The Petra Fallacy: Early Mosques do face the Sacred Kaaba in Mecca

by David A King Published on: 15th September 2020

3.2 / 5. Votes 9

No votes so far! Be the first to rate this post.

Dan Gibson, a Canadian amateur archaeologist, is the latest of a number of revisionist historians of early Islam who are desperate to show that Islam did not start in Mecca, and hence that early Islamic history was forged.

Gibson claims that early mosques face the MODERN direction of Petra rather than the MODERN direction of Mecca, and that this points to Petra as the cradle of Islam rather than Mecca. He does not realize that these early mosques could not face Mecca in the modern sense and that it is foolish to expect that they should. Nor could they face toward Petra in a modern sense. Anyway mosques do not face Mecca; they face the Kaaba. How does one face a distance edifice that one cannot see? The first generations of Muslims had their own way of doing this, which was fully in keeping with their knowledge of simple folk astronomy. The methods they used have been documented in scholarly literature. Gibson compares the HISTORICAL mosque orientations with MODERN directions of Petra and Mecca from these mosque locations. He does not realise that historical qibla directions cannot be the same as MODERN  qibla directions, except by coincidence. In fact, the earliest mosque orientations were not calculated at all, but relied on astronomical horizon phenomena, not least because the Kaaba itself is astronomically aligned. Astronomical alignments are found elsewhere in Arabia, particularly in monumental pre-Islamic Nabataean architecture. 

Having discovered to his satisfaction that the mosques were aligned to Petra rather than Mecca, Gibson was able to claim that these early mosques were deliberately laid out in the modern direction of Petra. This is, of course, ridiculous, but to 'prove' that he is correct Gibson is prepared to deliberately distort the history of science and claim that his Nabataean Muslims could determine the direction of Petra accurately from places between al-Andalus to China using sophisticated scientific methods. Now only in the 9th century could astronomers in Baghdad determine the qibla there using a mathematical formula. Reasonably accurate qibla directions for major centres based on modern geographical coordinates became available only in the 19th century.

In brief, the history of the ways in which Muslims have laid out their mosques over a thousand years ago is a very complicated subject. To understand it, we have numerous medieval Arabic treatises on the ways to find the qibla using non-mathematical means (sacred geography) and the ways to find the qibla using mathematical procedures (but with pre-modern geographical data). And we have medieval Arabic astronomical, historical and legal Arabic treatises detailing the different qiblas underlying mosques in Cordóba, Cairo and Samarqand. Gibson and a missionary group (in their numerous comedy videos) are happy to laugh at all this material because they do not understand any of it. This enables them to laugh at Muslims who they think have faced the wrong direction towards Mecca in their prayers for so many centuries. They even suggest that Muslims should start praying toward Petra again. In fact, the last laugh is on those who think that any Muslims of sound mind ever faced Petra in their prayers.

Figure 1. The Kaaba (literally “The Cube”) in an early photograph.” (Source)


Note of the Editor: This short article is a collection of paragraphs extracted from the full paper that was originally published here by David A. King. The original paper contains explanations of various early mosque orientations, showing that they have nothing to do with Petra. Prof. King has added some remarks dated September 2020, at the end of this shorter version.



For over 1,400 years, Islamic civilization has taken the orientation of sacred space more seriously than any other civilization in human history. The sacred direction towards the sacred Kaaba in Mecca is called qibla in the languages of the Muslim commonwealth. The ways in which Muslims have determined the qibla over the centuries constitute a complicated story, but several facts are known:

  • The Arabs before Islam had an intricate system of what we now call ‘folk astronomy’ based on what one can see in the heavens.
  • The Kaaba has a rectangular base which is astronomically aligned; its major axis points toward the rising of Canopus, the brightest star in the southern sky, and its minor axis is defined by summer sunrise and winter sunset. Its four corners point roughly in the cardinal directions.
  • The Muslims developed a sacred geography in which, over the centuries, various schemes were developed in which segments of the perimeter of the Kaaba corresponded to sectors of the world which had the same qibla, defined in terms of astronomical risings and settings. The first such schemes appear in Baghdad in the 9th century.
  • By the early 9th century, the Muslims had accessed the geographical and mathematical knowledge of their predecessors, which meant that for the first time they could calculate the qibla using (medieval) geographical coordinates and mathematical procedures. (Of course, this would not mean that they could find the MODERN direction of Mecca.)
  • From the 7th to the 9th century and also occasionally thereafter until the 19th century, Muslims used astronomical alignments to layout the qibla. From the 9th century to the present Muslims have also used mathematical methods to calculate the qibla.

