exists, in other words, between the tradition and one of the
Prophets Companions.
143
The reasons for eighth-century opposition by traditionists to tombstone

inscriptions are not entirely clear. An eleventh-century Malikite jurisprudent, Muhammad ibn Rushd, would claim that Malik (ca. 71295 c.e.),
the master of Medina, had abhorred the inscribed agstone because the
practice was one of the innovations (bida) by which powerful magnates
(ahl al-tawl ) seek glory and fame. In attributing this opinion to Malik, he
seems to have erred. Malik himself abhorred the practice of piling up
stones at the grave, surely in reference to the mound or pillbox cairns so
commonly spotted throughout Arabia. But he seems not to have preached
against tombstone inscriptions specically. And why should he have? If
tombstone inscriptions were not all that popular in eighth-century Medina,
as suggested by the tombstone record,52 then Malik may have had little
reason to be concerned with the innovation. Rather, it was an Egyptian
Malikite, Ibn al-Qasim (d. 806 c.e.), who abhorredapparently on his own
authoritythat an inscribed agstone (balata) be placed on the grave.
He declared that there was no problem with the use of stone or wood in
memorials for the dead, so long as nothing was written on these materials
by which a man might identify the graves inhabitant (wali).53 That this
52
A search for Medinan tombstones from the rst three centuries of Islam yielded none.
In this research, I consulted Janine Sourdel-Thomine, ed., Index gographique du rpertoire
chronologique d pigraphie arabes (Cairo, 1975); and the CD-ROM of the Fondation Max
Van Berchem: Thesaurus d pigraphie islamique, directed by Ludvik Kalus, 3e livraison
(Paris and Geneva, 2001). However, it is worth noting that a number of early Islamic tombstones have been discovered in the Arabian peninsula. In addition to the work of al-Faqih on
the epitaphs in the environs of Asham, see Madeleine Schneider, Stles funraires musulmanes du Ymen, Revue des tudes Islamiques 7, no. 1 (1979): 71100, where the earliest
tombstone (pp. 7273) has been dated to the century 250350 by the Hijri calendar. Recently
a team of Saudi archaeologists discovered a tombstone in Rabadhah, a way station on the
pilgrimage road from Kufa to Mecca, with archaic Kuc script, which they dated rather tentatively (but not altogether convincingly) to the turn of the rst century of Islam; see Sad
al-Rashid, Al-Rabadhah: A Portrait of Early Islamic Civilisation in Saudi Arabia (Riyadh,
1986), pp. 8788. A. Grohmann, Expdition Philby-Rychmans-Lippens en Arabie, pt. 2, Textes
pigraphiques, vol. 1, Arabic Inscriptions (Louvain, 1962), pp. 4 41, includes a number of
third-century a.h. epitaphs from the cemetery of Ikrima near al-Taif. See also Hasan alBasha, Ahammiyyat shawahid al-qubur ka-masdar li-tarikh al-jazira al-arabiyya fi al-asr
al-islami, in Tarikh al-Jazirah al-Arabiyah (Riyadh, 1979), vol. 1, pt. 1, pp. 81126; and
Sad al-Rashid, Kitabat islamiyya min Makka al-mukarrama (Riyadh, 1995), and Kitabat
islamiyya ghayr manshura min Ruwawat al-Madina al-munawwara (Riyadh, 1993); Moshalleh Moraekhi, A Critical and Analytical Study of Some Early Islamic Inscriptions from
Medina in the Hijaz, Saudi Arabia (Ph.D. diss., University of Manchester, 1995); Muhammad al-Salook, Analytical and Paleographic Study of Some Early Kuc Inscriptions from
Saoudi Arabia (masters thesis, University of Durham, 1998). Finally, it is worth noting the
Meccan epitaphs from the Coll. G. Wiet, an unpublished collection available for consultation through the Thesaurus d pigraphie islamique. It includes a Meccan tombstone dated to
175 a.h./791 c.e.; see che 14931.
53
Muhammad Ibn Rushd (10561126), Al-Bayan wa al-tahsil, vol. 2 (Beirut, 1984), KJ,
pp. 22021, where the opinion is attributed to Malik, though on the basis of a quotation from
144
opinion against inscribed agstones should have occurred to a Malikite

