Evidence that the term ‘Allah’ originated with Jewish and Christian Arabs
Rick Brown, ((The author acknowledges with gratitude the helpful feedback received from a great many reviewers.)) 19 October 2007
In talking with Muslims, it is essential to understand and affirm their names for God. In most languages spoken by Muslims, the term allâh is at least one of their names for God. Dudley Woodberry (1996a: 173) has pointed out that the name allâh “is of Christian Syriac origin and was in use long before Muhammad’s time.” Syriac-speaking Christians have always believed this, and scholars like Arthur Jeffrey (1938: 66) have noted this as well. But violent acts perpetrated by some militant Islamists in the name of allâh have led many people in the West to conclude that allâh must be someone besides God. ((See five articles in the Christian Century 2004, all entitled “Do Christians and Muslims Worship the Same God?” by Levenson, Sanneh, Woodberry, Ariarajah, and Abd-Allah.))
For example, I was recently at an academic conference where one of the speakers was noting that each of the languages of Africa has an indigenous name for the Supreme Being, the lord and creator of the universe, and that this local name is used by the Christians in their worship and in their translations of the Bible. ((In point of fact, as Lamin Sanneh (1989: 181) points out, long-time Islamized language communities like Hausa and Fulani have “allowed Allāh to displace the god or gods of pre-Islamic times,” with the result that some groups no longer even remember the name by which their ancestors invoked the Most High God.)) Suddenly, however, he was struck with some doubt, so he qualified his remark by saying, “Well, at least everyone south of the Sahara has a name for God.” He was uncertain whether the peoples of northern Africa had a name for God! This doubt stemmed from claims he had read that allâh, the Arabic word for God, does not refer to the Lord and Creator of the universe but to some demon or idol, such as the ancient Semitic moon god sīn. These claims are being made by a number of authors who have written that the term allâh denotes a pagan deity, and in particular a moon god. Their well-meaning but poorly substantiated claims have left many Western Christians fearful of the term allâh and opposed to its use. ((Opposition to use of this term in translations of the Bible is discussed and addressed in Kenneth J. Thomas (Thomas 2006a).)) Some Western Christians have even removed the term allâh from translations of the Arabic Bible and from other materials. ((An example is www.Arabbible.com, which has published the Van Dyck Arabic Bible online with all mentions of allâh removed.))
Dudley Woodberry, however, has long warned us about the dangers of such rejectionism. He entitled one of his articles “When Failure is Our Teacher: Lessons from Mission to Muslims,” and he made this observation (1996b: 122):
Many missionaries branded so-called Muslim forms of worship and religious vocabulary as wrong, without knowing that virtually all quranic religious vocabulary, including the name “Allah,” and virtually all the forms of worship, except those specifically related to Muhammad, were used by Jews and/or Christians before they were used by Muslims.
Phil Parshall (1989) makes a similar point. But when Muslims encounter Christian religious materials that have carefully avoided all mention of the name allâh, they often fear the materials are intended to lead them away from God. And if Western Christians “explain” to their Muslim friends that Muslims use the name allâh to invoke a demon or moon god, then they lose all credibility. And besides these fears and follies, there is the simple fact that if we are speaking to people in their own language yet reject the names they use to refer to God and the prophets, then we convey rejection of them personally. Such insults often prompt their rejection of our testimony before they have even considered it. ((Rejection of the audience’s language is just one part of a wider phenomenon of cultural denigration that characterizes polemical approaches. Unfortunately such approaches tend to antagonize people and to harden them in their positions rather than to open them up to the love of God in Christ (Sharkey 2004).)) Consequently, those who believe these myths regarding the term allâh are doomed to failure as witnesses to Muslims. Of course, missionaries who have lived closely with Muslims understand that the name allâh is simply one of the many names for God used by Muslims, but some of them encounter opposition to its use from people in their supporting churches or in their home offices, people who have misconceptions about the term.
Christians who are unaccustomed to religious diversity are often confused by the fact that different monotheistic religions teach different conceptualizations of God, and some Christians even suppose that adherents of different religions are referring to different gods, as if there were a pantheon to choose from. In the technical language of semantics, these people are confusing different “senses” (or “conceptions”) of a term with different “referents”. The referent is the person or entity to which one is referring, who in this case is God. The sense encompasses the characteristics that are attributed to God in their conception of Him. People can have different conceptions of the same referent. Even Christians differ among themselves in their conception of God. ((See (Baylor Institute for Studies of Religion 2006). For a description of the main features of the Biblical concept of God compared to features of Muslim concepts of God, see Brown (2006a).)) A person’s concept of God can change, but this does not happen simply by calling God a different name; it happens by grace when a person ponders the characteristics of God as He is presented in the Bible, and especially as He is revealed in the person of Jesus Christ. It happens when people hear the testimonies of believers, when they experience God’s grace in their lives, when they apprehend God in their inner life, and when they receive illumination from the Holy Spirit.
In (Brown 2006b) I argued that allâh was never the name of a moon god, and that the crescent symbol used in modern Islam does not come from an ancient moon-god religion but from a medieval symbol of Ottoman political domination. Kenneth Thomas (2006b) followed up with an article showing that Arabic-speaking Jews, Christians, and Muslims have always referred to the one true God as allâh, while Massey (2003) and Cox (2006) have emphasized that Arab Christians call God allâh, and that the term is related linguistically to Hebrew terms for God. Imad Shehadeh (2004), director of an Arab Christian seminary, notes the oldest extant Arab Christian translations of Scripture use allâh, and that this practice is documented from ancient times until the present. This fact is well exemplified in the essays in David Thomas (2006a), especially (Kachouh 2006). Shehadeh notes the total lack of evidence that anyone ever used the term allâh as the name of a moon god. Quoting Montgomery Watt, he says the claim that “Christians worship God and Muslims worship Allah” is as sensible as saying “Englishmen worship God and Frenchmen worship Dieu”. He goes on to say that “Muslims and Christians…believe in the same God as subject [but] the nature of God as conceived by Islam is not at all identical to the nature of God within the Judeo-Christian faith” (p. 26) The need, then, is for Muslims to encounter the nature of allâh as presented in the Bible.
