The Content And Context Of Early Arabic Inscriptions*
R. G. Hoyland
© Jerusalem Studies In Arabic & Islam, 1997, Volume 21, No. 3, pp. 77-102.
When Janine Sourdel-Thomime wrote her entry for the Encyclopaedia of Islam on Islamic epigraphy (s.v., "Kitābāt"), she struck an optimistic note, describing it as, "a science full of promise . . . a science indispensable to the scholarly exploitation of a whole category of authentic texts capable of throwing light on the civilisation in the context of which they were written." Yet though catalogues and corpuses of such texts abound, and indeed proliferate, the expected utilization of them has not materialised. There are almost no studies which use, as, opposed to collect, these writings; no monographs which consider what these writings can tell us, who wrote them and why. It is in the hope of provoking debate on these matters that I offer some comments here. In order to restrict the amount of material, I confine myself principally to those inscriptions dated or dateable to the first two centuries of Islam (AH 1-200 / AD 622-815). This is also a period which would benefit most from an injection of new historical data, since it is a period for which Muslim literary texts, or at least physical evidence thereof, are mostly lacking. The great majority of the extant inscriptions from this time are graffiti, which are mostly brief, though may run into several lines, and which are found in their thousands etched on rocks throughout the central Islamic lands from the 30s / 650s onwards. From the 170s / 790s epitaphs begin to be numerous, especially in Egypt, and the smooth surface of a tombstone, in contrast with the rough rock on which are carved most grafflti, allows a greater length of text and a greater range of formulae. Both of these two types are personal statements on behalf of an individual; a third type, the official inscription, is, on the other hand, a public declaration on behalf of the community and will be found on a multiplicity of media: buildings, milestones, coins, textiles, ceramics, seals, weights, measures and so on. Such inscriptions are relatively few and brief in our period, but increase in number and become more varied with the full establishment of the Abbasids, the expansion of the bureaucracy and the concomitant refinement of literary style. Though each type has expressions peculiar to it, they nevertheless draw upon a common repertoire of phrases which, though gradually becoming more extensive and elaborate, remains fairly limited for the first two or three centuries and exhibits a high degree of recurrence of set formulae. An overview of these will be given in the following section ("The Content"), and after that some introductory remarks on their origins and significance will be proffered ( "The Context" ).
It is to God that most Arabic inscriptions are direeted, usually, though not always, beginning with a simple exclamation (Allāhumma) or calling upon His narne (bismillāh / bismillāh al-rahmān al-rahīm). In addition He may be invoked, especially in early graffiti, by the term "Lord": "my Lord" / rabbī, "Lord of the Worlds" / rabb al-‘ālamīn, "Lord of all the people" / rabb al-nās ajma‘īn, or the Lord of one or a pair of prophets (e.g., rabb Mūsā wa-Hārūn, rabb Muhammad wa-Ibrāhīm). Though the word "Lord" occurs very often in the Qur'ān, it is linked with prophets only in the case of Moses and Aaron (7:122, 20:70, 26:48). The variety of prophets adduced in early graffiti is, therefore, unexpected - though the Qur'ānic doctrine of a succession of messengers of equal standing is assumed - and illustrates that these texts are not a mere rehashing of Qur'ānic pericopes. Occasionally, God is relegated to second place in an inscription and the owner first introduces himself / herself, simply stating "I am ... " / anā ... and then proceeding to give his / her name.
The presentation of some sort of petition to God is the most persistent feature of Arabic inscriptions. By far the most common plea is for forgiveness, mercy, blessing or approval (using some form of the roots ghfr, twb, ‘fw, rhm, brk, rdy); it is found already by the 40s / 660s and remains in use for centuries thereafter. Also very popular is the converse of this, namely seeking to be spared from bad things (usually expressed by ‘ādha bi ... min / ‘afā min). Requests for forgiveness may be elaborated upon by various epithets either qualifying the supplicant ("Your servant" / ‘abduka, "before his death" / gabla mawtihi, "in life and in death" / hayyan wa-mayyitan, "neither perished nor lost" / ghayr hālik wa-la mafqūd, "the day he dies and the day he is raised alive" / yawm yamūtu wa-yawm yub‘athu hayyan) or qualifying the transgressions to be taken into account (dhanbahu al-‘azīm, dhanbahu kullahu / dhunūbahu jam‘īan / kull dhanb adhnabahu, mā ta‘lamu / ahsayta min dhanbihi, mā taqaddama min dhanbihi wa-mā ta'akhkhara). Once the owner of the inscription has seen to his/her own needs, or sometimes beforehand, he / she may ask that the same benefits be granted to other parties, such as relatives, the Muslim community, prophets and angels, and often, in conclusion, the reader of the inscription and/or whoever says "amen, amen, Lord of the Worlds" or just "amen." For this purpose, the phrase "incline unto" / salli ‘alā is often used, especially for the Prophet Muhammad, the request being put just to God or to God and His angels and possibly His messengers.
While these simple demands do not cease, more elaborate entreaties begin to appear, displaying a greater confidence towards their divine addressee. The supplicants wish to be admitted into paradise, to be united with their Prophet, to be preserved from the torment of the Day of Reckoning and to be spared the punishment in the tomb, to be saved from the Fire and receive succour on the Day of Resurrection, to be instructed in His proof, to be rewarded for the best of their deeds, to have their devotions and good actions accepted, to receive God's favour and guidance, and to be granted good health, virtue and prosperity.
In the case of offlcial inscriptions, there will often be included an appeal to God to deal favourably with his deputies and their aides. For example, on a dam near Tā'if in the Hijāz, built in 58 / 678 at the command of the caliph Mu‘āwiya, an inscription calls upon God to "strengthen him and help him and let the faithful profit by him." On a bridge in Fustat, commissioned by the governor ‘Abd al-‘Azīz b. Marwān in 69 / 688, God is solicited to "bless him in all his affairs, strengthen his rule as You see fit, and cheer both him and his entourage." And at the palace of Muwaqqar, an invocation in favour of Yazīd II (720-24) calls upon God to "make him virtuous, protect him, extend his life and ease (al-yusr), and perfect for him His favour and generosity in this world and the next."
Otherwise there are found various short adjurations, such as "may God treat him generously" / akramaku Allāh, "may He prolong his life"/ atāla baqā'ahu, "may God guarantee his reward" / awjaba Allāh ajrahu, "may God preserve him and enhance his reward and give him good recompense" / hafizahu Allāh wa-‘azzama ajrahu wa-ahsana jazā'ahu, "may God lend him strong support" / a‘azza Allāh nasrah?
Occasionally, an inscription will contain a curse. This will usually be; one of two kinds: either the inscriber will be giving a warning to any potential vandal, such as "may the right hand of anyone who defaces it be paralysed" / shallat yamīn man mahābu, "may God requite anyone who defaces it" / ajzā Allāh man mahābu; or else he will ask God that favour be denied to such and such a person, that is, do not forgive them, do not bless them and so on. Though in the inscriptions of other cultures, such as Nabatean, curses are common on tombstones, in the case of Arabic they seem only to appear on graffiti.
Professions / Declarations
The second most common objective of inscribers is to convey some of the essentials of their faith and to pronounce their allegiance to it. This may be stated outright or else an introductory word will be used, either inna or one of the verbs shahida / āmana, sometimes together with a phrase emphasising the veracity of the profession. In many epitaphs, this testimony will then be concluded by the set formula: "by this he has lived, by this he has died and by this he will be raised alive, if God wills."
God and His attributes. Always in first place is some declaration about God. In the simplest graffiti, inscribers may say no more than that they believe / trust / take refuge / seek shelter / have confidence in God (using some form of the roots āmana / wathiqa / ‘ādha / i‘tasama bi / tawakkala ‘alā), or that He is their "Lord" / rabb, their "patron" / walī, their "might" / ‘izz. Very commonly, however, various epithets and predicate phrases will be assigned to Him, almost all corresponding to portions of Qur'ānic verses: "the Clement, the Generous" / al-halīm al-karīm, "Praiseworthy, Glorious" / hamīd majīd, "the Living, the Ever-lasting" / al-hayy al-qayyūm, "the Forgiving, the Compassionate" / al-ghafūr al-rahīm, "the Mighty, the Wise" / al-‘azīz al-hakīm, "the most Compassionate of the compassionate" / arham al-rahīmīn, "the Omnipotent" / ‘alā kulli shay'in qadīr, "the Lord of the Worlds" / rabb al-‘ālamīn, "the Manifest Truth" / al-haqq al-mubīn; "to Him belongs sovereignty and praise" / lahu al-mulk wa-lahu al-hamd, "He gives life and brings death"/ yuhyī wa-yumītu, "in His hand is the good" / bi-yadihi al-khayr. Very frequently also His unity will be affirmed, both by simple assertions that He is one (ahad, wāhid, wahdahu) and has no associate or offspring (lā sharīka lahu, lam yattakhidh sāhibatan wa-lā waladan), and by recourse to Qur'ān 3:18 ("God is witness that there is no god but He, as also are the angels and men of knowledge; He acts with justice, there is no god but He, the Mighty, the Wise") and Qur'ān 112 ("Say He is God the One, God the Eternal, He did not beget nor was begotten, and there is none equal to Him" ). Otherwise, a number of miscellaneous expressions are encountered, most drawn from the Qur'an, a few not so.