jurisprudent from Egypt is interesting, for it was in the lands of the eastern Mediterranean, and particularly in Egypt, that Muslims turned in large
numbers to commemorate their dead with tombstone inscriptions.54 As
an Egyptian transmitter of Medinan doctrine, Ibn al-Qasim was of course
highly aware of the discrepancies between this popular Egyptian practice
and the venerable Medinan tradition.
The controversy over plastering and building upon graves, which was
associated with writing on tombstones, centered on an egalitarian principle: the leveling of the graves (taswiyat al-qubur). In the words of a
North African jurist who ourished in the early eighth century, Bakr Ibn
Sawada, the graves are to be leveled with the ground (kanat al-qubur
li-tusawwa bi l-ard).55 Jurisprudents, who resented tombstone inscriptions as an innovation deviating from the custom of the Prophets Companions, argued eventually that the dead had no need (haja) for this form of
worldly embellishment (zina). Squandering capital without benet seemed
abhorrent (makruh) in their eyes.56
This austere-minded reaction to an innovative practice that had developed outside of Medina was not the only reason for the original opposi-
Ibn al-Qasim, who does not himself cite the master. Malik, KJ, both in Yahyas and in Suwayds recensions, mentions neither the topic of plastering nor inscribing. According to Sahnun (n. 28 above), KJ, p. 189, Malik abhorred plastering and building upon graves. Ibn Abi
Zayd (n. 28 above), KJ, pp. 65253, cites the Utbiyya manuscript representing Maliks abhorrence for building or stacking up stones at the grave, and Ibn al-Qasims independent abhorrence for writing on tombstones.
54
Thus I cited above the major collections of Egyptian epitaphs, as well as Imberts article (n. 32 above) on the tombstones of Qastal al-Balqa, in Jordan. The RCEA (n. 3 above)
includes a number of third-century tombstones from non-Egyptian sites, dated by script to
ca. 250 a.h. Busra in southern Syria, Homs in Central Syria, and other unspecied sites in
Syria are best represented: vol. 1, no. 114 (ca. 200 a.h.); vol. 2, nos. 542 46 (ca. 250 a.h.),
675 (264 a.h./87778 c.e.). For early Islamic inscriptions from Syria, see also Solange Ory,
Cimetires et inscriptions du Hawran et du abal al-Duruz (1989), nos. 1, 8, 9. S. Ory has
also studied epitaphs from Busra. In a personal communication, he has informed me of his
plans to edit these on-line, in a collection that will include two dozen epitaphs from the rst
three centuries of Islam. Given traditionist opposition to tombstones in Kufa and Basra, the
reader may wonder about the existence of Mesopotamian tombstones. However, this region
is not well represented in the epigraphic record. The RCEA cites only one early Islamic tombstone from Kufa or Basra, an epitaph dated to ca. 250 a.h. (vol. 2, no. 547). The reason for
the scarcity of tombstones here was the lack of stone in the alluvial plain. Muslims in Mesopotamia would have needed to import or recycle stone. Alternatively they could have engraved their memorials for the dead in clay. One way or the other such monuments would
have not survived the ravages of time as easily as in Egypt, where the aridity of the land
worked to preserve an extensive record.
55
Sahnun, KJ, p. 189, in the section on the plastering of graves, which does not specify
anything about writing.
56
See, e.g., the sharp formulation of the Hanafi jurist al-Kasani, K. al-Badai al-sanai
fi tartib al-sharai (Pakistan, 1989), 1:320.
145
tion. Another, equally important reason was suggested by Ibn al-Qasim.