These articles, however, have not assuaged the concerns of some who think that the term allâh has its origin as an Islamic invention or as a pre-Islamic demon or idol, and some people remain worried by the apparent similarity of the name allâh with that of the pagan goddess allāt. ((Except for geographical names and the word ałłâh, Arabic words have been transliterated in accord with DIN31635, which is identical to ISO 233 except for the long vowels. The exception in ałłâh is that the velarized el sound, which occurs uniquely in this word in Arabic, is represented on occasion with the “dark el” symbol ‘ł’, and the velarized (low-back) vowel sound that follows it is represented with the symbol ‘â’. The symbol ‘ā’ is used in accord with convention to represent the normal long /a:/ vowel in Arabic, as in ’ilāh “god”. Thus the name of the supposed goddess is transliterated as al-lāt, which rhymes with ‘cat’, whereas the name of God is transliterated as al-łâh, which rhymes with ‘law’. Transliterations of Hebrew, Aramaic, and Syriac consonants are in accord with ISO 259, which is followed by the Society of Biblical Literature. The seven vowel qualities represented in the Tiberian system, [i], [e], [ɛ], [a], [ɔ], [o], and [u], have been represented here as ‘i’ ‘ê’ ‘e’ ‘a’ ‘â’ ‘o’ ‘u’, and similarly for Syriac. Thus a low-back vowel sound has been represented in the same way in all of these languages, i.e., as ‘â’ rather than as ‘ā’, to maintain uniformity of representation for this sound.)) So following Luke’s example, it seemed good to me to investigate these things carefully, and to present in this article detailed evidence to support what Dudley Woodberry wrote, namely that allâh was the term used by Arab Christians for the God of the Bible before the rise of Islam and that it has its origin in the Aramaic term for God, which Jesus Himself would have used. In that sense the term allâh is freer of pagan history than is the Hebrew word ’el, which was used by the Canaanites as the name of the chief deity of their pantheon, ((See (Scott 1980). He notes that the word ’ēl is used across the Semitic languages both as a generic term for a god and as an epithet for the most high God. The latter meaning is sometimes made explicit in Hebrew by use of the phrase ’ēl ‘elyôn.)) or the English word ‘God’, which comes from a generic term for middle-rank Teutonic deities. ((See the entry for gheu(ə)– in (Watkins 2000). Some historical linguists think the English word ‘god’ originates in the name of a Teutonic king named ‘Gaut’ who was deified after his death.)) I will argue that Christianity pervaded all parts of Arabia prior to the rise of Islam, that most Christian Arabs used Aramaic Scripture and liturgy in which God was called alâh(â), that they borrowed this term into Arabic as allâh, and that even non-Christian Arabs identified allâh as the God of the Bible, the supreme being, who is creator and lord of all and above any other gods. I argue that in languages like Arabic where allâh is the normal term for God, its avoidance by Western Christians is unjustified. Similarly there is no reason to avoid calling our Lord Jesus Christ by his well-known Arabic epithet, kalimat allâh, the eternal “Word of God”, incarnate as a man, the visible image of the invisible God and the Lord and Savior of humankind.
Pre-Islamic Arab Christians Referred to God as allâh
In what follows I show first of all that Arabic-speaking Jews and Christians lived throughout Arabia for centuries before Islam. Therefore they would have had a term for referring to God. I then note the existence of pre-Islamic Christian names that incorporated the term allâh. I also show that ancient Arabic Bible translations and the Qur’an itself reflect pre-Islamic Jewish and Christian usage of allâh to refer to God. The conclusion is that pre-Islamic Jews and Christians referred to God as allâh.
Arabic-speaking Christians lived throughout Arabia for centuries before Islam
Although Muslim historians emphasize the paganism and depravity of pre-Islamic Arabia, this is actually an overstatement. It serves to exaggerate the religious transformation effected by the prophet of Islam on Arab culture. A more accurate description is that Judaism had been in Arabia from ancient times, with several tribes having converted, and this had been followed by a wave of conversions that made Christianity the dominant religion in most of Arabia.
Prior to the rise of Islam, there were Jewish tribes in Arabia. The town of Yathrib (later called “Medina”) had long been settled and dominated by Jews (Winder 1999). In the south of the Arabian Peninsula, the populations of Najran and Yemen included large numbers of Jews and proselytes (Shahid 1971). The witnesses of Pentecost included Arabic-speaking Jews and proselytes (Acts 2:11), and they would have taken the Gospel back to their homelands. Paul made a trip to Arabia as well (Gal 1:17), probably the kingdom of Nabataea, meaning “the peoples of the towns and villages that existed throughout the whole region east of a line from Aleppo to the Dead Sea” and including Sinai (Trimingham 1979: 72). So Judaism was present in Arabia before Christ was born, and the Gospel entered Arabia soon after His resurrection.