Muhammad and his attributes. After God, the name featuring most frequently is that of the Prophet. He is portrayed as "the messenger of God," "the servant of God and His messenger," "Your servant and Your prophet," "Your prophet and Your intimate," "the prophet of compassion," "a herald a warner" / bashīr nadhīr, "the illiterate Arab prophet" / al-nabī al-ummī al-‘arabī, "the seal of the prophets" / khātim al-anbiyā', "the prophet who follows and is guided along the right path" / al-nabī al-rashīd al-mahdī. His mission is described by recour to a variety of Qur'ānic verses, most often 9:33 ("Muhammad is the messenger of God whom He sent with guidance and the religion of truth to make it prevail over all religion, even if the associators are averse"), 37:37 ("he brought the truth and confirmed those (already) sent)," 36:70 ("to warn whoever lives and that the word may be fulfilled against the disbelievers" ), 33:45 ( "a summoner to God by His permission and a light-giving lamp"). Again, very occasionally a non-Qur'ānic expression is used: "he imparted the message and advised his community and served God until death came to him"/ ballagha al-risāla wa-nasaha li-ummatihi wa-‘abada Allāh hattā atāhu al-yaqīn, "Islam is as he described and the religion is as he laid down" (al-dīn ka-mā shara‘a), "You guided us by him to Islam.''
The Hour and the Resurrection. A testimony on a tombstone from Egypt dated 174 / 790 cites Qur'ān 22:7: "the Hour is coming, of that there is no doubt, and God will raise those who are in the graves," and, thereafter, the verse becomes a staple feature of epitaphs from this country. In addition, writers of graffiti often proclaim that they believe in "the last day" / al-yawm al-ākhir. The nascent Muslim community was much preoccupied by eschatological questions, when the End would come and what would be the events of the Resurrection, and such credal statements as the above and the Fitan chapters in hadīth collections bear eloquent witness to this speculation.
Heaven and Hell. Descriptions of Heaven and Hell are scattered throughout the Qur'ān and the two terms appear on inscriptions from the 80s / 700s onwards, most commonly as part of a request that the inscriber be admitted into the one and exempted from the other. They also feature in professions of faith where their reality is affirmed (al-janna haqq wa-l-nār haqq), and here they may be joined by other elements such as that "death (al-mawt) is a reality," "the resurrection (al-ba‘th / al-qiyāma) is a reality" and "the meeting with Him (liqā'uhu) is a reality." The nature and status of each, especially the last, was disputed in early Islam, and so the frequency of their mention on tombstones from the 170s / 790s onwards perhaps give us an indication of when their acceptance became part of mainstream belief and their denial seen as heretical.
God's Will. On an epitaph of 179 / 795, it is asserted of the deceased that "he believed in al-qadar, its good and its bad." Again, this is a reference to a much contested issue, the conflict between God's justice and man's responsibility on the one hand and His omnipotence on the other. But though there were many finer points to be resolved, by the date of this tombstone the matter had largely been settled and those who opposed the idea that God decreed the evil in the world as well as the good were branded as heretics.
The Qur'ān. Though cited directly or alluded to innumerable times, the Qur'ān is not itself mentioned on inscriptions as a separate entity until Abbasid times. On a second / eighth century graffiti from northern Arabia there appears the expression: "he believes . . . in every messenger He has dispatched and book He has sent down" / yu'minu . . . bi kull rasūl arsalahu wa-kitāb anzalahu. And there is a reference to "the Book of God" in the inscription of 135 / 752 commissioned by the caliph Saffāh for the mosque of Medina. The first explicit statement features on a tombstone from Egypt dated 195 / 810: "he testifies that the Book is truth which God sent down with His knowledge. Falsehood does not come to it from before it nor from behind it, a revelation from One wise, praiseworthy. He believes in what is in it, the sure and the doubtful, the abrogating and the abrogated, from its beginning to its end." On another epitaph of the same period, this time from Mosul, the owner bears witness that "the Qur'ān is the speech of God (kalām Allāh), sent down, uncreated (ghayr makhlūq).'' These are already quite elaborate characterisations, the product of a long period of debate, and so can only be used as a terminus ante quem for conceptions of the Qur'ān.
Islam and the Muslims. The term "Islam" features in inscriptions from a very early date, but its nature or content is never explained. It is simply stated that "Religion with God is Islam" (Dome of the Rock, 72 / 691), "our religion is Islam" (Mosque of Medina, 87 / 706), "my religion is Islam" (Jabal Usays, 119/737). One thing is clear, however, namely that Islam is closely connected with Muhammad. It is "the religion of truth" entrusted to him by God, it is what he guided his community to and what he described. The associated term "Muslim" does not seem to appear until 123 / 741 (inscription at ‘Ayn al-Jarr), the word "believer" being preferred before this. Thereafter it becomes more common and testifies to a greater sense of communal feeling among Muslims, evidenced also by such expressions as "the greatest calamity for the people of Islam is what befell them with (the death of) the prophet Muhammad." 
The primary source upon which early Arabic inscriptions draw is, without doubt, the Qur'ān. Occasionally an inscription may be composed of nothing more than a Qur'ānic citation. Sometimes a Qur'ānic verse may be quoted in full within the text of an inscription, and some examples of this have been given above. More common still, an inscription will be an eclectic blend of words and phrases taken from different verses of the Qur'ān, often slightly modified and / or supplemented as required or desired. For example, the text rabbī rabb al-samawāt wa-l-ard wa-mā baynahumā lā ilāha illā huwa fa-ana attakhidhuhu wakīlan is assembled from 26:24 (or 37:5, 38:66, 44:7, 38) and 123:9, with a small amendment to personalise the quotation (fa-ana attakhidhuhu rather than fa-ttakhiduhu). The text rabbī Allāh wa-dīnī al-islām ‘alayhi tatuakkaltu wa-ilayhi unību wa-ilayhi al-masīr borrows from 40:28, 11:88 (or 42:10) and 5:18, and inserts the non-Qur'anic phrase dīnī al-islām. And the text urzuqhu min fadlika wa-adkhilhu rahmataka wa-atimma ‘alayhi ni‘mataka wa-j‘alhu min al-muflihīn takes from 24:38 (paraphrased), 7:151, 48:2 (or 5:3), and concludes with a request to be made "one of the prosperous."
Even those inscriptions which do not explicitly cite or paraphrase the Qur'ān will mostly be comprised of words and concepts familiar from the Qur'ān. Yet this need not imply direct influence. Each of the words in the expression taqabbal min ... salātahu wa-sawmahu, which occurs in the text published in the Appendix below, do feature in the Qur'ān, and prayer and fasting are Qur'ānic concepts, but it is not in itself a Qur'ānic phrase and the most we can say is that the Qur'an is an indirect source. One possibility is that many of these texts, especially the graffiti, reflect prayers of supplication (da‘awāt) that were current in this period. "Supplicate to me (ud‘unī) and I will answer you" (40:60), "I answer the supplication of the supplicant (da‘wat al-dā‘ī)" (2:186), says the Qur'ān, and this is reinforced by the Prophet, who declared that "supplication is the marrow of worship" and recommended a whole host of occasions when such activity would be appropriate. Certainly, the formulae in some inscriptions do find their counterpart in later collections of da‘awāt. For example, the first part of a graffito dated 64 / 683 found near Karbala in Iraq conforms to one of the prayers said at the Muslim festival of the ‘Id (Allāhu akbaru kabīran wa-l-hamdu lillahi kathīran wa-subhāna Allāhi bukratan wa-asīlan); refuge from the punishment in the grave (‘adhāb al-qabr) is sought in both epitaphs and prayers; and the request to be reunited with (‘arrif bayna ... wa-bayna ...) someone in the hereafter is frequently encountered in physical and literary texts. Moreover, the frequency with which inscriptions conclude with a blessing for "the one who reads (this text)" and then "says amen" or "supplicates on behalf of (da‘ā li)" the owner conveys an impression of some sort of ritualistic activity, whether private prayer or communal recitation. The latter must often have been the case, for qara'a would seem to mean "read out loud" and not just "read," as is suggested by such expressions as: "O God, forgive ... the one who reads (this text) and the one who hears, then says amen," and furthermore many of the formulae in inscriptions have a marked rhythmic quality, as for example: sallī ‘ala abīna Ibrahīm | wa-‘alā malā'ikatika al-muqarrabīn | wa-anbiyā'ika al-mursalīn | amīn rabb al-‘alamīn. This oral context would, in addition, explain why a few set phrases tend to feature again and again in so many of our inscriptions.
A third possible inspiration for these texts that should be considered is foreign influence. There is no doubt that one can adduce parallels between our texts and the writings of other peoples. In Greek Christian inscriptions, for example, one finds "O Lord accept ... "/ Kyrie prosdexe (cf. Allāhumma taqabbal), "O Lord have mercy . . . " / Kyrie eleêson (cf. Allāhumma irham), "O Lord forgive his sins" / Kyrie synchôrêson tas amartias autou (cf. Allāhumma ighfir li ... dhanbahu), " ... Your servant" / ... ton doulon sou (cf....‘abdaka). In Thamudic inscriptions, it is common to assert "I am . . . " / 'n . . . followed by one's name (cf. ana . . . ); in Safaitic graffiti, inscribers will seek refuge in the deity (‘wdh b h'lh; cf. ‘ādha billah) and curse those who damage their work; and regard for "the one who reads this writing" is found in Thamudic, Nabatean, Safaitic and Greek texts. But such parallels on their own demonstrate little. Petitions and offerings to a divine being, declarations of one's own identity and appeals to the passerby recur in the inscriptions of numerous societies, and one would have to carry out a very careful comparative study of the content, function and structure of a representative sample of phrases before one could even begin to posit borrowing. And, in general, it must be said that it is the distinctiveness rather than the derivativeness of Arabic inscriptions that is most striking.
