Dan Gibson is a historian who has worked to fit the geography of the Islamic sources to other historical and archeological data. This approach has been quite successful in other contexts but has largely been neglected with Islam because of the seeming impossibility of the task. Simply put, Islamic geography seems so confused that it is unlikely to be able to be made fully coherent. The lapse of hundreds of years of oral transmission is, of course, a large part of that–little details like exact locations are easy to lose when the people telling the story live far away from its origins. Nevertheless, Gibson took on this work with an admirable amount of seriousness.
In his early work, Gibson discovered that the earliest qibla of existing mosques is definitively, without question, in Petra, not Mecca. This led him to conclude that the ancient city of Muhammad as also Petra, not Mecca. Certainly, if the Kaaba were in Petra, a lot of the assertions surrounding it would be a great deal more plausible than if it is in Mecca, as it has been historically taught.
However, I believe that Gibson is wrong. This is not to denigrate the excellent and difficult work he has done. He is correct about any number of things–which I will cover later. However, the ancient city of Muhammad was, in fact, Mecca/Makkah and no other place. Here, I will explain why.
1. General linguistic Issues
The language of the Quran has too many loan words from Abyssinian languages and Old South Arabian languages and relatively few loan words from Greek and Aramaic, on a general level.
Many Muslims have been taught that the Quran is written in “pure classical Arabic.” This is utter hogwash, of course. There is no such thing as a “pure language.” Every language is constantly changing, and every language borrows from its neighbors.
One of the most obtuse defenses of this idea is that the Quran actually somehow defines “pure classical Arabic”–whatever the source of the words, by being incorporated into the Quran in a certain way, they have then been transformed by Allah into this sacred “pure classical Arabic” language. This same defense is used for the many grammatical infelicities and outright errors in the text of the Quran: because a word is used in a certain form in the Quran, that then defines the correct way to use that word.
This idea is so absurd on the face of it that it hardly deserves a response, but I will give one, anyway. The Quran was written in a language that contains real words spoken by real people. The usage in the Quran is inconsistent at times with a usage that would have seemed correct by these real people. Additionally, the ultimate origins of a number of these words came from sources that were not Arabic and sometimes not even Semitic.
The direction of these borrowings make sense in the geographical context of Mecca. They do not make sense in the geographical context of Petra.
It should be noted that the idea of “perfect classical Arabic” is an innovation. There is no basis for this doctrine in the early Islamic sources. Muhammad made a point that the Quran was being given in the plain, everyday language of the people around him. That language, like any language, would include borrowings from neighbors.
2. Pre-Islamic Historical Knowledge
The early Islamic sources attempt to capture a history of pre-Islamic Arabia, and in so doing, they show a very strong orientation toward Yemen, a moderate knowledge of Abyssinia, only a peripheral knowledge of Syria, and almost complete ignorance of the Byzantines and Sassanids. This awareness fits with the geographical location of Mecca, and it makes no sense with the geographical location of Petra.
3. A Yemenite View of Religion
The Himyarite tribe of Yemen saw religion as inherently political. This is a common belief with many pagans, but it was taken to a new level by the Himyarites, who viewed Christian missionaries as an act of political aggression. While it is true that the Byzantines found it easier (and also more morally acceptable) to have friendly relationships with Christian countries, missionary conversions did not mean that states were being forced into a hegemonic relationship with the Byzantines. For example, Axum converted to Christianity, and though they maintained friendly relations with Byzantium, they were never under the Byzantine sphere of control in any way. However, the Himyarite view of power was fundamentally hegemonic and tribal. Their rejection of Byzantine control is inextricably linked to their conversion to Judaism and their bloody pogroms against Christians in Yemen.
The view of religion being inherently political is absolutely fundamental to Islam. It makes sense in terms of a Jewish/Yemenite context, with the twin influences of Himyarite power structures and Talmudic law, but it isn’t a worldview that would be compatible with the experience of Petra.
4. The Origin of Al-Rahman
One of the most common words for Allah in the Quran is al-Rahman. This word does not mean “the Merciful One,” as it is sometimes translated. That is “al-Rahim.” In fact, all of the translations of this name are in fact an interpolation or a fabrication, and that the Arabs of Muhammad’s day were not sure what the word meant.