The church quickly grew. Origen, the third-century theologian and commentator, gave theological lectures in Petra in 213 or 214 at the invitation of the governor (Preuschen 1953 : 268–273). Origen returned again to “Arabia” to correct Beryllus, bishop of Bostra, (((Eusebius: Hist. Eccl. 6:33). According to Eusebius, Beryllus had been teaching that Jesus was not preexistent. This was also a doctrine of the Ebionites, a Jewish Christian sect that might have influenced some of the Arab Christians.)) and returned again in 246 to settle theological disputes in the Arab church synod, which was “of no small dimensions”. (((Eusebius: Hist. Eccl. 6:37). According to Eusebius, Origen sought to correct an unorthodox doctrine that had developed in Arabia, namely that the soul died with the body and was restored with it at the resurrection. It might be noted that a form of this doctrine survives in Islam, as does the view of Beryllus regarding the mere humanity of Jesus Christ.)) In the introduction to his Hexapla edition of the Old Testament, Origen wrote that he consulted Bible translations in several languages, including Arabic (Beeston 1983: 22). This suggests that at least portions of the Old Testament had been translated into Nabataean Arabic by the third century, presumably using Nabataean script, although it is possible that it was a translation into Nabataean Aramaic. In 244 AD an Arab Christian, Philip the Arab, became emperor of Rome, indicating the degree to which Arab Christians were involved in the Roman Empire. (((Eusebius: Hist. Eccl. 6:34). This would make Philip the first Christian emperor. It is also recorded that Origen corresponded with him. Philip stopped the persecution of Christians, but he did not give favoured status to Christianity and he maintained certain imperial Roman religious traditions (Grant 1985: 155).)) Their status in the church is indicated by the presence of Arab bishops at the Council of Nicaea in 325 and at the later councils as well. ((See ‘Arabia’ in (The Catholic Encyclopedia 1913: 668), where it is reported that six bishops participated in the Council of Nicaea from the Roman province of “Arabia.” There were also bishops from Mesopotamia, which was outside of the Roman Empire, and which would have included the Christian Arab Lakhmid kingdom.))
By the early fourth century, northern Arabia and the Arabian Gulf were ruled by the Christian Arab King Imrul Qays (288–328 AD), whose capital was the town of Hira in Mesopotamia and who ventured as far south as Najran. His Lakhmid dynasty of Christian Arab kings continued until 602, when their kingdom was destroyed by the Persians. According to Bellamy (1990), it was this Christian Lakhmid kingdom that fostered the development of the Arabic alphabet and the writing of Arabic poetry, some of which survives. He notes that according to Arab traditions, three Christian Arabs, Muramir, Aslam, and ‘Amir, developed the Arabic alphabet from the Syriac alphabet and taught it to the people of the Lakhmid kingdom. It is said that the alphabet was brought to Mecca by Bishr ibn ‘Abd al-Malik. Prior to this the Meccans and South Arabians had used the Musnad alphabet, but it was very different from the Syriac script to which people in the rest of Arabia had become accustomed.
As for north-western Arabia (modern-day Syria and Jordan), it was ruled by the Arab Nabataean kingdom. In 106 it was annexed to the Roman Empire and became the province of “Arabia”. Then from 363 this whole region was ruled by a succession of Orthodox Christian Arab Monarchs who were outside the empire but were federated with it. Māwīya, Queen of the Saracens, ruled 363–378 AD, and she lobbied successfully for the appointment of Moses of Sinai as bishop of the Saracens (Langfeldt 1994: 53). ((The word ‘Saracen’ is from the Arabic word šarqiyīn “easterners”. See the extensive footnotes in Langfeldt’s article for his sources.)) Moses was famous for the miracles that attended his ministry/ He evangelized the bedouin and was later recognized as a saint. Māwīya was eventually succeeded by King Zokomos (Dhuj‘um), who converted to Christianity in response to an answered prayer. Zokomos began a dynasty of Christian Arab kings (Shahid 1989: 3–8), with the result that according to Langfeldt (1994: 53), “The indigenization of Christianity among Arabian tribes proceeded rapidly from the late fourth and early fifth centuries.” By the sixth century, the Christian Arab Ghassanid kingdom covered most of Syria, Palestine, and Jordan, and extended south almost to Yathrib (Medina). It competed with the Christian Arab Lakhmid kingdom in Mesopotamia and the Gulf.
As for the people in southern Arabia and Yemen, which the Romans called “Arabia Felix,” they had converted to Judaism in the fourth century, but by the sixth century large numbers of them had become Christians. The church building in Najran was so large that their Jewish persecutors were able to force 2000 people inside before burning it down (Shahid 1971; Tardy 1999) and (Brock & Harvey 1998: chap. 4, esp. p. 105). In Sanaa (Yemen) there was an even larger cathedral, built by King Abraha (Guillaume & Ibn Ishaq 2002 : 21), the site of which remains to this day.
Langfeldt provides further detail on the extent of Christianity (1994: 53):
A brief summary of the 4th–7th centuries shows a great many of the tribal groupings in the areas now called Jordan, Syria and Iraq becoming Christian, including the Tanukhids, the Kalb confederation of tribes, the Tamim, the Taghlib, Banu Ayyub, and the majority of the tribes in the Hijaz, Nafud, Najd, Yamama and Bahrain sections of present day Saudi Arabia. A large portion of the Kinda tribe, having left the Yemeni Hadramawt in the 4th C and migrating to the Najd, by the 5th C, had forged alliances with the Ma’add; this “federation” stretched from a point two day’s journey east of Mecca, north and east to include the entire heart of central Arabia. As part of an alliance with the Byzantine Empire in the opening years of the 6th century the Kinda federation adopted Christianity. Many of the Yamama centering in the area of modern Riyadh were Christian (since the middle of the fourth century), as was the great tribal grouping of the Bakr ibn Wa’il in the central and eastern regions.