Arabic inscriptions are found throughout the lands ruled by the Muslims in our period. Most of the known graffiti are from the arid zones of northern Arabia, the Negev, Jordan and Syria, and most of the epitaphs are from Egypt, but that may simply reflect the extent of work carried out so far. The archaeology of Islamic Iraq and Iran has received little attention, but the few inscriptions that we do have from these countries still accord well with those found further west. There is, of course, some regional diversity, but perhaps more impressive is the degree of homogeneity encountered from a very early date. In the first place, all inscriptions are in the same standard Arabic, and most show close affinity with the style and vocabulary of the Qur'ān. Secondly, all are in the same Kufic script, which does exhibit variations depending on the surface chosen, the tools used, and the skill and social status of the inscriber, but which is nevertheless remarkably uniform. Thirdly, all evidently draw upon a common stock of words and phrases, for one sees certain expressians crop up time and time again in many different places. For example, the request that God "forgive him his former and his latter sins" is found in an inscription of 64 / 683 in southwest Iraq, one of 92 / 710 in eastern Jordan, one of 110 / 728 near Palmyra, one of 112 / 730 in the Negev, one of ca. 115 / 733 in western Jordan, one of 121 / 739 in northern Arabia, one of 185/801 in Egypt, as well as in countless undated inscriptions.
A similar picture is presented by two other phenomena, namely the Arabs' use of a new era and of a new name for themselves ("emigrants" / muhājirūn), both of which are likewise attested throughout the Middle East from a very early date. All of this suggests that the early Muslims formed a homogenous elite who were united by a shared ideology and common religious idiom. A measure of individual creativity is discernible in the inscriptions, but it is subject to clear limitations. Presumably, this idiom was controlled from the centre and thence disseminated to the periphery, from the imperial court to the regional capitals to the outlying districts, since otherwise it is difficult to see how its relative uniformity could have been preserved. Certainly this would seem to be borne out by the frequency with which the earliest dated occurrence of a phrase in graffiti will be a couple of decades later than its earliest dated occurrence in an imperial inscription. And it is more frivolously confirmed by the following anecdote, which makes the point that fashion and taste were set from above: "When people met in Walīd's time they would talk about nothing but building and construction; next Sulaymān came to power ... and they would ask one another about copulation and slave girls; and then when ‘Umar b. ‘Abd al-‘Azīz held office, people would meet and discuss their night prayers, their memorisation and recitation of the Qur'ān and their fasting.''
The majority of the people that are named in inscriptions are not known to us from literary sources. Many others are given in too brief a form, only first name and father's name, for us to be sure which, if any, of a number of persons is intended. Inevitably, the few that are known to us from elsewhere are those of high status: official inscriptions, of course, bear the names of caliphs and their governors, but graffiti too seem often to have been commissioned by figures of high standing, then often carved for them by trained scribes. Thus graffiti from the area around the Umayyad castle of Jabal Usays in southeastern Syria have among their owners four of the sons of the caliph Walīd I (705-15) as well as others close to the ruling family, and many of the texts from the Ruwāwa district of Medina were penned by dignitaries of the city including religious scholars. It must, however, also be the case that many graffiti, especially those roughly scratched with a flint or knife on loose rocks and rock faces out in the desert, are the work of semi-nomadic or nomadic peoples, able just to write their name and a few basic phrases, consciously or unconsciously carrying on the practice of the Thamudic and Safaitic tribes that roamed these lands before them.
Whether this material would be able to inform us in some way about the persons named therein and their milieu is difficult to say while it remains little studied and scattered across diverse publications. But certainly a number of questions spring to mind for which it might be of some help. Does, for exarnple, the persistence of pagan names/proliferation of Muslim names say anything about the process of Islamisation, or the rate of conversion to Islam? What can the concentration of inscriptions found in an area tell us about settlement patterns, economic trends, population density, and so on? How closely is the skill of execution of an inscription linked to the social status of its owner, and could this then tell us something about the degree to which literacy penetrated down the social ladder? In other words, can this material be used to write social history, and if so in what ways?
The content of early Arabic inscriptions, as outlined above, shows that the ostensible aim of most inscribers was to win God's indulgence and to affirm their adherence to His religion and their allegiance to His Prophet and His community. On tombstones, it is natural that the deceased should wish to have a statement of their faith and their trust in God. And on the various objects of officialdom - monuments, coins, seals, weights, etc. - it is expected that there should be some affirmation of the principles by which the community is governed. But what about graffiti? Why did people feel impelled to scribble on rocks?
While traversing Syria on a military expedition into Byzantine territory, the caliph Ma'mūn (813-33) stopped off at one point to admire an ancient church and observed to his companion: "It is the wont of travellers abroad and whoever is far from family and friends that, when they enter upon a remarkable place or a celebrated site, they leave there for themselves some trace (athar), seeking a blessing by appealing to others away from home." Surveys of pilgrimage routes tend to yield a good crop of graffiti, and the same is true for strategic sites such as watering holes and high places, so it is indeed likely that many of these texts were written by those on a journey.
Epigraphic surveys of a specific area often reveal that one person or one lineage (parent, children, grandchildren) is responsible for a number of inscriptions. This could be an indication that that person or lineage was of some importance in the local community. If a scribe was commissioned, this would particularly be the case, since presumably such a transaction would involve a sum of money. To inscribe a text on rocks full of rough surfaces, bulges and cracks in such a manner that it is clear, well-formed and properly joined, even aesthetically pleasing, demands much time, care and skill. A scribe may, therefore, be able to ask a considerable fee for a well-executed graffito. In other words, the writing of inscriptions could be a way of demonstrating status as well as of exhibiting affiliation.
The frequency with which God's unity and Muhammad's divine mission are proclaimed, in all types of inscriptions, suggests that at least some inscribers had apologetic aims. This is most evident in certain monumental texts, in particular that on the Dome of the Rock in Jerusalem, which blatantly asserts that "religion with God is Islam" and upbraids Christians for saying God is "three." But it is also manifest in many epitaphs and graffiti. Both, for example, commonly cite Qur'ān 9:33 ("Muhammad is the messenger of God whom He sent with guidance and the religion of truth to make it prevail over all religion even if the associators are averse"), a verse that first appears on ‘Abd al-Mālik's reform coinage of 77 / 696 and one that is overtly confrontational. A text did not need, however, to be polemical in content to convey a message to the non-Muslim population, who would have been almost everywhere a majority in the period under consideration here. The fact that Arabic writings, even if unintelligible to the reader, were now found in the places and on the objects where once would have been inscribed Greek or Persian, was a clear sign of the Muslims' possession of political dominion, which in its turn buttressed their claim to possession of religious truth.
Accompanying a number of the graffiti are drawings of various domestic and wild animals (horses, bulls, ibex, lions, camels, dogs, etc.), as well as of humans (men, women and children). These are usually thought to have been executed by pre-Islamic peoples (Thamudic, Nabatean, Safaitic), and many, perhaps the majority, were so. Some, however, definitely belong to the Islamic period. This is clear from inscriptions such as that in the Appendix below where the animal has been drawn on top of the Arabic text. And it has also been proved to be the case by an interesting inscription from the Jawf region of northern Saudi Arabia, which stands next to a sketch of a lion and states: "Ya‘lā b. Yazīd made this lion in the year one hundred and forty-four" (761). These drawings have never been studied, and yet would probably provide interesting insights into the continuity of life and culture of the tribes of this region and into Muslim attitudes towards images, and so on.
Much more could be said, but at this stage I have many more questions than answers, so it is perhaps best to wait until further research has been done. Having considered what this material might be able to tell us, I would like to conclude by emphasising what it cannot tell us. It has been inferred from the relatively limited stock of specifically Muslim concepts present in the early inscriptions that either Islamicisation was as yet shallow, or that Islam itself was as yet little developed. But, aside from the fact that the medium imposes severe constraints upon the complexity and variety of a message, these texts were never intended as catechisms of Islamic doctrine. Moreover, to say that "the inscriptions lack typical Islamic expressions" or "exhibit indeterminate monotheism" just because they do not mention Muhammad is to misconstrue Islam, which is not primarily Muhammadanism, but rather subordination to an omnipotent and unique God. So the very common formula lā ilāha illa Allāh wahdahu lā sharīka lahu, though not incompatible with Judaism or Christianity, can nevertheless be said to be specifically Islamic. This is not to say that Islam did not develop over time, but that one cannot hope to isolate the different stages of its growth from inscriptions; firstly because only about three percent or so are dated, and secondly because very many have remained simple and "un-Muhammadan" in their expression right up until the present day. Archaeology may well offer the "hard facts" of material remains, which ... "represent what somebody once did, not what some contemporary or later writer says they did." But one is often less interested in what someone did than in why they did it, how, with what consequences, and so on; and for answers to these questions we are in some ways worse off than with literary texts, since we have to supply our own interpretations rather than evaluate those already given.
The following inscription was found by the late Bill Jobling, of the University of Sydney, in Wadi Shireh, which runs into Wadi Rabigh, itself close to Jabal Ramm in southern Jordan. It is a fine example of an Umayyad graffito and is clearly legible except for the sixth and seventh lines, which are partially obscured by the drawing of a horse mounted by a lance-bearing rider.
The translation of the inscription is:
- In the name
- of Allah, the Merciful,
- the Compassionate; O Allah
- accept from ‘Abd al-‘Alā' bin Sa‘īd
- his prayer and his fasting, keep him
- among his family, and act as his deputy for his ...
- and make him virtuous. You have power over all things.
