According to the Qur'an, the direction of prayer (the Qibla), was canonized (or finalized) towards Mecca for all Muslims in or around 624 AD (see Sura 2:144, 149-150).
Yet, the earliest evidence from outside Muslim tradition regarding the direction in which Muslims prayed, and by implication the location of their sanctuary, points to an area much further north than Mecca, in fact somewhere in north-west Arabia (Crone-Cook 1977:23). Consider the archaeological evidence which has been and is continuing to be uncovered from the first mosques built in the seventh century:
According to archaeological research carried out by Creswell and Fehervari on ancient mosques in the Middle East, two floor-plans from two Umayyad mosques in Iraq, one built at the beginning of the 8th century by the governor Hajjaj in Wasit (noted by Creswell as, "the oldest mosque in Islam of which remains have come down to us" - Creswell 1989:41), and the other attributed to roughly the same period near Baghdad, have Qiblas (the direction which these mosques are facing) which do not face Mecca, but are oriented too far north (Creswell 1969:137ff & 1989:40; Fehervari 1961:89; Crone-Cook 1977:23,173). The Wasit mosque is off by 33 degrees, and the Baghdad mosque is off by 30 degrees (Creswell 1969:137ff; Fehervari 1961:89).
This agrees with Baladhuri's testimony (called the Futuh) that the Qibla of the first mosque in Kufa, Iraq, supposedly constructed in 670 AD (Creswell 1989:41), also lay to the west, when it should have pointed almost directly south (al-Baladhuri's Futuh, ed. by de Goeje 1866:276; Crone 1980:12; Crone-Cook 1977:23,173).
The original ground-plan of the mosque of ‘Amr b. al ‘As, located in Fustat, the garrison town outside Cairo, Egypt shows that the Qibla again pointed too far north and had to be corrected later under the governorship of Qurra b. Sharik (Creswell 1969:37,150). Interestingly this agrees with the later Islamic tradition compiled by Ahmad b. al-Maqrizi that ‘Amr prayed facing slightly south of east, and not towards the south (al-Maqrizi 1326:6; Crone-Cook 1977:24,173).
If you take a map you will find where it is that these mosques were pointing. All four of the above instances position the Qibla not towards Mecca, but much further north, in fact closer possibly to the vicinity of Jerusalem. If, as some Muslims now say, one should not take these findings too seriously as many mosques even today have misdirected Qiblas, then one must wonder why, if the Muslims back then were so incapable of ascertaining directions, they should all happen to be pointing to a singular location; to an area in northern Arabia, and possibly Jerusalem?
We find further corroboration for this direction of prayer by the Christian writer and traveler Jacob of Edessa, who, writing as late as 705 AD was a contemporary eye-witness in Egypt. He maintained that the ‘Mahgraye’ (Greek name for Arabs) in Egypt prayed facing east which was towards their Ka'ba (Crone-Cook 1977:24). His letter (which can be found in the British Museum) is indeed revealing. Therefore, as late as 705 AD the direction of prayer towards Mecca had not yet been canonized.
Note: The mention of a Ka’ba does not necessarily infer Mecca (as so many Muslims have been quick to point out), since there were other Ka’bas in existence during that time, usually in market-towns (Crone-Cook 1977:25,175). It was profitable to build a Ka’ba in these market towns so that the people coming to market could also do their pilgrimage or penitence to the idols contained within. The Ka’ba Jacob of Edessa was referring to was situated at "the patriarchal places of their races," which he also maintains was not in the south. Both the Jews and Arabs (‘Mahgraye’) maintained a common descent from Abraham who was known to have lived and died in Palestine, as has been corroborated by recent archaeological discoveries (see the earlier discussion on the Ebla, Mari and Nuzi tablets, as well as extra-Biblical 10th century references to Abraham in McDowell 1991:98-104). This common descent from Abraham is also corroborated by the Armenian Chronicler, Sebeos, as early as 660 AD (Sebeos 1904:94-96; Crone-Cook 1977:8; Cook 1983:75).
According to Dr. Hawting, who teaches on the sources of Islam at the School of Oriental and African Studies (SOAS, a part of the University of London), new archaeological discoveries of mosques in Egypt from the early 700s also show that up till that time the Muslims (or Haggarenes) were indeed praying, not towards Mecca, but towards the north, and possibly Jerusalem. In fact, Dr. Hawting maintains, no mosques have been found from this period (the seventh century) which face towards Mecca (noted from his class lectures in 1995). Hawting cautions, however, that not all of the Qiblas face towards Jerusalem. Some Jordanian mosques have been uncovered which face north, while there are certain North African mosques which face south, implying that there was some confusion as to where the early sanctuary was placed. Yet, the Qur'an tells us (in sura 2) that the direction of the Qibla was fixed towards Mecca by approximately two years after the Hijra, or around 624 AD, and has remained in that direction until the present!