South west Arabia had a strong Christian enclave in Najran where some 2000 believers were massacred in A.D. 523. There was also a Christian presence in the Hijaz. In the process of hurling invectives at the Umayyad poet Jamil ( ca. 701), a Christian of the ‘Udra tribe, Ja‘far ibn Suraqa testified to Christian monks living in the Wadi al-Quara near Medina. The ‘Udra were Christianized, probably by the 5th century, and maintained that faith well into the Islamic period. There is evidence of Christian monasteries located at strategic locations on the caravan routes and functioning as caravanserai. The writings of al-Muqaddasi, al-Azraqi and other Islamic sources record a) a Christian cemetery (Maqbarat al-Nasara) and Christian stopping place (Mawqif al-Nasrani) in or very near Mecca, and b) the mosques or praying places of Maryam (Masajid Maryam) outside of Mecca on the road to Medina – quite likely a church turned mosque since the Qur’an accepts the Virgin Mary. In the Ka‘ba itself in 630 when Muhammad captured the city, paintings of the Virgin Mary and Jesus occupied positions on the pillars along with Abraham and the prophets.
So as Langfeldt observes, Christianity dominated the Arab religious scene in most of pre-Islamic Arabia and was “the primary religious allegiance of the vast majority of the population”, even after the rise of Islam (Ibid.).
Daniel Potts concludes his two-volume history of The Arabian Gulf in Antiquity with a similar observation (1990: 353):
As we have seen, Christianity was widespread both amongst the tribes of northern Arabia and in the settled communities along the coast.
It is not incorrect to say that, in one sense, the Nestorian Church, for the space of over three centuries, united a region which secular rulers from Sargon to Šapur had never mastered so completely.
So by the time Islam appeared, Christianity was present throughout Arabia, and Christians dominated the major kingdoms into which Arabia was divided: Ghassanid, Lakhmid, Himyarite (Yemen) and Kindite (Southern Arabia). Christians had the weakest presence in the towns that fell outside these kingdoms, notably Mecca and Yathrib (Medina), the very places that gave birth to Islam. Yet the ‘Udra tribe in Mecca was Christian, and in Yathrib (Medina) there were three or more Jewish tribes. Since Christianity was widespread across the various Arab tribes and Judaism was present as well, their name for the God of the Bible, the creator of the universe, would have been well known to all of the Arabs. In what follows we will discover what name they were using for God.
Pre-Islamic Christian names incorporated the term allâh in reference to God
There has been speculation that some of the pre-Islamic Arab churches would have developed an Arabic-language liturgy and lectionary in the fourth or fifth century. Irfan Shahid (1989: 528f.) entertains this as a likelihood. He affirms with confidence, however, that there was pre-Islamic Christian Arabic poetry, as does Kenneth Cragg (1991). Trimingham (1979) lists five of the poets by name. ((Nābigha al-Dhubyānī (died 604 AD), Jarīr ibn ‘Abd al-Masīḥ (died 580 AD), Abu Du‘ād al-Iyādī, Aws ibn Ḥajar, and Maimūn ibn Qais (AKA al-A‘shā, died 625 AD). Trimingham (pp. 177, 201) notes that the poetry which survives does not focus on Christian themes.)) These pre-Islamic Arab Christians would of necessity have had a word for God that they used when speaking Arabic; the poetry that survives, from Nābigha al-Dhubyānī, shows that he used the term allâh.
The hardest pre-Islamic evidence comes in the form of stone inscriptions that bear theophoric Arab names, i.e., Arabic names that incorporate a word for deity. The word one finds most often in the surviving inscriptions is ’lh, pronounced [ałłâh], ((Since doubled consonants were not usually marked in ancient Syriac and Arabic inscriptions, ’lh could be pronounced alâh or allâh. Ryckmans (1934–35) vocalized them as ilāh, evidently under the influence of Wellhausen’s thesis that henotheism developed later in history. Winnett (1938: 247) objects, however, presenting the linguistic evidence allâh:
Against this theory it may be urged that when we meet names like W-h-b-’-l-h and W-h-b-’-l-h-y in Nabataean no one doubts that the theophorous element is Allah. It cannot very well be ilāh, because the Nabataean word for “God” is allāhā which would require a final alif after the ha in the inscriptions. The Greek transliterations of these Semitic names are a further proof that the theophorous element should be read as Allah. If we admit that these names are Allah-names when they appear in Nabataean, on what ground shall we deny them the same interpretation when they appear in Lihyanite or Thamudic?
When Arabic names are found in Greek texts and inscriptions, the letter lambda is doubled. This means the Arabic lām must have been pronounced with doubling, as allâh. For example, the common Arabic name الله وهب ‘wahab allâh’ “Gift of God” is found written as in Ancient Arabic as whb’lh, but in Greek as Ουαβαλλας, showing that the lām was pronounced doubled at an early time. See also (Jaussen & Savignac 1914: 264), cited in (Macdonald 1999b: 275).)) and sometimes the shortened or Hebraic form, ’l. ((There is literary evidence for the pre-Islamic use of the phrase al-’ilāh “the god” to designate the supreme being, but this does not appear in the pre-Islamic epigraphic evidence or in pre-Islamic names. Later one does find the name ‘abd al-’ilāh عبد الاله “servant of the god”, meaning servant of the one who is truly God.)) There is no evidence for a significantly different term for God used in place of this, such as Greek theos or Hebrew adonai or elohîm, although Yhwh is found on occasion, probably as part of a Jewish name. ((Jane Taylor (2001: 168) notes that among the 7,000 Nabataean inscriptions in the Sinai (mostly from the 1st and 2nd centuries AD), none of the theophoric names mention traditional Nabataean deities. There are names incorporating allâh, as well as names with ’el and even names with ba‘al (which might mean the Canaanite god “Baal” but more likely has its normal meaning of “Lord” or “husband”). But she also notes the names šm‘yw, ‘bdyw, and ‘abd’hyw, which appear to “relate to the worship of YHWH”. In addition to Nabataea, Winnett (1952) cites epigraphic evidence in North Arabia for theophoric names that end with -yah (i.e. Yhwh) as well as allâh. One notes that any or all of these names could be Jewish, including those with ba‘al, and so they might belong to Jewish, Christian, or Jewish Christian Arabs.)) Harding’s Index and Concordance of Pre-Islamic Names and Inscriptions includes the following observation (1971: 907): “A feature which emerges very clearly from these lists [of theophoric names] is the overwhelming popularity of ’l, ’lh.” So while many inscriptions bore theophoric names that incorporated the names of pagan deities, there was an “overwhelming” number of theophoric names that incorporated ’lh [ałłâh] and the shortened form ’l. The widespread usage of these terms in the two centuries before Islam correlates with the well-documented spread of Christianity throughout most of Arabia that during that same period (Guillaume & Ibn Ishaq 2002 : 18). ((It would be useful for someone to make a map of pre-Islamic Arab sites, correlating what is known of the distribution of Christians with the epigraphic evidence of theophoric names using allâh. This is complicated, however, by the fact that the documentation on inscriptions is spread over a large corpus, such that the list of relevant books and articles (Kitchen 1994) runs to 821 pages, and the index of names (Harding 1971) runs to 943 pages.))