The eighth-century discourse seems primarily interested in preventing the
earmarking of graves, hence the use of the verb to be given a recognizable mark (yuallamu). Thus, Abu Hanifa (d. 767 c.e.), a key jurisprudent from Kufa, found reprehensible the distinguishing of a grave by a
sign (alama).57 In other words, the other key problem with inscribed
tombstones was their capacity to identify the graves of particular persons.
Why pietists found this development a matter of concern it is not difcult
to guess, given their opposition to intercession on behalf of the dead, to
women visiting tombs, and to the cult of the saints. Ibn al-Hajj (d. 1336
c.e.), a Malikite Egyptian scholar, explained this concern well in his tirade against epitaphs. He railed against the practice of people seeking
fame, who commission inscriptions for the graves of saints, ancient scholars, and pious ancestors, in order to spur graveside tourism. In this practice,
born out of religious ignorance and godlessness, he identied a source of
corruption.58
Such persuasive preaching against the idolatrous practice of visiting
tombs did, under the right conditions, culminate in violent action. The
razing of the cemetery of al-Baqi instigated by Wahhabi zealots serves
as a recent, memorable example.59 In early Islamic times, opposition to
tombstones could also become, on occasion, forceful, as indicated by a
story about political action against a tombstone of high symbolic value.
While Marwan ibn al-Hakam (684750 c.e.) was the governor of Medina,
he passed by a rock in the cemetery of al-Baqi, which commemorated
the burial site of Uthman ibn Mazun (d. 624 c.e.), an early convert to
Islam who fought at the Battle of Badr. According to Muslim tradition
the prophet Muhammad himself had commanded the placement of this
rock in honor of the deceased man, so that I will recognize him by it.
This was the rst burial of a Muslim at al-Baqiand from this point onward the Muslims of Medina, following the Prophets example, would bury
their dead at the same cemetery.60 Disregarding this history, Marwan ibn
al-Hakam ordered the removal of the monument: By God, there will not
be in the grave of Uthmam b. Mazun a rock by which he can be recognized (hajar yurafu bihi). The Banu Umayyaa category representing
57
Ibid., p. 320.
Ibn al-Hajj, Al-Madkhal (Cairo, 1929), 3:274.
59
For a poignant description of the cemeterys demolition, see Eldon Rutter, The Holy
Cities of Arabia (London, 1928), 2:25658.
60
Ibn Shabba, Tarikh al-Madinah al-munawwarah/Akhbar al-Madinah al-nabawiyah
(Beirut, 1996), 1:68, no. 315, with Abd al-Aziz ibn Umar relating from Muhammad ibn
Qudama. Ibn Abi Shayba, KJ, pp. 33435, with Abu Bakr al-Hanafi relating from Kathir
ibn Zayd (d. ca. 158 a.h./775 c.e.), on the authority of al-Muttalib ibn Abdallah b. Hantab.
On Uthman ibn Mazun, see Ibn Sad, Al-Tabaqat al-kubra (Biographien Muhammeds), ed.
E. Sachau (Leiden, 1904), vol. 3, pt. 1, pp. 28691.
58
146
the Sufyanid dynasty (66184 c.e.), soon to be displaced from power

complained to the Medinan governor and eventual founder of the Marwanid dynasty about the removal of a memorial set in place by the Prophet.
It is wretched what you have done, they declared. To which the caliphto-be replied, No way, since I lifted it, it cannot be returned.61
The story is complicated, as are many oral traditions, and not necessarily truthful. Possibly it symbolizes the Abbasid perception of a shift in
policy that occurred between Sufyanid and Marwanid times. Alternatively
it may represent Shiite memories of the destruction of commemorative
monuments by Abbasid governors, or veiled popular criticism of the
Abbasids. Governmental demolitions of specic monuments and of unclaimed graves did occur in the late eighth and early ninth centuries.62
What matters for our purposes is the fact that violent opposition to tombstones had not only its supporters, but also its detractors.
This last point brings up questions of some signicance in the history
of religion: How uniform was the opposition of Muslim scholars to
tombstone inscriptions? Were they all impractical idealists who abhorred
popular practices that deviated from the traditional customs of Medina?
When they observed that ordinary Muslims were decorating their tombstones with Quranic quotations and orthodox statements of belief, did
some of them not begin to approve of this practice and to see in it the potential to spread the word of Islam? Or was the tension between popular
practices and orthodox ideals so sharp that Muslim scholars, overwhelmed
by tradition, were unable to perceive religious change in a positive light?
To answer these questions, we will need to consider later developments
in Sunni and Shiite law.
The History of Medina by Ibn Shabba (789875 c.e.) includes a number of storiesignored or deliberately neglected by traditionistsrepresenting tombstone inscriptions without censure. In this account we hear
of the serendipitous discovery of epitaphs commemorating the death of
personages closely associated with the prophet Muhammad. Thus, when
Aqil ibn Abi Talib (d. ca. 675 c.e.) was digging a well in his house, he
found an inscribed stone (hajar manqush) that revealed: The grave of
61
Ibn Shabba, 1:69, no. 318.