Thus, according to Crone and Cook and Hawting, the combination of the archaeological evidence from Iraq along with the literary evidence from Egypt points unambiguously to a sanctuary [and thus direction of prayer] not in the south, but somewhere in north-west Arabia (or even further north) at least till the end of the seventh century (Crone-Cook 1977:24).
What is happening here? Why are the Qiblas of these early mosques not facing towards Mecca? Why the discrepancy between the Qur'an and that which archaeology as well as documents reveal as late as 705 AD?
Some Muslims argue that perhaps the early Muslims did not know the direction of Mecca. Yet these were desert traders, caravaneers! Their livelihood was dependant on traveling the desert, which has few landmarks, and, because of the sandstorms, no roads. They, above all, knew how to follow the stars. Their lives depended on it. Certainly they knew the difference between the north and the south.
Furthermore, the mosques in Iraq and Egypt were built in civilized urban areas, amongst a sophisticated people who were well adept at finding directions. It is highly unlikely that they would miscalculate their qiblas by so many degrees. How else did they perform the obligatory Hajj, which we are told was also canonized at this time? And why are so many of the mosques facing in the direction of northern Arabia, or possibly Jerusalem? A possible answer may be found by looking at archaeology once again; this time in Jerusalem itself.
In the centre of Jerusalem sits an imposing structure (even today) called the Dome of the Rock, built by ‘Abd al-Malik in 691 AD One will note, however, that the Dome of the Rock is not a mosque, as it has no Qibla (no direction for prayer). It is built as an octagon with eight pillars (Nevo 1994:113), suggesting it was used for circumambulation (to walk around). Thus, it seems to have been built as a sanctuary (Glasse 1991:102). Today it is considered to be the third most holy site in Islam, after Mecca and Medina. Muslims contend that it was built to commemorate the night when Muhammad went up to heaven to speak with Moses and Allah concerning the number of prayers required of the believers (known as the Mi'raj in Arabic) (Glasse 1991:102).
Yet, according to the research carried out on the inscriptions on the walls of the building by Van Berchem and Nevo, they say nothing of the Mi'raj, but state mere polemical quotations which are Qur'anic, though they are aimed primarily at Christians. The inscriptions attest the messianic status of Jesus, the acceptance of the prophets, Muhammad's receipt of revelation, and the use of the terms "Islam" and "Muslim" (Van Berchem 1927:nos.215,217; Nevo 1994:113). Why, if the Dome of the Rock were built to commemorate that momentous event, does it saying nothing about it? Perhaps this building was built for other purposes than that of commemorating the Mi'raj. The fact that such an imposing structure was built so early suggests that this and not Mecca became the sanctuary and the centre of a nascent Islam up until at least the late seventh century, (Van Bercham 1927:217)!
From what we read earlier of Muhammad's intention to fulfill his and the Hagarene’s birthright, by taking back the land of Abraham, or Palestine, it makes sense that the caliph ‘Abd al-Malik would build this structure as the centre-piece of that fulfillment. Is it no wonder then, that when ‘Abd al-Malik built the dome in which he proclaimed the prophetic mission of Muhammad, he placed it over the temple rock itself (Van Berchem 1927:217).
According to Islamic tradition, the caliph Suleyman, who reigned as late as 715-717 AD, went to Mecca to ask about the Hajj. He was not satisfied with the response he received there, and so chose to follow ‘abd al-Malik (i.e. traveling to the Dome of the Rock) (note: not to be confused with the Imam, Malik b. Anas who, because he was born in 712 AD would have been only three years old at the time). This fact alone, according to Dr. Hawting at SOAS, points out that there was still some confusion as to where the sanctuary was to be located as late as the early eighth century. It seems that Mecca was only now (sixty years after the Muhammad’s death) taking on the role as the religious centre of Islam. One can therefore understand why, according to tradition, Walid I, who reigned as Caliph between 705 and 715 AD, wrote to all the regions ordering the demolition and enlargement of the mosques (refer to `Kitab al-`uyun wa'l-hada'iq,' edited by M. de Goeje and P. de Jong 1869:4). Could it be that at this time the Qiblas were then aligned towards Mecca? If so it points to a glaring contradiction to the Qur'an which established Mecca as the sanctuary and thus direction for prayer during the lifetime of Muhammad some eighty to ninety years earlier (see Sura 2:144-150).
And that is not all, for we have other archaeological and inscripted evidence which point up differences with that which we read in the Qur'an.