Few people know anything about this these days. Indeed, most Muslims think that all mosques face Mecca. Yet if they would investigate just a few historical mosque orientations they would be surprised. For medieval mosques face the Kaaba rather than Mecca. There is a subtle, but highly significant difference. How can they ‘face’ a distant edifice that is not visible? How these mosques actually ‘face’ the Kaaba is something we moderns have to learn. And the matter of the qibla is not only about mosques: it is about every Muslim at home and abroad, in life and in death, who follows the prescriptions relating to the sacred direction of Islam.

* * * 

One of my concerns over the past 50 years has been to attempt to document – mainly for the first time – the ways in which Muslims over many centuries have used astronomy in the service of their religion:

  • to regulate the lunar calendar through the sighting of the crescent;
  • to organize the times of the five daily prayers; and
  • to determine the qibla or sacred direction toward the Kaaba.

To do this I first read what my teachers Karl Schoy (1877-1925) and Ted Kennedy (1912-2010) had written about these subjects using medieval Arabic sources. Particularly important were Kennedy’s translations of and commentaries on the writings of al-Bīrūnī, the greatest scientist in early Islamic history, which dealt with the second and third of these topics.

I spent many years looking at thousands of medieval Arabic manuscripts and hundreds of scientific instruments in libraries and museums around the world. Since nobody had ever looked at most of these manuscripts for centuries, I inevitably found things that were new. Some of my results took some Muslim colleagues by surprise. Western colleagues are, I find, becoming less and less interested in anything to do with classical Islamic Studies. And that field is plagued by revisionists who think that no medieval Arabic texts are trustworthy and who eagerly rewrite a chapter of Islamic history relying instead on the ramblings of some early Christian bishop in Armenia (I exaggerate, but not much).

Some of my publications in the history of Islamic astronomy include studies of the following subjects:

  • the astronomical alignments of the rectangular base of the Kaaba;
  • the methods with which Muslims from the earliest period could have determine the qibla by simple folk astronomy;
  • the notion of a sacred geography about the Kaaba, with sectors of the world having the same qibla defined by astronomical horizon phenomena;
    the methods by which the Muslim scientists could calculate the qibla for a given locality;
  • the geographical tables showing longitudes and latitudes of hundreds of localities from al-Andalus to China together with their qiblas in degrees and minutes;
  • the extraordinarily sophisticated mathematical tables displaying the qibla for any locality with which the user enters its (medieval) longitude and latitude in the table and reads the value of the (medieval) qibla;
  • the remarkable cartographical grids produced by Muslim scientists enabling the user to reading off the (medieval) qibla on a circular scale and the distance to Mecca on a diametrical scale.
  • the medieval Arabic texts discussing the palettes of accepted directions for the qibla and for mosque orientations in specific localities, which partly explains the wide range of mosque orientations in these places (notably Córdoba, Cairo and Samarqand).

Over the past few decades numerous colleagues have published papers on various mathematical procedures proposed by individual Muslim scholars for finding the qibla, and some of my colleagues and former graduate students have written on the procedures involving folk astronomy and astronomical alignments. The interested reader can survey what has been written on historical qibla-determinations in the bibliography appended to this paper.

We have left it to others to write on such controversial topics as the conflict regarding the qibla – is it south-east or north-east? – amongst Muslims in North America. Frequently over the years, other folks have introduced the factor that the Earth is not a sphere into the qibla discussion, which is not helpful.

In 1999 I published a book dealing with the way Muslims have determined the sacred direction over for some 1,400 years. This presented an overview of the earliest procedures of using astronomical alignments to face an astronomically-aligned Kaaba, with different means of calculating the qibla using geographical coordinates and trigonometric or geometric methods. But the book focusses on the mathematical tables that were devised giving the qibla as an angle in degrees and minutes to the local meridian for the whole Muslim world; the geographical tables giving for the principal localities in the Muslim world the qibla and distance to Mecca; and the cartographical Mecca-centred grids which enable the user to read off the qibla and distance to Mecca for any locality in the (classical and medieval) world.