The Arabs used a number of scripts, but what we now call “Arabic” script was not developed until the fifth or sixth century. The earliest dated Arabic-language inscription in this “Arabic” script is the Zebed inscription. It was inscribed onto a Christian martyrion in 512 AD, where the texts are in Greek, Syriac, and Arabic. ((For dates and literature, see (Gruendler 1993). There is also an inscription on a church in Jabal Ramm which is thought to be from the fourth or fifth century, but it is not dated and does not include a word for God. There are two earlier Arabic inscriptions written in the Nabataean alphabet, namely the ‘En ‘Avdat inscription of the second century AD and the Namarah funerary inscription of the Christian Arab King Imrul Qais of Hira, dated 328 AD, but neither inscription includes a reference to deity. Earlier Arabic inscriptions exist, such as a first century BC inscription in Musnad script at Faw, in southern Arabia, but they do not include the term allâh or any other reference to God. For a catalogue of early Arabic inscriptions and dialects, see (Macdonald 2000).)) The Arabic text includes a name or statement in which God is referred to as alâh or allâh. ((Photographic plates are found in (Grohmann 1971: 6–8). The term allâh is found at the beginning of a list of names of Christian martyrs, but it is not clear if it is part of name, “Help of God”, or is a statement, “By the help of God”. The spelling of allâh is phonetic, with the long second vowel indicated by an alif: ’l’h (Arabic script: الاه ), which is normal for long vowels in Arabic. But elsewhere the term is spelled without marking the long vowel, evidently on analogy with the spelling in Jewish Aramaic and Christian Syriac.)) This shows that pre-Islamic Christians were using this term in reference to God in Arabic, just as they used alâh(â) to refer to God in Syriac.
This archaeological evidence is corroborated by historical sources as well. For example, a leader of the Christians who was martyred in Najran in 523 AD is said to have been ‘Abdullah ibn Abu Bakr ibn Muhammad. Not only does he bear a theophoric name that means “servant of allâh”, he is also said to have worn a ring that said “allâh is my Lord” (Guillaume & Ibn Ishaq 2002 : 18). Similarly when four of the leading pre-Islamic men of Mecca pledged to renounce idolatry, worship God alone, and seek the true religion, it was allâh whom they acknowledged, and three of them found Him in Christianity (Ibid. pp. 99–100). ((According to this account, although Muhammad began his mission soon after this pact and interacted with the four men, only one of them accepted Islam, and afterwards he converted to Christianity. Muhammad added the man’s widow to his harem.)))
Of course, pagans might give their children Christian names, so the existence of theophoric names like ‘Abdullah and Daniel does not imply that their bearers were necessarily named by Jewish or Christian parents. Muhammad’s father, for example, was named ‘Abdullah, and there is no reason to think that the one who named him, his father ‘Abdul Muttalib, was a monotheist rather than a polytheist or henotheist. There is evidence, however, that henotheism had become widespread among the pagan Arabs, i.e., that they acknowledged that the God of the Bible was the Lord and Creator of the universe, while continuing to fear and appease lesser beings instead of God alone. This is reflected in the Qur’an in verses like ‘Ankabūt 29:61, 63, which speaks of pagan Arabs who refused the message of Muhammad:
If indeed thou ask them who has created the heavens and the earth and subjected the sun and the moon (to his Law), they will certainly reply, “Allah”. How are they then deluded away (from the truth)? … And if indeed thou ask them who it is that sends down rain from the sky, and gives life therewith to the earth after its death, they will certainly reply, “Allah!” Say, “Praise be to Allah!” But most of them understand not. (Yusuf Ali translation)
In summary, there is epigraphic evidence that the pre-Islamic Arab Christians were using allâh as the name of God, and there is no evidence that they were avoiding this name and using some other name instead. As Bob Cox (2006) has noted, allâh is the only word in Arabic for God, it is cognate with the Hebrew and Aramaic terms used in the Bible, and it has been used by Arabic speaking Jews and Christians to refer to God for as long as we have records. ((It might be noted that in proto-Hebrew, the word for a deity would also have been ’ilāh or ’elāh, but long ā vowels shifted to /o/ vowels in stressed syllables, resulting in the form ’eloh. This is a common word for “god” in Biblical Hebrew, as in Ps. 18:32 (“Who is god besides YHWH”), although in Job and Proverbs 30 it is used as a name for God. In Aramaic the vowel shift was less pervasive and less pronounced, from a long low-front /ā/ vowel to a low-back /a/ vowel, written here as /â/. In some dialects this was pronounced [ɔ] as in British ‘ought’ and later [o] as in ‘coda’. In Arabic the word for “god” remained ’ilāh. To refer to the one true God in Arabic, the Aramaic word ałâh was used. At an early stage the el sound was doubled, resulting in the word ałłâh, but with either pronunciation the word was written as ’lh, at least until the seventh century.))