Al-Shai, KJ, no. 327576, on how he observed the rulers destroying graves in Mecca,
over which no one claimed ownership. He reports that no jurisprudent censured this act. In the
same chapter, KJ, no. 3289, al-Shai mentions Muhammads curse on the Jews and Christians who adopted the graves of their prophets as mosques, let neither religion remain in the
land of the Arabs. On legal opposition to tomb structures, see Thomas Leisten, Between
Orthodoxy and Exegesis: Some Aspects of Attitudes in the Sharia toward Funerary Architecture, Muqarnas 7 (1990): 1222. Yusuf Raib, Les premiers monuments funraires de
lIslam, Annales Islamologiques 9 (1970): 336, mentions on p. 27 that Shiites accused
Harun al-Rashid of demolishing al-Husayns mashhad.
62
147
Umm Habiba, the daughter of Sakhr b. Harb. This refers to Ramla bint
Abi Sufyan (d. 664 c.e.), a Meccan woman who married Muhammad
upon her return from an excursion to Abyssinia. Another story recounts
how, as a grave digger was digging the grave for Muhammad ibn Zayd
ibn Ali (a son of the founder of Zaydi Shiism), he discovered, at the
depth of eight hand spans, a broken stone, upon which something was
written. On it he managed to decipher the inscription: Umm Salama,
wife of the Prophet, who died in the year 59 a.h. (679 c.e.).63 Apocryphal or not, these stories served to link close members of Muhammads
family to tombstone inscriptions, and in this manner signaled implicit
approval for the practice.
It is said that Fatima (d. 632 c.e.), the daughter of the Prophet and a gure positively associated with the cult of the dead, used to visit the grave
of Medinas martyr, Hamza, which she recognized by a stone (hajar).64
Fatima was the key female gure and the noblest ideal in the tradition of
the Shiites. A tradition in her name condoning the use of a commemorative stone is not altogether surprising, as Shiite traditions diverge sharply
from Sunni ones in accepting the use of tombstone inscriptions.
Zaydi and Ismaili Shiites joined Twelver Shiites in this permissive
attitude toward tombstones. Thus the Ismaili chief judge of the Fatimid
caliphate, al-Numan (d. 974 c.e.), related that Muhammad had placed a
stone (hajar) at the head of Uthman ibn Mazuns grave, approving of its
use as a sign (alam) for the burial of kin.65 The Zaydi-Mutazilite al-Hadi
ila l-Haqq (d. 911 c.e.), founder of the Zaydi imamate in Yemen, considered it unproblematic for a boulder (sakhra) engraved with the deceaseds name to be placed at the head of a grave. Boulders were preferable to tablets (alwah), he claimed, but the latter were acceptable.66 The
Zaydi Yahya ibn Hamza (d. 1348 c.e.) commented on a new trend: the
use of two stones to mark womens graves. This practice was deemed unproblematic by the Zaydi-Mutazilite Ibn al-Murtada (d. 1437 c.e.), if its
purpose was to distinguish the grave. Inscribing a boulder with the name
of the deceased was perfectly acceptable in the Zaydi tradition, so long
as this was done in a simple manner, without ornamentation (zakhrafa).
Key early Zaydi authorities, including al-Qasim ibn Ibrahim al-Rassi
63
Ibn Shabba, 1:17980, nos. 359 and 360. For the identication of Muhammad as a son
of the famous Zayd ibn Ali, see Ibn Hazm, Jamharat ansab al-Arab (Cairo, 1962), p. 58. It
is of course unlikely that Aqil would have discovered Umm Habibas tombstone given that
he died shortly after she did.
64
Ibn Shabba, 1:86, no. 382.
65
Al-Numan, KJ, Daaim al-Islam (Cairo, 1963), 1:238.
66
Al-Hadi ila l-Haqq, Kitab al-ahkam f i al-halal wa al-haram (n.p., 1990), 2:166.
148
(d. 860 c.e.), approved of this practice, which they justied by citing
Muhammads action with Uthman ibn Mazun.67
Sahl ibn Ziyad (al-Razi), a ninth-century jurist associated with the
eleventh Imam of the Twelver Shiites, related a story about the seventh
Imam: When Abu al-Hasan Musa (al-Kazim, d. 799 c.e.) was returning
from Baghdad to Medina, his daughter died in Fayd, a village on the way
between Kufa and Mecca. So he buried her, and ordered one of his nonArab clients (mawali), to plaster her grave and inscribe upon a tablet
(lawh) her name, and place it within her grave ( l-qabr). The purport
of this Shiite tale, the collector of the tradition makes clear, was to lift
the interdiction (hazr) from writing on tombsby way of this sanction
(rukhsa).68
Unlike Shiites, Sunni jurisprudents were unable to strike down the
prohibition against inscriptions by referring to the authoritative example
of an Imam. For Sunnis, divine revelation had ended with the Prophets
death, and, for this reason, they were less capableor rather, they faced
a greater challengethan did Shiites in accommodating religious change.
Bound as they were to the oral tradition holding that Muhammad had
prohibited writing upon the grave, Sunni jurisprudents remained down
the ages virtually unanimous in considering this practice abominable
(makruh).69
In certain cases, however, they did sanction the use of grave markers
by which to recognize an individuals burial site, following the Prophets
example with Uthman ibn Mazun.70 Rarely did they consider inscriptions acceptable. In an exceptional ruling, Ibn al-Hazm (d. 1064 c.e.), the
famous Andalusian scholar of the Zahiri school of law, declared that he
did not abhor the engraving of the deceaseds name on a stone.71 Unfortunately he did not explain the grounds for his reasoning.
67
Ibn al-Murtada, K. al-Bahr al-zakhar al-jami li-madhahib ulama al-amsar, 2d ed.,
vol. 3 (Beirut, 1975), KJ, pp. 13132. On the margins of this work the commentator claries
that the Prophet had not placed a written mark on Ibn Mazuns grave but simply a rock.
Nevertheless, he continues, by analogical reasoning inscriptions appear justied.
68
Al-Tusi, KJ, Al-Istibsar, vol. 1 (Beirut, 1991), KJ, no. 768. See also al-Kulayni, Al-Furu
min al-kaf i, vol. 3 (Teheran, 1957), KJ, p. 202, no. 3; note use of in the grave rather than
on the grave. Ibn al-Mutahhar al-Allama al-Hilli, Mukhtalaf al-Shia (Qum, 1992), 2:315,
shows divergence in Twelver Shiite views regarding building on graves but does not mention tombstone inscriptions.
69
Raib, Les pierres de souvenir (n. 5 above), pp. 32223, includes a short but useful
discussion on Sunni laws regarding epitaphs. It does not mention Shiite views.
70
Al-Hattab, Mawahib al-jalil li-sharh mukhtasar Khalil (Tripoli, Libya, 1969), 2:243.
Compare al-Ramli, Nihayat al-muhtaj ila sharh al-minhaj f i al-qh ala madhhab al-Imam
al-Shai (Cairo, 1967), 3:34, considers it acceptable to engrave the deceaseds name in case
of need.
71
Ibn Hazm, Al-Muhalla (n. 28 above), KJ, p. 133, no. 577.
149
With good humor, the famous traditionist al-Hakim al-Naysaburi (d.