At-Tirmidhi recorded from ‘Abdur-Rahman ibn ‘Awf that he heard the Messenger of Allah say, “Allah the Exalted said, ‘I am al-Rahman. I created the raham [womb] and derived a name for it from my name. Hence, whoever keeps it, I will keep ties to him, and whoever severs it, I will sever ties with him.'” This is the only explanation of al-Rahman that exists in the early Islamic sources that actually comes from Muhammad.
Al-Qurtubi claimed that this statement shows that al-Rahman comes from rah’mah, which means “the mercy,” but it actually shows no such thing. Al-Rahman being similar to “womb” (ramah) does not at all show evidence that it’s derived from the word for “mercy” (rah’mah).
Al-Qurtubi also claimed, “The Arabs have denied the name al-Rahman because of their ignorance about Allah and his attributes.” This seems to blame the Arabs for being so confused! He also claimed, “It was said that Ar-Rahman and Ar-Rahim had the same meaning, such as the words Nadman and Nadim.” There are various other narrations of this sort, all of which attempt to connect al-Rahman with al-rah’mah, with statements that try to infuse the word with mystical importance, such as, “al-Rahman, which is used exclusively for Allah, is a name which encompasses every type of mercy that Allah has” in comparison to the generic and therefore less elevated “al-Rahim.” But this does not avoid the fact that al-Rahman was simply a foreign and unfamiliar word to the Arabs.
The confusion of Muhammad’s contemporaries was evident in another narration. Suhail ibn Amr, one of the foremost pagans of Mecca, told Muhammad that he should change his insignia (the bismillah) from “By the name of God, al-Rahman, the Merciful One” to “By your name, O Allah!” because, as Suhail said, “As for ‘Rahman,’ by Allah, I do not know what it means.”
However, “Rahman” is almost certainly a loan from Old South Arabian. RHMNN–that may be “Rahamanan”–was used in Himyarite Jewish inscriptions to refer to the God of Israel. It was also used in the Christian king Abraha’s inscriptions to refer specifically to God the Father.
Before the Himyarite converted to Judaism, however, RHMN was used as generic title used for chief pagan gods (including Marduk) and for human kings, as there is epigraphic evidence of the Semitic root RHMN used this way as far north as Syria. Though it does nevertheless seem to be ultimately derived from a Semitic root that means compassion or mercy, the word itself befuddled early Muslims, which means that it could not have been known as an Arabic word. As it seems very unlikely that Muhammad embraced a word directly from a pagan context, that makes Yemen and the Old South Arabian language the only likely place for the borrowing to have occurred. It seems impossible that this usage could have reached Muhammad in Petra rather than in Mecca.
5. The Ignorance of Christianity
Petra was an area where its inhabitants would have had substantial contacts with Christians on a regular basis. While there were some Christians in Yemen, Mecca was quite isolated from them in general. If Muhammad had lived in Petra, it seems impossible that he would have misunderstood the doctrine of the trinity, that he would have confused Mary the mother of Jesus with the sister of Aaron and Moses, or that he would have gotten the name of Jesus wrong. Yet all of these things are in the Quran. And not only does the Quran get them wrong, but Muhammad is not made aware even of the slightest mistake until he goes to Yathrib. If Yathrib has more Christian contacts than his place of origin, his place of origin fits Mecca and not Petra.
His confusion about the nature of the trinity is also reinforced by the Yemenite formula for the trinity. Often, they named only the Father and Son as Rahmanan and “his son, the Messiah.” When they named all three, the formula was Rahmanan, the Messiah, and the Ruach (or Spirit). Since the Jews used the term Rahmanan for the God of Israel, it was possible that the Yemenite terminology, which seemed like someone was being added to or associated with the God of Israel, could be a source of confusion for Muhammad. Nothing except the basest kind of ignorance could explain his ignorance otherwise.
6. The Inaccessibility of the Scriptures
Related to this is the fact that the Hebrew and Christian scriptures were simply not available to Muhammad at any time. He got a few things correct, but his use of what amounts to Jewish folklore and Christian “fan fic” is much more extensive than his access to the actual scriptures, and Muhammad lacked the sophistication to know the difference.
I find it highly improbable that Muhammad’s environment could have been so highly ignorant if he had lived in the cosmopolitan city of Petra. Of course, the glory days of Petra were behind it, but Petra was still not a backwater, and Muhammad’s knowledge of Hebrew and Christian scriptures are not compatible with a place as sophisticated as Petra.