Arabic Bible translations reflect pre-Islamic Christian usage of allâh to refer to God
Bruno Violet (1901) published a fragment of Psalm 78 [77 in LXX], discovered in Damascus, in which the Greek text is in one column and the parallel column contains an Arabic translation in Greek characters. Michael Macdonald, a palaeographer and an expert on Ancient Arabic, makes the following evaluation of this text (2004: 50):
Following a detailed study of this text I am convinced that it is pre-Islamic. This is the most valuable text in Old Arabic so far discovered since the Greek transliteration seems to have been made with great care and consistency from an oral source, and thus is uncomplicated by the orthographic conventions of another script. ((Sydney Griffith (1985: 134), in line with his general thesis, had suggested an eighth century date.))
In this fragment, the Greek term for God, ho theos, is found in verses 22, 31, and 59. It is translated there into Arabic as αλλαυ (= Arabic allâh) (where the Arabic /h/ has been transliterated with a Greek upsilon, as is the custom in this manuscript). This provides further evidence that pre-Islamic Arab Christians were using allâh to refer to God. One also notes that the Greek letter lambda is doubled; this demonstrates that the Arabic letter lām must have been pronounced double by this time as well. Given the practice in ancient Arabic of not writing doubled letters twice or an internal /ā/ vowel at all (Macdonald 1999b: 271), this Greek evidence provides further support for Winnett’s claim (1938) that ’lh in the epigraphic evidence was pronounced as allâh.
The New Testament or parts of it were translated many times into Arabic. Kachouh (2006; personal correspondence) has compared 210 different ancient and medieval translations, and he discerns among them 22 different translation traditions. The extant manuscripts date from the post-Islamic period, but there is evidence for pre-Islamic translations of the Gospel, although scholars disagree on the matter. ((Orientalist Anton Baumstark (1934) argued that the Gospel and Psalter were translated into Arabic prior to Islam. But Griffith (1985: 166) disagreed, noting that since no dated Arabic manuscripts survive from the pre-Islamic period, “All one can say about the possibility of a pre-Islamic, Christian version of the Gospel in Arabic is that no sure sign of it’s actual existence has yet emerged.” Griffith’s judgment seems a bit too dismissive, however, because few manuscripts were dated in that period in any language, and in later manuscripts that do have a date, it is usually the date of copy that is noted rather than the date of the original translation.)) It is said that Waraqah ibn Nawfal translated the Gospel and other portions of the Bible into Arabic in Mecca in the sixth century, which is well before John of Sedra (see note 40). Ibn Isḥāq (died 761) wrote that in 570 AD one of the stones of the Ka‘ba was found to have writing on it, and the words he quotes are clearly taken from Matthew 7:16 (Guillaume & Ibn Ishaq 2002 : 86). Irfan Shahid (1971: 249–250) presents evidence that before 520 AD the Christians of Najran had the Gospel in their language, meaning their dialect of Arabic, written in Musnad script. Trimingham (1979: 225) cites Michael the Syrian’s 12th century Chronicle to the effect that John of Sedra, Patriarch of Antioch, arranged in the early 7th century for “the first translation of the four Gospels” into Arabic for use by Muslim scholars. The Patriarch’s translation does not survive, except perhaps, for a passage from John that is “quoted” by Ibn Isḥāq. ((For discussion of the quote in Ibn Isḥāq, its use relation to the Palestinian Syriac Lectionary, and its possible relation to the translation sponsored by John of Sedra, see (Griffith 1985: 137), (Guillaume & Ibn Ishaq 2002 : 104), and (Guillaume 1950: 289–296). According to Michael, the Muslim scholars asked John of Sedra to use terminology in the translation that was acceptable to them; Michael says that John resisted, but the passage cited in Ibn Isḥāq is clearly contextualized, in that ‘Father’ is translated as Rabb, which then meant “Sustainer, Patriarch, Paterfamilias”, whereas the extant Christian Arabic manuscripts use Rabb to translate κύριος “Lord”.).
Although Trimingham cites Michael’s note as part of his argument that Arab churches used Aramaic liturgy and Scripture rather than Arabic, it is not clear what lies at the origin of this tradition. It could represent the first translation of the four Gospels into the Arabic language, or the first that used the new Arabic script, or the first authorized by the Syrian Orthodox Church, or the first for Muslim scholars, or the first that included the “Four Separated Gospels”, in contrast with the Diatessaron. (The Diatessaron Gospel harmony had been the standard form of the Gospel in Syriac until the 5th century. The Arabic Diatessaron that survives today was translated or revised by Abdullah ibn al-Tayyib in the 10th century).))