1014 c.e.) conceded, after declaring sound two traditions against tombstone inscriptions, The practice (amal ) is not in accordance with the
traditions, for both in the East and the West the graves of the Imams of
the Muslims are inscribed; this is a practice that the descendants of the
ancestors adopted. 72 From a strict traditionist perspective, the sound
nature of the traditions entailed an unequivocal prohibition against writing. Disarmingly al-Hakim al-Naysaburi acknowledged that the practice
of the descendants had diverged from that of the ancestors and in this
way recognized, if not condoned, a change in the practice of Muslims.
Entertaining doubts about the soundness of Sulayman ibn Musas addendum, Abu Bakr Ibn al-Arabi (d. 1148 c.e.), a commentator on traditions from Seville, formulated a more permissive position than did alHakim al-Naysaburi. His view was that people ought to be treated kindly
or indulged forgivingly for daubing graves with clay or commissioning
tombstone inscriptions, since there is no benet in marking the burial site
except to prevent its falling into oblivion.73
Aware of the opinions of both al-Hakim al-Naysaburi and Ibn alArabi, al-Wansharisi (d. 1508 c.e.) judged writing on tombstones permissible, yet imposed one condition: the memorials should be placed
vertically, rather than horizontally, lest people trample on them. In the
context of a discussion on the legality of decreeing in ones last will the
wish to be buried with a copy of the Quran or of al-Bukharis authoritative collection of traditions, al-Wansharisi returned to our subject. Inscribing a headstone with Gods name, Quranic verses, and the Muslim
profession of faith that there is but one God and Muhammad is his messenger (shahadatan) seemed unproblematic, in his view, for this practice
does not harm ones (dead) neighbors at the cemetery, and in any case, it
had become part of the customary law (urf ) in Muslim cities.74
Al-Wansharisis relatively permissive attitude toward Quranic inscriptions is surprising, given the traditionist rejection of tombstones. Notably,
however, al-Wansharisi did not recommend such inscriptions. He merely
considered them unproblematic. It is striking that it took over seven centuries after the rise of the practice for a Sunni scholar to come up with
such a lukewarm acceptance of Quranic inscriptions.
72
Al-Hakim al-Naysaburi (n. 40 above), KJ, no. 1370.