- May God grant him peace. And peace be upon him
- and mercy of God and His blessings. And
- it was written in Ramadān
- of the year one hundred and nine.
lines 4-5: "accept ... his fasting": note that it is Ramadān, so ‘Abd al-‘Alā' is emphasising to God his observance, which is more notable if, as is likely, he is on a journey.
lines 5-6: "keep him among his family": this is a very interesting expression, unique among graffiti as far as I know, it being a request to enjoy something of this world rather than the next. Note that in Qur'ān 84:13 the one who "lived joyously among his people" (fī ahlihi masrūran) was one too rooted in this life, who "thought that he would never return (unto God)," so there is something of the Jāhilī spirit in this request, though it is, of course, a very natural request to make for one away from home.
line 6: "act as his deputy for his ... ": cf. Ahmad b. Hanbal, Musnad (Cairo, 1895), V, p. 83:
(and see A.J. Wensinck, Concordance et indices de la tradition musulmane [Leiden, 1933-88], II, p. 63, for further examples), This expression, used by on anticipating death or away from home, also occurs in ‘Ushsh, "Jabal Usays," no. 55 (ukhlufhu fī ahlihi), but it remains only a tentative suggestion, since our inscription is damaged at this point. For the final word, one would expect something like banīhi ("children," "tribe").
line 8: "peace be upon him": one could read is 'alhu ‘alayhi ("ask Him for him ... "), but this would require reading the very clear final mīm of 'slm as a hā and discounting the wa / "and" at the beginning of line 9. It is probably better to assume the inscriber has left out the lām of the definite article from al-salām, which is very understandable since it is not pronounced.
lines 10-11: this date corresponds to December 727-January 728. I read tis‘a / "nine" rather than sab‘a / "seven," as the first tooth seems to be slightly higher than the following three.
*I am much indebted to the late Yehuda Nevo, and to Nitzan Amitai-Preiss, Frédéric Imbert and Michael Macdonald for thought-provoking discussion on the subject of epigraphy. Note that editions of inscriptions will be cited in brief in the text of the article and in full at the end.
 Exceptions are Fred M. Donner, "The Formation of the Islamic State," JAOS 106 (1986): 283-96 (though chiefly using papyri); Solange Ory, "Aspects religieux des textes épigraphiques du début de l'Islam," Revue du monde musulman et de la Méditerranée 58 (1990): 30-39; Yehuda Nevo, "Towards a Prehistory of Islam," JSAI 17 (1994): 108-41. Some of the corpuses do also have a good analytical section, in particular see that of Imbert, Jordanie; and the entry on "Kitābāt" in EI2 is helpful.
 Early Arabic inscriptions can often be assigned an approximate date, for the so-called Kufic script in which they are written did change over time, but the study of such developments is still in its infancy. See especially A. Grohmann, Arabische Paläographie (Vienna, 1967-71), esp. II, pp. 71-231; Beatrice Gruendler, The Development of the Arabic Scripts (Atlanta, Georgia, 1993); Yasser Tabbaa, "The Transformation of Arabic Writing," Ars Orientalis 21 (1991): 119-48, and idem, 24 (1994): 119-47.
 The earliest dated literary papyrus is of 229 / 843 (R.G. Khoury, "L'importance d'Ibn Lahī‘a et son papyrus conservé a Heidelberg," Arabica 22 : 11-12), though some fragments belonging to the mid- to late eighth century are found in A. Grohmann, Arabic Papyri from Khirbet el-Mird (Louvain, 1963), nos. 71-73.
 Personal statements can be found on handmade objects too, notably oil lamps, gemstones (then usually talismans), pottery, textiles, etc. For some general comments on inscriptions and the variety of media that bear them, see Heinz Gaube, "Epigraphik," in W. Fischer, Grundriss der arabischen Philologie, Band 1: Sprachwissenschaft (Wiesbaden, 1982), pp. 210-25.
 I am interested here in giving a general impression of the material to the non-initiate, and so concentrate on the frequently occurring formulae. The reader should, however, be aware that the full range of expression is quite extensive and a number of unusual inscriptions will be encountered: e.g., Sharon, "Rehovoth and Sinai," no. 1 ("any friend who is not (a friend) in God, then his friendship is aberrant, lifeless, empty, and his attachment ephemeral," kull sadīq laysa fī Allāh wudduhu ghayy kall khālin wasluhu ghayr dā'im); Rashid, Makka, nos. 17 (dated 98 / 716), 19 (189 / 805) and 28, which contain lines of poetry.
 The latter two expressions appear together in Nevo, Negev, MA4210 and MA4265 (dated 85/704).
 The former is found in an inscription of 92 / 710 (Abbott, "Kasr Kharāna"), the latter in one of 117 / 735 (Nevo, Negev, HS3155) and one of 121/739 (Muaikel, Jawf, no. 1). A graffito of 64 / 683 near Karbala in Iraq has "Lord of Gabriel, Michael and Serafiel" (Sanduq, "Hafnat al-Abyad"). For other examples see the list in Nevo, Negev, pp. 141-42.
 E.g., Rashid, Medina, nos. 38 (anā Mansūr ibn ‘Atā' ibn Rabī‘a āmantu billāh wa-rasūlihi, 140 / 757), 53 (anā ‘Uthmān ibn Hafs u'minu billāh al-‘azīm, 120/738); ‘Ushsh, "Jabal Usays," nos. 7, 49, 58; Rashid, Makka, nos. 10, 20, 22, 30-33, 37, 41-42, 49 (a few, e.g., nos. 5-8, do not even mention God, but just say "I am so-and-so").
 Most often they will be used as verbs in the imperative or optative perfect, but sometimes as nouns after sa'ala ("to ask"); e.g., "'‘āliya asks God His forgiveness His mercy and His approval" / inna ‘āliya yas'alu Allāh maghfiratahu wa-rahmataHu wa-ridwānahu (Muaikel, Jawf, no. 15).
 Sharafaddin, "Darb Zubayda," plate 49 (rahmat Allāh wa-barakatubu ‘alā ... 40 / 660); Grohmann, AI, Z202 (ighfir li . . ., 46 / 666). A request for forgiveness and mercy is also found on a tombstone dated 31 / 651 (Hawary-Rached, Stèles, no. 1).
 The inscription of ‘āliya quoted in n. 9 above continues: "and he seeks refuge in Him from His wrath and His aversion and seeks refuge in Him from the Fire" / wa-ya‘ūdhu bihi min sakhtihi wa-maqtihi wa-ya‘ūdhu bihi min al-nār. For examples of the use of the root ‘afw, see nn. 18 and 20 below.
 There are a number of variants of the latter expression,
some given by Nevo, Negev, pp. 8-9.
 E.g., Imbert, Jordanie, no. 114 ("O God, grant forgiveness to Khālid ibn al-Sawwār al-Yamāmī and to his parents and to those whom they begot, amen, and to the Muslim men and women, and to Muhammad your servant and your messenger," second-third / eighth-ninth century); Sauvaget, "Glanes," no. 10 (... ighfir li ... wa-li-wālidayhi wa-li-l-wālidāt wa-li-l-mu'minīn wa-li-l-mu'mināt, Jabal Druze, Umayyad); Nevo, Negev, HS 3155 ( ... ighfir li ... wa-li-wālidayhi wa-ma waladā, 117 / 735); Rashid, Medina, no. 21 (... yastaghfir Allāh al-karīm lahu wa-li-man wal-adahu mu'minan wa-li jamī‘ al-muslimīn, second / eighth century); Baramki, "Al-bādiya al-sūrīya," nos. 8, 34, 37, 48-49, 52, 63, 67-68, 77, 86, 91, 93.
 E.g., Sanduq, "Hafnat al-Abyad" (64/683); Abbott, "Kasr Kharāna" (92/710); Cantineau, Palmyre, no. 39 (110 / 728); Couroyer, "Beit Gibrin" (first / seventh-eighth century); ‘Ushsh, "Jabal Usays," nos. 1, 58, 74; Nevo, Negev, EL200C, SC301, SC305, GM389, MA453, MA4283, MA4319, HR510; Imbert, Jordanie, nos. 1, 5, 11, 22-23, 72, 82, 106, 151, 156; Moraekhi, Medina, B11, L4a, L17, R8; Baramki, "Al-bādiya al-sūrīya," nos. 22, 33, 56, 65, 71, 77. There may also be a request for favour towards "the one who wrote this writing," who is often patently distinct from the one who commissioned the inscription.
 E.g., Kessler, "Dome of the Rock," has five examples, including Qurān 33:56 which is also found on graffiti (e.g., Miles, "Tā'if," p. 241); Ory, "'Ayn al-Garr," no. 1 ("May God incline unto all the Muslims," 123 / 741); Nevo, Negev, HR508 ("O God, may You and Your angels incline unto him who wrote this writing"); Hawary-Rached, Stèles, no. 13 ("O God, incline unto Muhammad the Prophet, Your servant and Your messenger, Your prophet and Your intimate, just as You inclined and gave blessing unto Your friend Abraham and the family of Abraham," 188 / 804). Nevo, Negev, p. 86, lists thirty-three occurrences of the tasliyya, most on behalf of the inscriber; see further Imbert, Jordanie, pp. 532-38. Throughout I translate salla as "incline unto," reflecting the Aramaic sense of the root, in order to avoid saying that God "prays."