Many translations were lost, largely due to the destruction of monasteries, but copies of many translations have survived and can be viewed in various libraries and museums. The following chart lists the principal ancient and medieval Arabic translations that I have examined, ((I am grateful to Kenneth Bailey for loaning me photocopies and microfilms of many of these manuscripts so that I could duplicate them, and to Berend-Jan Dikken as well, for providing digital versions of some manuscripts. Microfilms of some manuscripts were ordered from libraries.)) showing the dates of the surviving manuscripts and the evident origin and source language of each translation. ((For discussions of provenance see (Graf 1905), (Metzger 1974); (Bailey 1989); (Griffith 1985).)) The translations that appear to be earliest in origin are presented first:
|Name of Arabic Version||Source Language||Place of Origin||Date of Origin||Date of ms|
|The Palestinian Gospels ((The Palestinian Gospels are also called “Mt. Sinai Family A”. This family of manuscripts includes Sinai 72 (copied 897 AD), Sinai 16, Sinai 74, Vatican Borgia 95, and Berlin 1108 (copied 1046 AD). This version is quoted in some patristic quotations. A study of this version is presented in (Baumstark 1929-1930). According to Griffith (1985: 153–154), Baumstark shows that all of these manuscripts are marked for the Sunday lectionary readings according to a liturgy that was used in Palestine prior to the rise of Islam but not afterwards. He reasons, therefore, that this translation was made prior to the rise of Islam.))||Greek||Palestine||prob. pre-Islamic||9th to 11th|
|The Elegant Gospels ((The Elegant Gospels survive in Leiden 2378 (OR 561), Vatican Arabic 17 and Vatican Arabic 18. The translation uses rhyming prose, with names and terms similar to those in the Qur’an. It translates both the Greek epithet (ho) theos “God” and the proper noun Kurios “LORD” as al-lâh. For a critical edition with textual commentary see (Nasr 2000). She writes that Vatican 18 was copied in Cairo in AD 993, while Vatican 17 was copied in AD 1009. The date of the Leiden 2378 manuscript is unknown.))||Syriac||Uncertain||pos. pre-Islamic||10th|
|Vatican Arabic 13 Gospels ((Bailey and Staal (1982) date Vatican Arabic 13 to the eight or ninth centuries but say the translation itself is earlier. It is not a single work but a collection of works from five different scribes. The Gospel of Matthew appears to be the oldest. Bailey and Staal think it was translated from Greek, in part at least, but Syriac influence is evident as well. For example, it uses the Aramaic loan-word sali-h? for apostle (as well as the term h?awa-ri- for disciple), and this is a characteristic of Syriac-based translations. Kachouh has compared Vatican Arabic 13 diligently with other versions and is convinced that the Gospels were translated from Syriac and the Epistles from Greek.))||Syriac||Uncertain||pos. pre-Islamic||8th or 9th|
|Vatican Arabic 13 Gospels||Greek||Uncertain||prob. post-Islamic||9th|
|Treatise on the Triune God ((The so-called “Treatise on the Triune Nature of God” is not a Bible translation, but it contains many Biblical quotations in Arabic. The terminology resembles the usage in the Qur’an and in the Elegant Gospels, but the wording is different from the Elegant Gospels. Part of the Treatise is found as part of Mt. Sinai 154 and was published in (Gibson 1899). Additional fragments of this treatise exist in other locations but have not been published.))||Palestine||776||9th|
|Mt Sinai 151 ((Mt. Sinai Arabic 151 consists of the Acts and Epistles. It was written in Damascus at different stages. The Epistles of Paul are dated 867, and the dates of the Acts and other Epistles are assumed to be near that.))||Syriac||Damascus||867||867|
|Vatican 71 ((Other manuscripts similar to Vatican 71 include Vatican Arabic 467, Leiden 2376, Leiden 2377, and St. Petersburg Asiatic Museum D226.))||Greek & Syriac||Damascus||10th century||11th century|
|Abdullah ibn al-Tayyib ((Ibn al-Tayyib translated Tatian’s Diatessaron, using the text of the Syriac Peshitta, and he produced a translation of the four Gospels with a running commentary that is still used today. Manuscripts can be found in many libraries.))||Syriac||Baghdad||980||many|
|Alexandrian Vulgate ((The Alexandrian Vulgate is represented in Vatican Coptic 9 (Coptic and Arabic). It became the normative Arabic translation of the New Testament in Egypt. A later version was printed in Rome in 1591, and afterwards in the Paris and London Polyglots.))||Coptic||Alexandria||uncertain||1202|
|Mt. Sinai 76 (Sinai Family B)||Greek?||uncertain||uncertain||13th|
|Lectionary of Abdishu ((This Lectionary of Abdishu (“servant of Jesus”) is translated in rhyme. It was widely used in the Levant.))||Syriac||Levant||1299||many|
Years ago I tabulated the key terms used in these translations in a comparative fashion. They exhibit such a diversity of wording that one is forced to conclude that they represent several independent traditions of translation. In other words, the earliest ones seem to have been translated independently of one another by different churches in diverse locations from different source texts. One of the things they have in common, however, is that they all use the word allâh to refer to God. Since the Arab Christians were spread over a vast region and belonged to diverse and warring churches long before the rise of Islam, the fact that all of them used allâh to refer to God in the earliest surviving translations is an indication that the term allâh must have been in widespread use by Arab Christians in pre-Islamic times. ((It is not known when the first Jewish translations of the Bible were made into Arabic, but ancient fragments survive that use the pre-Islamic spelling of allâh with one el letter, as אלה (see note 49). The principal Medieval Arabic Jewish versions were made in the tenth-century by Saadia Gaon and by the Karaites. They used al-lâh, sometimes written in Hebrew script as אללה , to translate both (hâ-)elohîm “God” and YHWH (Polliack 1997).))
More recently Hikmat Kachouh has studied Codex Sinaiticus Arabicus (Sinai Arabic New Finds 8 & 28 Parch), a newly discovered Arabic manuscript of the Gospels. He shows in an article forthcoming in Novum Testamentum that it represents a translation made from a Greek vorlage whose unique text-type lies between Sinaiticus and Beza. Since it is highly unlikely that a translator would base his work on a source text that was no longer in use, and since by the sixth century the Byzantine/Syrian text-type of the Gospels had become the standard in the Middle East and had replaced the previous text-types, this Arabic translation must almost certainly have been made before then, at a time prior to Islam. Since the translation uses allâh for the name of God, it is another witness to the usage of that term by Arabic-speaking Christians.