Ibn al-Arabi, Aridat al-ahwadhi bi sharh Sahih al-Tirmidhi (Beirut, 1997), vol. 4,
comment on chap. 58, hadith 1052, pp. 21617. As precedent for this opinion, Ibn al-Arabi
cites al-Hasan al-Basris concession to daubing graves with clay. See also al-Hattab, p. 247,
and al-Wansharisi, Al-Miyar al-murib wa al-jami al-mughrab an fatawa ulama Ifriqiya
wa al-Andalus wa al-Maghrib (Beirut, 1981), 1:318.
74
Al-Wansharisi, 1:31719, and 9:39496.
73
150
Still, by nding acceptable tombstones of this kind, al-Wansharisi diverged from the opinion of other Sunni jurisprudents. Al-Adhrai (Shihab
al-Din Ahmad ibn Hamdan, d. 1381 c.e.), a Shaite jurisprudent, turned
to analogical reasoning in order to declare such tombstones unlawful
(tahrim). In his view, inscribing the Quran on graves exposed Muslim
scripture to trampling underfoot, not to mention the impurity owing from
corpses.75 Prey to similar fears, the Malikite scholar Ibn al-Hajj (d. 1336
c.e.) conjured up a terrifying image. By perpetrating the innovation of
engraving a tombstone with Quranic quotations, which would include
one of Gods names, one risked debasing Islam. These tombstones might
fall into oblivion, a robber might plunder them, and they might end up
face down in a toilet. There, a Muslim might trample upon them unknowingly, a Jew or a Christian audaciously.76
iii. conclusion
The forceful rejection of inscribing tombstones with Quranic verses reveals the workings of a traditionist mentality. Opposition to this popular
practice was grounded, of course, in the traditionist view that the Muslim
customs of Medina in Muhammads age were perfect and all subsequent
religious developments were in essence irreligious. However, this
opposition did not derive from abstract contemplation of a golden past,
for despite themselves traditionists lived in their own age. Their opinions
emerged in reaction to a practice that arose in the eighth century, long
after Muhammads death, and thus bore no relation to the Prophets own
preoccupations with the matters of his day.
Due to traditionist opposition to religious change, a gap arose in Islamic
history between orthodox ideals and novel practices. In the traditionist
discourse against the innovation of tombstones, one can notice such a
gap and, at the same time, sense the distance that began to open in the
eighth century between the orthodox vision of Islam and the reality of
popular Islam. In certain cases, where tombstones were sanctioned, the
distance, though palpable, was easily bridgeable. In other cases, however, a gulf between the two shores emerged with the separation manifesting itself in an elitist, learned rejection of practices not rooted in
Muslim tradition.
This gap should not be exaggerated, however, as it privileges the
traditionist construct of a basic distinction between universal Muslim
75
Al-Ramli, 3:34. For the identication of this al-Adhrai, see Carl Brockelmann, Geschichte der Arabischen Litteratur, vol. suppl. 1 (Leiden, 1996), p. 680, no. 6, and vol.
suppl. 2, p. 108, no. 16a; Ibn Qadi Shuhba, Tabaqat al-Shaiyya (Hyderabad, 1979), vol. 3,
no. 678, pp. 19094. I thank members of the scholarly discussion network, Middle East Medievalists, for answering my query regarding the identity of al-Adhrai.
76
Ibn al-Hajj (n. 58 above), 3:273.
151
customs and innovative local practices, and it suggests, in this respect,