 E.g., Nevo, Negev, MA4265
(adkhilhu al-janna, 85 / 704); Ory, "'Ayn
al-Garr," no. 1 (adkhilhum jannāt al-na‘īm,
123/741); ‘Ushsh, "Jabal Usays,"
no. 18 (adkhil Muhammad ibn al-Walīd bi-dhimmat Allāh,
mid- to late Umayyad); Nevo, Negev, MA4339 (yas'alu
Allah al-janna, 160 / 776); Hawary-Rached, Stèles,
nos. 13 (sayyirhu ilā jannatika, 188 / 804), 14 (askinhu
jannāt al-na‘īm, 190 / 806); Baramki, "Kufic
Texts," no. 1 (akhrijnī min al-dunyā sāliman
‘āliman wa-adkhilnī al-janna āminan). Both terms,
janna (lit. "garden") and jannāt al-na‘īm
(lit. "gardens of bliss / delight"), occur frequently in the Qur'ān
(137 and 10 times respectively). The expression wassi‘ ‘alayhi
madākhilahu, commonly found in epitaphs (e.g., Hawary-Rached, Stèles,
nos. 4, 13, etc.; Imbert, "Qastal al-Balqā',"
nos. 6-8; cf. Grohmann, AI, Zll: adkhilhā
madkhalan karīman), is presumably also meant to refer to paradise
(cf. the use of madkhal in Qur'ān 4:31, 17:80, 22:59).
 Found on numerous epitaphs from 179 / 795 onwards, most commonly expressed as ‘arrif / ijma‘ baynahu wa-bayna nabiyyihi or alhiqni bi-nabiyyihi (see Hawary-Rached, Stèles, nos. 3-4, 13; Imbert, "Qastal al-Balqā'," nos. 2, 7-8, 14, 16; Grohmann, AI, Zl9, Z23; cf. Qur'ān 26:83: alhiqni bi-l-sālihīn).
 E.g., Ory, "'Ayn al-Garr," no. 1 (‘āfāhu min sharr yawm al-hisāb, 123/741); Grohmann, AI, Z73 (amminhu min al-‘adhāb yawm al-hisāb).
 Hawary-Rached, Stèles, no. 4 (180/796); on the punishment in the tomb see A.J. Wensinck, The Muslim Creed (Cambridge, 1932), pp. 117-21. Grohmann, AI, Z178, contains a request to be spared ‘adhāb Allāh, a very common phrase in the Qur'ān.
 E.g., ‘Ushsh, "Jabal Usays," nos. 72 (as'alu al-janna wa-a‘ūdhu bihi min al-nār, by someone who is also the owner of an inscription dated 113 / 731), 58 (a‘udhu bihi min al-nār yawm al-ma'āb); Grohmann, AI, Z73 (ikbit lī wa-lahu min al-nār), Z143 (‘āfāhu min al-nār); Rashid, Makka, no. 11 (uktub lī barā'a min al-nār). An inscription of 154 / 771 from Shubayka asks for "deliverance (al-najāt) from the Fire and immunity (al-amān) from the Punishment and aid (al-‘awn) on the Day of Resurrection" (Imbert, Jordanie, no. 105); and an epitaph of 186 / 802 (RCEA, no. 67) cites Qur'ān 3:16 ("Forgive us our sins and spare us the punishment in the Fire").
 This phrase (laqqinhu / laqqihi hujjatahu) occurs frequently in epitaphs (e.g., Hawary-Rached, Stèles, nos. 3 - dated 179 / 795, 10, 13, etc.; Imbert, "Qastal al-Balqā'," nos. 2, 6-8, 10). Cf. Qur'ān 6:83 ("This is Our proof which we bestowed upon Abraham") and 6:149 ("To God belongs the conclusive proof").
 Thus an epitaph of 188 / 804 (Hawary-Rached, Stèles, no. 13); cf. Moraekbi, Medina, L13 (as'alu Allāh na‘īm al-ākhira wa-husn thawāb al-dunyā).
 E.g., Abbott, "Kasr Kharāna" (as'aluka an taqbala minhu salātahu wa-haybatabu); Muaikel, Jawf, nos. 24 (taqabbal hasanāt ... ), 25 (taqabbal ... salātahu); Hamed, Arabie Saoudite III, plate 224 (taqabbal minhu hajjatahu, 91 / 710) and see Appendix below.
 Nevo, Negev, SC305 (atimma ni‘mataka ‘alayhi wa-'ahdihi sirātan mustaqīman, 112/730; a paraphrase of Qur'ān 48:2); Baramki, "Al-bādiya al-sūrīya," no. 61 (an‘im ‘alayhi al-hudā wa-l-nūr).
 E.g., ‘Ushsh, "Jabal Usays," no. 42 (aslih ‘abdaka, 113 / 732); Grohmann, AI, Z48 (ahsana Allāh bihi wa-as‘adahu wa-akramahu, first-second / seventh-eighth century); Sharon, "Rehovoth and Sinai," no. 3 (as'alu al-‘afw wa-l-‘āfiya fī 1-dunyā wa-l-ākhira lī wa-li-ahl baytī, second / eighth century); Moraekhi, Medina, H3 (ij‘alhu min al-sālihīn); Rashid, Medina, no. 10 (yas'alu Allāh al-shahāda wa-l-sa‘āda); Baramki, "Kufic Texts," no. 4 (al-kifāya al-hasana wa-1-‘āfiya).
 Miles, "Tā'if," pp. 237-41 = Grohmann, AI, Z68.
 RCEA, no. 8 (citing Ahmad ibn ‘Alī al-Maqrīzī). Though this inscription is known only from a literary attestation, the lack of any religious content (the date is not even said to be according to the Hijra) makes it likely that it has been transmitted accurately.
 Imbert, Jordanie, no. 56.
 E.g., Mittwoch, "Bauinschrift" (Grand Mosque of San‘a, 136/753).
 E.g., on a lead weight dated 128 / 745 by Nitzan Amitai-Preiss, "An Umayyad Lead Seal," al-Qantara 18 (1997): 234; it states that it was commissioned fī khilāfat ‘Abd Allāh Marwān amīr al-mu'minīn.
 RCEA, nos. 27 (Qasr al-Hayr al-Gharbī, 109 / 727 - not Qasr al-Milh as stated in RCEA), 42 (Ascalon, 155 / 772, for which see now Sharon, CIAP1, p. 144); Sauvaget, "Bosra," no. 2 (128 / 745).
 Moraekhi, Medina, L1, L13; ‘Ushsh, "Jabal Usays," nos. 58, 87 (dated 119 / 737); cf. Baramki, "Al-bādiya al-sūrīya," no. 33.
 Muaikel, Jawf, nos. 26-29; Grohmann, AI Z38, Z50, Z130; Baramki, "Al-bādiya al-sūrīya," no. 19.
 The root shahida does not occur in a dated inscription until 157 / 774 (Moraekhi, Medina, W1; cf. ibid., B1, dated 165 / 781, and Nevo, "Prehistory," EKI261 and ST640, dated 164 / 780 and 170 / 786), but is found on graffiti thought by Grohmann to be "possibly 1st century AH" (e.g., Al, Z250, Z287; thus also Rashid, Makka, nos. 3, 23, 27, 29, 56). The earliest dated use of āmana is in an inscription of 83 / 702 from the Hijaz which contains the interesting statement āmantu bi-mā kadhdhaba bihi ashāb al-Hijr (Hamed, Arabie Saoudite, III, plate 238).
 A rather elaborate example is found in an epitaph of 190 / 806: "This is what ‘Alī ibn Salama ibn al-‘Alā' testifies - and God testifies to it too, along with His angels, the bearers of His throne, His prophets, His messengers and all of His creation, and God suffices as a witness - he testifies that ... " (RCEA, no. 77).
 This occurs very often on Egyptian tombstones from 174 / 790 onwards. It also occurs occasionally in graffiti (e.g., Moraekhi, Medina, W2, Hl).
 This doublet most often occurs in citations of the first line of the Throne verse (Qur'ān 2:255): Allāh lā ilāha illā huwa al-hayy al-qayyum; e.g., ‘Ushsh, "Jabal Usays," nos. 16 (dated 93 / 711), 26, 63, 67; Imbert, Jordanie, nos. 3, 9, 58, 71, 88.
 This last phrase is part of Qur'ān 72:3 and is used in an epitaph of 191 / 807 (Hawary-Rached, Stèles, no. 18). Lā sharīka lahu, from 6:163, features in all types of inscriptions, often in some version of the profession lā ilāha illa Allāh wahdahu lā sharīka lahu Muhammad rasūl Allāh (found in papyrus protocols from 86 / 705 - A. Grohmann, "Protokolle," Corpus papyrorum Raineri III [Vienna, 1924], xxxiii-iv; on a cistern in southeast Syria constructed by order of the caliph Hishām, recorded by Rihaoui, "Deux inscriptions arabes," no. 1; on the Grand Mosque of San‘a', dated 136 / 753 and published by Mittwoch, "Bauinschrift"; and on numerous epitaphs from 174 / 790 onwards; cf. the Negev graffito of 164 / 780 in Nevo, "Prehistory," EKI261).
 Both are already cited on the Dome of the Rock (72 / 691) and Qur'ān 112 also on post-reform coinage (from 77 / 696) and papyrus protocols (from 86 / 705). In epitaphs they first appear in 180 / 796 and 190 / 806 onwards respectively (Hawary-Rached, Stèles, nos. 5, 16). Qur'ān 3:18 is also cited on a graffito, probably Umayyad (Ory, "'Ayn al-Garr," no. 55).
 An epitaph of 102 / 720 cites 67:1 (Tawab, Stèles, no. 1); on one of 193 / 808 we find most of 2:255, and on another of 193 / 809 the verses 16:32 and 35:34-35 (Hawary-Rached, Stèles, nos. 21-22). On a graffito, probably Umayyad, 42:53 is cited (Ory, "'Ayn al-Garr," no. 56).