The Qur’an reflects pre-Islamic Christian usage of allâh to refer to God
Prior to his mission, Muhammad interacted with a number of Arabic-speaking Christians and Jews, notably the Jordanian monk Bahira (Guillaume & Ibn Ishaq 2002 : 89–91) and the Meccan Nestorian monk Waraqah ibn Nawfal, who was an older cousin of Muhammad’s wife Khadija. ((The Hadith (Sahih al-Bukhari, volume 9, book 87, number 111) and Aghani state that Waraqah ibn Nawfal, the cousin of Muhammad’s wife Khadija, was a Christian who studied the Bible and who translated parts of one or more Gospels into Arabic. Even if it was an apocryphal Gospel that he translated, like the Infancy Gospel of Thomas, (which has parallels with some Qur’anic stories of Jesus), such an activity would suggest that Muhammad had access to Christian Arabic terms for God. There is no evidence that Waraqah ever became a Muslim, and some authors have suggested that he was one of the disputants.)) He also attended lectures by an unnamed Christian teacher near Mecca. ((The usual speculation is that this teacher was Bahira, as noted in Hughes (2001 : 30):
Sprenger thinks that Baḥīrā remained with Muhammad, and it has been suggested that there is an allusion to this monk in the Qur’ān, Sūrah xvi. 105: “We know that they say, ‘It is only a man who teacheth him.’” Ḥusain the commentator says on this passage that the Prophet was in the habit of going every evening to a Christian to hear the Taurāt and Injīl.
)) After the commencement of his mission he often debated with Jews and Christians, including a delegation of Christians from Najran. The participants in such discussions must have used mutually intelligible names for God. During the initial stage of the prophet’s mission, as reflected in the Meccan suras of the Qur’an, he presented his Qur’anic prophecies as an affirmation of the Bible and as a continuation of the Jewish and Christian prophetic tradition. He could not have made this claim if he had been proclaiming in allâh a different god or if he had been using radically different terminology from that used in the Arab Christian tradition that he claimed to be affirming and continuing.
There are also scholars who argue that many of the Meccan suras are based on pre-Islamic Arab Christian hymns. This is based in part on the presence of Syriac words that were used by Christians but were not used or understood by non-Christian Arabs. Luxenburg (2004) and Lüling (2003) show that when the words are interpreted in accord with their meaning in Syriac, it is possible, with some further editing, to recover fragments of Christian hymns and poetry. It is also based on similarities between pre-Islamic poetry and verses of the Qur’ān, as shown by (Abul Kasem 2007). Lüling (2003: 1) states the thesis of his book quite forcefully: “The text of the Koran as transmitted by Muslim Orthodoxy contains, hidden behind it as a ground layer and considerably scattered throughout it (together about one-third of the whole Koran text), an originally pre-Islamic Christian Text.” It might be noted that medieval Christian sources claim that parts of the Qur’an were written by the Nestorian monk Bahira (Abel 1999). If many of the Meccan suras were indeed drawn from Christian poetry, then their terminology, including the name allâh, would seem to have its origin in Christian Arab sources.
In later stages of his work, the prophet of Islam faced increasing resistance and disputation from Christians. There are a number of passages in the Qur’an that cite these disputes. Some of these passages quote statements made by the Christians, and it might be noted that the Christians are quoted as using the term allâh. Examples include their claim that “allâh is Jesus” (Al-Ma’ida 5:17), that Christians are “sons of allâh” (Al-Ma’ida 5:18), and that Jesus is a “son of allâh” (At-Tawba 9:30). Nowhere in the Qur’an is there any indication that Arab Christians and Jews referred to God by a name different from those used in the Qur’an. All of the disputation passages reflect situations in which the same God is in view and is referred to in the same basic ways.
In light of this evidence from inscriptions, historical documents, and Arabic translations of the Bible, we can conclude that allâh was the term used by pre-Islamic Jewish and Christian Arabs to refer to God. ((Christians in Yemen and other parts of southern Arabia also called God raḥmān-an, which equals northern Arabic al-raḥmān, but this was in reference to God as the Father. In 541 AD, King Abraha, the Christian ruler of Yemen and southern Arabia, placed an inscription on the dam at Marib (in Musnad script) that began with an expression of the Trinity: “By the power and grace of the Raḥmān and his Christ and the Holy Spirit” (Academie des Inscriptions et Belles-lettres 1911: fig. 541, p. 278). Abraha also placed an inscription on a cliff at Mureighan that begins “by the power of the Raḥmān and his Christ” (Wickens et al. 1954). The Jewish Aramaic term rahṃân-â רחמנא was a common epithet for God among Jews, and one finds Jewish Arabic inscriptions in South Arabia that use this term (Abdallah 1987), so this is the evident source of the Arabic term. The root rḥm means “womb”, and raḥmān describes a male who is compassionate like a father. The Aramaic term raḥīm-â means “beloved one”. It is possible that the term raḥīm was used by Christians in Southern Arabia to refer to Jesus as God’s Beloved and the bearer of his love. In that case the basmala might be derived from an originally Christian formula that meant “in the name of God, the Compassionate One and the Beloved One and the Holy Spirit”. No evidence, however, survives to verify or falsify this speculation.)) In the next section I will argue that Jews and Christians introduced this term themselves into Arabic from Aramaic.
Next Week: Who was ‘Allah’ before Islam? (2)
Used by permission from the author. All rights to this material are reserved. Materials are not to be distributed to other web locations for retrieval, published in other media, printed for distribution or mirrored at other sites without written permission from the author(s)