that a tremendous cultural barrier divided the religion of the traditionists
from the religion of the vulgar. In fact, the barrier between the two sides
was not as high as diehard traditionists imagined it.77
Consider, in this respect, the position of tombstone inscribers. They were
by no means unlearned. They were literate. They referred to the power of
God in dogmatic terms. They understood orthodox Muslim views of salvation and the afterlife. They issued perfectly orthodox confessions of
faith and they quoted the Quran accurately. Culturally speaking, their inscriptions were subject to both upward and downward pressures: a downward pressure because they agreed, despite their learning, to indulge the
Muslims who insisted on memorializing the dead; an upward pressure
because they possessed the knowledge not to deviate in their formularies
from Muslim dogma.
Perhaps for this reason traditionists directed their barbs not so much
against ignorant persons as against innovative practices. Yet it was precisely because they contained a mixture of elements mediating between
high and low that these practices worked effectively as a force of acculturation: they served on the one hand to diffuse Islamic notions to the
population at large and, on the other hand, to dene by opposition the
contrarious, reactionary movement of the traditionists.
Ironically, traditionists rejected not only dangerous innovations but
also the perfectly pious practices of reciting the Quran at the cemetery
and inscribing memorials to the dead with Gods name and the profession of faith. Such practices, though foreign to the Medina of Muhammads age, became an essential mechanism by which Islam, over the
course of the eighth century, spread to and grew rooted in South Arabia,
Mesopotamia, the Mediterranean world, and other regions where a culture of tombstone inscriptions had existed before the rise of Islam. In
77
Arnaldo Momigliano, Popular Religious Beliefs and the Late Roman Historians, in
his Essays in Ancient and Modern Historiography (Middletown, Conn., 1977), pp. 14159,
on 148, 150, and 15556, argues against any clear separation between upper- and lowerclass Christian culture in the late fourth and early fth centuries. Similarly, in The Cult of the
Saints (Chicago, 1981), pp. 1222, Peter Brown effectively criticizes the two tiered model
of religious history, which separates the views of an enlightened minority from the religion
of the vulgar. For Momigliano, this Christian abolition of the internal frontiers between
the learned and the vulgar implies, in part, that cultured persons had accepted unsophisticated beliefs. The early Islamic case seems radically different, in light of tombstone inscriptions, for three reasons. First, we are not dealing here with unsophisticated beliefs, as
epitaphs display a sophisticated belief system. Second, the Islamic discourse divides not the
learned from the vulgar, but sound practices from innovative ones. Third, instead of abolishing frontiers, traditionists created them. Their views cannot be dismissed as a product of
eighteenth-century rationalism. That said, it is important not to adopt uncritically the traditionist scheme of separation, which must be recognized as a construct.
152
fact, it is difcult to imagine how Islam could have become a world religion had Muslims such as Abbasa ibnat Jurayj and Rabia ibn Maslama
not developed novel practices that contradicted traditionist notions of
Islam while transforming pre-Islamic forms. Here lies buried the paradox
of Islamization.
Texas A&M University

The Paradox of Islamization Tombstone in

Uploaded by

Copyright:

Available Formats

The Paradox of Islamization Tombstone in

Uploaded by

Document Information

Original Description:

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

The Paradox of Islamization Tombstone in

Uploaded by

Copyright:

Available Formats

Leor Halevi

Twenty or so years after the death of the prophet Muhammad in 632,

2004 by The University of Chicago. All rights reserved.

This inscription was written in Jumada II of the year 31 (January or February

One Line Short

i. the process of islamization

This inscription of the Muslim daughter of Jurayj, or Little George, a man

we have an unequivocal confession of faith. The use of the imperfect

I have chosen this particular inscription because, though conventional or

acy.)26 In any case, without a reader, it is difcult to imagine much of a

complex in social and anthropological terms, and deeply meaningful. But

One Line Long

they dismissed all tombstone inscriptions as an anti-Islamic innovation

Sulayman ibn Musas tradition conveys the image of a single, thin

originally named Solomon ben Moses, garb in Muslim form a rabbinical

on another front,47 decreed on his will: O, my son, do not write on my

The reasons for eighth-century opposition by traditionists to tombstone

opinion against inscribed agstones should have occurred to a Malikite

tion. Another, equally important reason was suggested by Ibn al-Qasim.

the Sufyanid dynasty (66184 c.e.), soon to be displaced from power

Ibn Shabba, 1:69, no. 318.

With good humor, the famous traditionist al-Hakim al-Naysaburi (d.

Al-Hakim al-Naysaburi (n. 40 above), KJ, no. 1370.

customs and innovative local practices, and it suggests, in this respect,

You might also like