 E.g., Hawary-Rached, Stèles, no. 29 ("In God is a consolation for every disaster and a compensation for every loss"/ fī Allāh ‘azā' min kull musība wa-khalaf min kull hālik, 199 / 815).
 Imbert, Jordanie, no. 105 (Shubayka, 154 / 771); Baramki, "Al-bādiya al-sūrīya," nos. 5, 12, 39, and Nevo, Negev, BR5116, just have al-nabī al-ummī, as does an epitaph of 195 / 811 (RCEA, no. 89). On this term see Isaiah Goldfeld, "The Illiterate Prophet (nabī ummī): An Inquiry into the Development of a Dogma," Der Islam 57 (1980): 58-67. 1
 Imbert, Jordanie, no. 24 (on a wall of Qasr Kharrana); Baramki, "Al-bādiya al-sūrīya," nos. 5, 39, 44, 47.
 On an epitaph of 183 / 799 (Hawary-Rached, Stèles, no. 8).
 This first appears on the reform coinage from 77 / 696 onwards, and is thereafter a common formula in graffiti and epitaphs.
 The latter three appear on epitaphs of 192 / 808, 199 / 814 and 199 / 815 respectively (Hawary-Rached, Stèles, nos. 20, 28-29).
 These two occur on epitaphs of 195 / 811 and 199 / 814, respectively, (RCEA, nos. 89, 98); the final clause of the first = Qur'ān 15:99.
 Imbert, Jordanie, no. 105 (Shubayka, 154 / 771).
 Hawary-Rached, Stèles,
no. 2; cf. Donner, "Al-Hanākiyya,
Saudi Arabia," W3 (inna l-sā‘a lā rayba fīhā,
from Qur'ān 18:21).
 E.g., Moraekhi, Medina, L14-15.
 Most recently see Suliman Basher, "Apocalyptic and Other Materials on Early Muslim-Byzantine Wars," JRAS ser. iii, 1 (1991): 173-207; Michael Cook, "The Heraclian Dynasty in Early Muslim Eschatology," al-Qantara 13 (1992): 3-23; idem, "Eschatology and the Dating of Traditions," Princeton Papers in Near Eastern Studies 1992: 23-47; idem, "An Early Islamic Apocalyptic Chronicle," JNES 52 (1993): 25-29; Lawrence I. Conrad, "Portents of the Hour," forthcoming in Der Islam.
 On two inscriptions of Hakīm b. ‘Amr,
one dated 85 / 704 (Nevo, Negev, MA450,
 This expression appears in epitaphs from Egypt from 179 / 795 onwards.
 The latter two appear in an epitaph of 181 / 797 (Hawary-Rached, Stèles, no. 6). The "meeting with Him" also appears in an epitaph dated either 172 / 789 or 192 / 807 (Tawab, Stèles, no. 2).
 They each receive comment in the early ninth century creed, the Wasīyat Abī Hanīfa, articles 20, 23-24; see the commentary thereon in Wensinck, Muslim Creed, pp. 165-67, 177-80.
 Hawary-Rached, Stèles, no. 3.
 See Wensinck, Muslim Creed, pp. 49-57, 107-109, and most recently J. van Ess, Theologie und Gesellschaft im 2. und 3. Jahrhundert Hidschra, I (Berlin, 1991), pp. 72-135. See also M. Cook, Early Muslim Dogma (Cambridge, 1981), pp. 145-52 (esp. p. 151 where he speaks of 'a widespread determinist mood').
 Muaikel, Jawf, no. 12; this appears inverted in B probably slightly later inscription (ibid., no. 18); cf. Hamed, Arabie Saoudite, III, plate 255 (huwa mu'min bi-mā jā'a min Allāh wa-rasūlihi, 160 / 776).
 RCEA, no. 38, citing Ibn Rustah, who says that he read it himself in 260 / 874.
 RCEA, no. 89, the second sentence ("Falsehood ... ") = Qur'ān 41:42.
 F. Sarre and E. Herzfeld, Archaologische Reise im Euphrat-und Tigris Gebiet, Band II (Berlin, 1920), p. 283 (= RCEA, no. 117), who date it to the late second / eighth century from a comparison with Egyptian tombstones of that period.
 Kessler, "Dome of the Rock," p. 6 (inner face); RCEA, no. 18 (citing Muhammad b. Shākir al-Kutubī); ‘Ushsh, "Jabal Usays," no. 87. Note that in an epitaph of 205 / 820 a distinction seems to be made between Islam and dīn: "he testifies that . . . Islam is what He has sent, that dīn is what He has laid down, that truth is what He has described and that justice is what He has ordered" (RCEA, no. 132).
 See the section on Muhammad and his attributes above.
 Ory, "Ayn al-Garr," no. 1; though not dated, very likely earlier is an inscription from Qusayr Amra seeking forgiveness for "the crown prince of the Muslim men and women" (Imbert, Jordanie, no. 165). Note that "Muslim" as a name occurs earlier; see, for example, C.J. Kraemer, Excavations at Nessana, Volume 3: Non-Literary Papyri (Princeton, 1958), no. 58, and Grohmann, AI, Z36, Z167, Z194, Z204.
 This first appears on a tombstone of 71 / 691 (H. Hawary, "The Second Oldest Islamic Monument Known," JRAS : 290-91), which should possibly be read 171/788, since the script is elaborate and this phrase is not otherwise seen until 190 / 806 (Hawary-Rached, Stèles, no. 16).
 E.g., Fahmi, "Makka," nos. 1 (= 4:87; dated 80 / 699), 2 (ya Dāwūd inna ja‘alnāka khalīfatan fī l-ard ... = 38:26, though note the variant li-tahkuma rather than fahkum; dated 80 / 699); Rashid, Makka, nos. 2 (2:21 with variants from 4:1 and 2:189 / 3:130 / 3:200 / 5:100; dated 84 / 703) and 62 (65:3; 98 / 716); ‘Ushsh, "Jabal Usays," nos. 20 (inna rabbakum Allāh alladhī khalaqa al-samāwāt wa-l-ard = part of 7:54) 23 (most of 30:4), 41 (most of 19:30 and beginning of 19:31); Ory, Hawrān, no. 1 (37:61); Rashid, Medina, nos. 5 (most of 27:19), 7 (part of 18:110); Donner, "Al-Hanākiyya, Saudi Arabia," W1 (10:90 and part of 3:67). ‘Ushsh, "Jabal Usays," nos. 36 and 43, and Moraekhi, Medina, L22, begin ighfir li + name, then add the last part of 27:19 (adkhilhu bi-rahmatika fī ‘ibādika al-sālihīn). For further examples see Imbert, Jordanie, pp. 527-31.
 Rashid, Medina, no. 21.
 ‘Ushsh, "Jabal Usays," no. 87 (dated 119 / 737).
 Nevo, Negev, SC301; the term "the prosperous" occurs thirteen times in the Qur'ān, but the request itself does not appear there.
 Taqabbal min occurs three times in the Qur'ān, salāt eighty-three times, sawm once (the root is used fourteen times).
 There are sections on da‘awāt in many major hadīth collections, and there exist digests of them, such as the Al-adhkār al-muntakhaba by Abū Zakariyyā al-Nawawī (cited in the next footnote) and the Kitāb al-adhkār wa-l-da‘awāt by Ghazalī (which is the ninth book of his Ihyā' ‘ulūm al-dīn and which has been translated by K. Nakamura [Tokyo, 1973 and Cambridge, 1990]). Otherwise they are scattered about various literary works; a number of interesting examples are found in the second chapter of Abū ‘Alī al-Tanukhī's Al-faraj ba‘da shidda (note, for instance, the expression yā rabb al-nabīyīn kullihim ajma‘īn, which, along with other examples, reflects the graffiti's preoccupation with other prophets besides Muhammad - ed. ‘A. Shāljī [Beirut, 1978], I, p. 190). See also EI2, s.v., "Du‘ā'."
 Compare Sanduq, "Hafnat al-Abyad," with, for example, Abū Zakariyyā al-Nawawī, Al-adhkār al-muntakhaba (Beirut and Damascus, 1971), p. 156. The graffito contains two small modifications: it begins Allāh w-kbr, presumably reflecting the case ending of Allāhu, and it ends with the addition wa-laylan tawīlan, drawing on Qur'ān 76:25-26.
 E.g., Hawary-Rached, Stèles, no. 4 (dated 180 / 796); Bukhārī, Sahīh. (Būlāq, 1894), IV, p. 66 (Da‘awāt 37).
 Musil, "Arabia Petraea," no. 1, publishes a graffito of an Umayyad official containing this request and cites Tabarī's Ta'rīkh al-rusul wa-l-mulūk (ed. M. de Goeje [Leiden, 1879-1901], II, p. 353) for its use by Husayn b. ‘Alī before his death in 61 / 680.
 See n. 15 above for examples.
 This phrase is most common in epitaphs; e.g., Hawary-Rached, Stèles, nos. 10 (rahima Allāh man qara'a wa-da‘a li-sāhib hādha l-qabr bi-l-maghfira), 14 (rahima Allāh man qara'a hādha wa-rahhama ‘alayhi).
 Nevo, Negev, EL200C, GM389. Cf. Bukhārī, Sahīh, IV, 70 (Da‘awāt 65): "the Prophet said: 'When the one reading (al-qāri') says amen, then you should all say amen.' "
 Imbert, Jordanie, no. 105 (Shubayka, 154 / 771). Note that this text resembles the du‘ā' said after the final tashahhud of the prayer (see, for example, Nawawī, Adhkār, P. 63)
 See the comments of Nevo, Negev, pp. 4-5, 7-8.
 Pierre-Louis Gatier, Inscriptions de la Jordanie, tome 2: Région centrale (Institut français d'archéologie du Proche-Orient [Paris, 1986]), nos. 6, 97-98, 100, 114, 143; 52, 90, 99; 54c; 46, 54b, 57ac, 90, 107, 113. These inscriptions are of the sixth-seventh centuries. Out of the many collections of Greek inscriptions, I choose this one simply because it has been recently and well edited, but any other collection could profitably have been selected.
 A. van den Branden, Les inscriptions thamoudéennes (Louvain, 1950), p. 35 (fifty-three examples listed, others are found in the collection); see also EI2, s.v., "Thamudic."
 Examples and references given in EI2, s.v., "Safaitic."
 See J.T. Milik, "Nouvelles inscriptions sémitiques et grecques du pays de Moab," Liber Annuus 9 (1958-59): 351-52.
 For example, though one could find a few words and phrases common to both Nabatean and Arabic inscriptions, it is evident from even a brief perusal of a number of catalogues that there is no real point of contact between the two. None of the most frequent Nabatean expressions (e.g., "remembered be" / dkyr) features in Arabic inscriptions (even though dkyr is an important root in Muslim Arabic religious vocabulary), and those that do (e.g., calls for "peace" / shlm) are too commonplace to be significant.
 Epitaphs from outside of Egypt are now beginning to be collected and studied; most recently see Giovanni Oman, La necropoli islamica di Dahlak Kebir (Napoli, 1976); Khaled Moaz and Solange Ory, Inscriptions arabes de Damas, les stéles funéraires. 1. Cimetiere d'al-Bāb al-Saghīr (Damascus, 1977); Chérine Gebara, Les inscriptions funéraires arabes de la ville de Der'a en Syrie (Ph.D. dissertation; Aixen-Provence, 1980); Hasan al-Basha, "Ahammiyyat shawāhid al-qubūr ka-masdar li-ta'rīkh al-jazīra al-‘arabiyya fl l-‘asr al-islāmī (ma‘a nashr majmū‘at al-shawāhid bi-l-mathaf al-atharī bi-kulliyyat al-ādāb jāmi‘at al-Riyād)" in A.M. Abdalla and S. al-Zakkar, eds., Studies in the History of Arabia I 1 (Riyād, 1979), pp. 81-126; Ahmed bin Umar al-Zayla‘i, Tombstones in Dar al-Athar al-lslamiyyah Kuwait (Kuwait, 1989). The texts in these collections, however, fall outside our period of study, being mostly of the third sixth /ninth-twelfth centuries.
 For some comments on the language of inscriptions see Imbert, Jordanie, pp. 549-59.
 For further information consult the studies listed in n. 2 above.
 Sanduq, "Hafnat al-Abyad"; Abbott, "Kasr Kharāna"; Cantineau, Palmyre, no. 39; Nevo, Negev, SC305; Musil, "Arabia Petraea," no. 1 (the owner of the inscription is known to have corresponded with Sulaymān I; Muaikel, Jawf, no. 1; Hawary-Rached, Stèles, no. 10. Nevo, Negev, pp. 95-96, lists twenty occurrences. This invocation is said to have been used by the Prophet himself (Bukhārī, Sahīh, IV, 69-70 [Da‘awāt 62]) and is taken from the second verse of the Victory Surah (48), which was allegedly his favourite surah, since in it God absolved him of all his sins.
 The Arabic name and / or its Greek rendering magaritai, plus the new era, appear in numerous papyri of Egypt from 642 onwards, the earliest firmly dated example is PERF 558 (A. Grohmann, "Aperçu de papyrologie arabe," Études de papyrologie 1932, pp. 41-43), which gives Pharmouthi indiction 1 and Jumādā I "of the year 22" (= April 643). The earliest literary attestation of the name is in Syriac (rendered mhaggrê) from a letter of the catholicos Isho‘yahb III (Liber epistularum, ed. R. Duval, CSCO 11 [Paris, 1904], p. 97), written in Iraq ca. 640s; and the dating first appears in the preamble to a Nestorian synod which took place "in the year 57 of the rule of the Arabs" (J.B. Chabot, ed. Synodicon orientale ou recueil de synodes nestoriens [Paris, 1902], p. 216).
 A number of Christian and Muslim reports suggest that the caliph ‘Abd al-Mālik (685-705) and his governor Hajjāj b. Yūsuf sponsored a revised edition of the Qur'ān, which was then dispatched to the provincial capitals (references given in my Seeing Islam as Others Saw it [Princeton, 1997], pp. 500-501). Perhaps this made the Qur'ān more widely available; this would then explain its frequent citation in graffiti after the 70s / 690s.
 Tabarī, Ta'rīkh, 11, pp. 1272-73 (slightly free translation). For a longer exposition of this theme see al-Ya‘qūbī's Mushakalat al-nās li-zamānihim (translated by W. Millward, "The Adaptation of Men to their Times," JAOS 84 : 333-44).
 The scribe is usually only referred to as "the one who wrote it/this writing," but may occasionally be named; e.g., ‘Ushsh, "Jabal Usays," no. 31: Allahumma aslih Jabal ibn al-Walīd wa-anbithu nabātan hasanan wa-kataba Salmān ibn Nāfī‘ (the phrase "give him good growth" comes from Qur'ān 3:37, indicating that Jabal is a young boy and Salman perhaps his tutor).
 See the studies of ‘Ushsh, "Jabal Usays," and Rashid, Medina.
 Cf. G. Rex Smith, "Studies on the Tihāmah" (review article), JRAS (1986): 32: "The presence of firmly pre-lslamic names still in use in the second and third century of Islam is proof positive of the slow seepage, rather than the sweeping flood, of Islam into the Yemen"; Sharon, "Sede Boqer (168)," p. xxxli, n. 3: "The character of the names that appear in these inscriptions is clearly Jahilite. Most of them are ancient Arabic names on which Islam had not yet made an impression (Jusham, ‘Idāh, Hamdān, etc.)." However, Simon Hopkins informs me that pre-Islamic names are still common among the bedouin of Jordan today.
 Names from biographical dictionaries have been used by Richard Bulliet to this end (see his Conversion to Islam in the Medieval Period [Cambridge, Mass., 1979]).
 Cf. Sharon, Har Nafha, 11*: "From the inscriptions, as well as from the abundant archaeological remains ..., it is clear that the semi-arid Negev supported a considerable population"; Enno Littmann, Semitic Inscriptions (New York and London, 1905), p. 195: "There are a great number of Safaitic inscriptions at il-Hifneh in the Harrah. But not so much writing was shown here during Mohammedan times, doubtless because travel and commerce between the Haurān and the Ruhbeh had decreased."
 Abū l-Faraj al-Isfahānī, Kitāb
adab al-ghurabā', S. al-Munajjid, ed. (Beirut, 1972), p.
23; compare the words of Abū l-Faraj himself (ibid., p. 21): "This
(writing on walls) has become a custom of travellers in every country and destination,
a hallmark (‘alāma) amongst them in every meeting place and visiting
site." This work is very rich in anecdotes about graffiti, but treats only
those of a poetic nature, which are rare among extant graffiti; see the forthcoming
translation and analysis by Patricia Crone and Shmuel Moreh.
 See, for example, Rashid, Darb Zubaydah, ch. 8; Thenayian, Yemeni Highland Pilgrimage Route, ch. 6; Hamed, Arabie Saoudite.
 Making invocations in wadis and high places is approved of in hadīth (e.g. Bukhārī, Sahīh, IV, 68 [Da‘awāt 52-53]). Cf. Miles, "Tā'if," p. 242: "I doubt that these Arabs (who inscribed these graffiti) were other than travelers obeying the universal and perennial impulse to record their names in remarkable places."
 For example, at Sede Boqer Hakīm b. ‘Amr and Khālid b. Humrān own seven and twenty graffiti respectively, while in the vicinity of Nessana Bishr b. Tamīm owns ten graffiti (Nevo, Negev, pp. 117-18). Muaikel, Jawf, nos. 14-24, belong to one family; Moraekhi, Medina, pp. 78, 87, 118, 159, 179, 197, 289, is even able to draw up genealogical charts from the graffiti he publishes.
 Especially as some were inscribed in places difficult of access; two lengthy graffiti, very likely by descendants of the Prophet, are written in clear large letters very high up, "like a landmark to the site" (Moraekhi, Medina, H1-2).
 Cf. R. Ettinghausen, "Arabic Epigraphy: Communication or Symbolic Affirmation?" in D.K. Kouymjian, ed., Near Eastern Numismatics: Studies in Honor of George C. Miles (Beirut, 1974), pp. 279-317; S. Blair, "Legibility versus Decoration in Islamic Epigraphy: The Case of Interlacing" in I. Lavin, ed., World Art: Themes of Unity in Diversity: Acts of the XXVI International Congress of the History of Art (University Park, Pennsylvania, 1989), pp. 329-31.
 For some examples and discussion of such drawings, see A. van den Branden, Les textes thamoudéens de Philby (Louvain, 1956), I, pp. 6-16, and II, pp. xxiii-xxviii; Hans Rhoterd, Transjordanien: vorgeschichtliche Forschungen (Stuttgart, 1938), pp. 159-212; F.V. Winnett and G. Lankester Harding, Inscriptions from Fifty Safaitic Cairns (Toronto, 1978), pp. 22-27; M.C.A. Macdonald, "Hunting, Fighting and Raiding: the Horse in Preislamic Arabia," in D. Alexander, ed., Furūsiyya (Riyād, forthcoming), pp. 72-83.
 Muaikel, Jawf, no. 4.
 Though many of the sketches are highly visible on well-travelled routes, I have not seen any defaced, presumably they escaped Islam's antipathy towards images because they do not occur in a religious context.
 Ory, "Aspects religieux," p. 36: "Ces textes témoignent d'une diffusion certaine de la pensée islamique dans les terres conquises et d'une pénétration des dogmes principaux de la foi musulmane ..., mais cette islamisation, telle que la révèlent ces inscriptions, ne semble pas dépasser ces vérités premières."
 Sharon, "Sede Boqer (167),"
pp. 31-35* (cf. idem, "The Umayyads as ahl
al-Bayt," JSAI 14 ); Nevo,
 Sharon, Har Nafha, p. 12*; Nevo, "Prehistory," p. 110 (cf. idem and J. Koren, "Methodological Approaches to Islamic Studies," Der Islam 68 : 103-104). See also Fred M. Donner, "From Believers to Muslims: Confessional Self-ldentity in the Early Islamic Community," in L.I. Conrad, ed., The Byzantine and Early Islamic Near East IV: Patterns of Communal Identity (Princeton, forthcoming). The introduction of Muhammad into inscriptions was nevertheless very significant; for discussion see my Seeing Islam, pp. 545-59.
 See n. 38 above for some instances of its use. Note that though the phrases might seem "of a general monotheistic nature" (Nevo and Koren, "Methodological Approaches," p. 104), they are often perceived to be characteristic of one confession by other confessions. In modern-day Lebanon, for example, the phrase alhamdulillah, though seemingly unpartisan, would be seen as typically Muslim by Maronite Christians and, especially on certain occasions like after a meal, would not be used by them (I am grateful to Bernard Haykel for this point). Unfortunately, no attention has been given to Christian Arabic inscriptions (usually recognisable by a cross placed at the beginning and end of the text; e.g., Sharon, "Rehovoth and Sinai," no. 5, "dating no later than the second half of the second / ninth century"), so we do not know if these evolved an expression distinct from their Muslim counterparts.
 Baramki, "Al-bādiya al-sūrīya," nos. 79 (dated 608 / 1211), 86 (628 / 1231), 93 (724 / 1324), 95 (767 / 1365), 96 (779 / 1377), 100-101 (724 / 1324), are all just simple requests for forgiveness, beginning with the same Allahumma ighfir as those graffiti written six centuries or so earlier. Imbert, Jordanie, p. 656, cites an inscription of a Muslim worker on a petrol line, dated 1988, that only contains the basmala.
 Nevo and Koren, "Methodological Approaches," pp. 90-91, citing A. Snodgrass.
 See B. Jobling, "Report of the Eighth Season of the ‘Aqaba-Ma‘an Archaeological and Epigraphic Survey (January-February 1988)," Liber Annuus 39 (1989): 255: "On a rock surface on the hill to the west of the mosque there is another large Kufic inscription which bears the date 107 AH." The site was later visited by Alison McQuitty, Director of the British School at Amman for Archaeology and History, to whom I am much indebted for providing me with a photograph of this inscription.
Abbott, Nabia. "The Kasr Kharāna Inscription of 92 H. (710 AD): A New Reading," Ars Islamica 11-12 (1946): 190-95.
Baramki, Dimitri. "Kufic Texts," Annual of the Department of Antiquities in Jordan 1 (1951): 20-22.
. "Al-nuqūsh al-‘arabiyya fī l-bādiya al-sūrīya," al-Abhāth 17 (1964): 317 46.
Cantineau, J. Inventaire des inscriptions de Palmyre, fascicule IX: le sanctuaire de Bel. Beirut, 1933.
Couroyer, B. "Inscription coufique de Beit Gibrin," Revue biblique 71 (1964): 73-79.
Donner, Fred M. "Some Early Arabic Inscriptions from al-Hanākiyya, Saudi Arabia," JNES 43 (1984): 181-208.
Fahmi, S. ‘A. "Naqshān jadīdān min Makka al-mukarrama," al-Manhal 48 (1407 / 1987): 346 61.
Grohmann, A. (AI=) Arabic Inscriptions. Expédition Philby-Ryckmans-Lippens en Arabie, IIe partie: texrtes épigraphiques, tome 1. Louvain, 1962.
Hawary, Hassan, and Hussein Rached. Catalogue général du musée arabe du Caire. Stèles funéraires, tome premier. Cairo, 1932.
Hamed, ‘Ali Ibrahim. Introduction a l'étude archeologique des deua; routes syriernne et égyptienne de pélérinage au nord-ouest de l'Arabie Saoudite. Ph.D. dissertation; Aix-en-Provence, 1988.
Imbert, Frédéric. "La nécropole islamique de Qastal al-Balqa' en Jordanie," Archéologie islamique 3 (1992): 17-59.
. Un corpus des inscriptions arabes de Jordanie du nord. Ph.D. dissertation; Aix-en-Provence, 1996.
Kessler, Christel. "'‘Abd al-Mālik's Inscription in the Dome of the Rock: A Reconsideration," JRAS (1970): 2-14.
Miles, George C. "Early Islamic Inscriptions near Tā'if in the Hijāz," JNES 7 (1948): 236 42.
Mittwoch, Eugen. "Eine arabische Bauinschrift aus dem Jahre 136 H," Orientalia 4 (1935): 235-38.
Moraekhi, Moshalleh K. al-. A Critical and Analytical Study of some Early Islamic Inscriptions from Medina in the Hijāz, Saudi Arabia. Ph.D. dissertation; Manchester, 1995.
Muaikel, Khaleel Ibrahim. A Critical Study of the Archaeology of the Jawf Region of Saudi Arabia. Ph.D. dissertation; Durham, 1988.
Musil, Alois. "Zwei arabische Inschriften aus Arabia Petraea," WZKM 22 (1908): 81-85.
Nevo, Yehuda, Zemira Cohen and Dalia Heftman. Ancient Arabic Inscriptions from the Negev, Volume 1. Midreshet Ben-Gurion, Negev, 1993.
Ory, Solange. "Les graffiti umayyades de ‘Ayn al-Garr," Bulletin du musée de Beyrouth 20 (1967): 97-148.
. Cimetières et inscriptions du Hawrān et du Gabal al-Durūz. Paris, 1989.
Rashid, Sa‘d b. ‘Abd al-‘Azīz al-. Darb Zubaydah: the Pilgrim Road from Kufa to Mecca. Riyād, 1980.
. Kitābāt islāmiyya ghayr manshūra min ‘Ruwāwa' al-Madīna al-munawwara. Riyād, 1993.
. Kitābāt islāmiyya Min Makka al-mukarrama. Riyād, 1995.
RCEA = Et. Combe, J. Sauvaget and G. Wiet, Répertoire chronologique d'épigraphie arabe, I. Cairo, 1931.
Rihaoui, Abdul Kader. "Découverte de deux inscriptions arabes," Annales Archéologiques de Syrie 11-12 (1961~2): 207-11.
Ruggieri, Vincenzo. "I graffiti umayyadi a Cnidos," Orientalia Christiana Periodica 58 (1992): 549-51.
Sanduq, ‘Izz al-Dīn al-. "Hajar Hafnat al-Abyad," Sumer 11 (1955): 213-17.
Sauvaget, J. "Les inscriptions arabes de la mosquée de Bosra," Syria 22 (1941): 53-65.
. "Glanes épigraphiques," REI 15-16 (1941-48): 17 29.
Sharafaddin, A.H. "Some Islamic Inscriptions Discovered on the Darb Zubayda," Atlāl 1 (1977): 69-70 (and plates 49-50).
Sharon, Moshe. "Arabic Inscriptions from Sede Boqer," Appendix to Rudolf Cohen, Archaeological Survey of Israel. Map of Sede Boqer East (168) 13-03. Jerusalem, 1981, pp. xxxi-xxxii, 87-88.
. "Arabic Inscriptions from Sede Boqer: Map 167," Appendix to Rudolf Cohen, Archaeological Survey of Israel. Map of Sede Boqer West (167) 12-03. Jerusalem, 1985, pp. 31-35* (introduction), 94-102 (transcription and translation), 103-106 (photographs).
. "Arabic Rock Inscriptions from the Negev," in Israel Antiquities Authority, Archaeological Survey of Israel. Ancient Rock Inscriptions. Supplement to Map of Har Nafha (196) 12-01. Jerusalem, 1990, pp. 9-35* (English), 9-36 (Hebrew), 47-59 (photographs).
. "Five Arabic Inscriptions from Rehovoth and Sinai," Israel Exploration Journal 43 (1993): 50-59.
. CIAP 1 = Corpus inscriptionum arabicarum Palaestinae, vol. 1 -A-. Leiden, 1997.
Tawab, ‘Abd al-Rahman M. ‘Abd al-. Stèles islamiques de la Nécropole d'Assouan I. Cairo, 1977.
Thenayian, Mohammed A. Rashed al-. An Archaeological Study of the Yemeni Highland Pilgrim Route between San‘a and Mecca. Ph.D. dissertation; Durham, 1993.
‘Ushsh, Muhammad Abū l-Faraj al-. "Kitābāt ‘arabiyya ghayr manshūra fī jabal Usays," al-Abhāth 17 (1964): 227-316.
Back To The Arabic & Islamic Inscriptions