None of these materials was known 50 years ago. And inevitably none of them are mentioned in uninformed popular accounts of the qibla such as one finds in Wikipedia. I never thought while preparing all my research that someday someone would come along and announce that all early mosques are oriented toward a location other than Mecca. No serious scholar, Muslim or non-Muslim, would ever have thought that mosques might have been deliberately oriented toward somewhere other than Mecca. If they had, they would rightly be considered to be deranged.

Figure 2.  The orientation of the Kaaba mentioned in medieval texts and confirmed by satellite images, taking into consideration the surrounding skyline. Canopus (Suhayl) is the brightest star in the southern sky. The direction of the rising of Canopus is conveniently perpendicular to the axis between summer sunrise and winter sunset for the latitude of Mecca. In pre-Islamic folklore the walls of the Kaaba were associated with the four ‘cardinal’ winds. Note that if one standing in front of the SW wall one is facing (istaqbala) the qabūl wind, also called ṣabā’; in this position one is facing summer sunrise with (formerly) fortunate Yemen (al-Yaman) on the right and ominous Syria (al-sha’m) on the left. Some revisionists have claimed that the orientation of the Kaaba (with al-ḥijr!) may have been altered on one of the several occasions when the edifice was rebuilt after destructive floods. Revisionists have to be very innovative when confronted with an edifice that is as ancient as the Kaaba.  Not to scale. (Source)

Revisionist fascination with N. W. Arabia

Some 50 years ago some over-enthusiastic London-based Arabists – John Wansbrough and his students Michael Cook & Patricia Crone – came up with the idea that Islam began not in Mecca but somewhere unspecified in N. W. Arabia. This was a curious idea, not least because there were no obvious potential sites. One of the principal and most convincing arguments for their bold assertion was the ‘fact’ that the earliest mosques in Egypt and Iraq do not face Mecca, but rather some locality in N. W. Arabia. Some 25 years ago I pointed out to Michael Cook the folly of this assertion, explaining that the earliest mosque in Egypt faces winter sunrise and the earliest mosque in Iraq faces winter sunset; so, of course, these mosques do not face (the MODERN direction of) Mecca. Nor were they deliberately aligned towards anywhere in N. W. Arabia. They were deliberately aligned to face toward the Kaaba. Cook reacted to this information by saying, most appropriately: “It’s a bit late”.

Yes, the earliest Muslims in Egypt and Iraq used winter sunset and winter sunrise, respectively, for the qibla, not because they were stupid, but because they were smart. How else to face an edifice they could not see: all savvy ancient peoples have used astronomical alignments for one reason or another. From al-Andalus to Central Asia early mosques were built in astronomical directions later referred to as qiblat al-ṣaḥāba or qiblat al-tābiʿīn, ”the qibla of the first or second generations of Muslims”.

My present intention is simple: it is to warn the unsuspecting reader that the only other person ever to have written generally on the subject of mosque orientations

(a)  has no qualifications to correctly interpret the available data;

(b)  has no understanding of the fact that MODERN directions from one place to another cannot be used to investigate the reasons underlying the orientation of PRE-MODERN architecture;

(c)  seems oblivious to the fact that there is a well-established discipline called archaeoastronomy and has no understanding of astronomical alignments;

(d)  has erred monumentally in his interpretation of mosques that were built on pre-existing religious architecture or to fit with pre-Islamic city plans;

(e)  has no understanding of how mosques were laid out over the centuries;

(f)  has no control over any of the numerous medieval Arabic sources – legal, astronomical, folk astronomical, and mathematical, geographical – relating to the determination of the qibla; and

(g)  prefers to refrain from citing the vast existing bibliography on the subject.

Worse still, he

(h) has settled on a nice-enough locality, Petra, as the focus of early Islam where in the early 7th century there were no Arabs, no Muslims, and no Jews, and, in brief, there was not much going on.

And worse than that,

(i) both his activities in a field which he does not master and his false conclusions have already contributed to somewhat dubious causes.

Figure 3. A schematic representation of the fallacy propounded by Cook & Crone. They observed that the earliest mosques in Egypt and Iraq appeared to be aligned toward a place in N. W. Arabia rather than toward Mecca. This, they wrongly thought, confirmed their theory that the origins of Islam were somewhere in N. W. Arabia rather than in Mecca. In fact, the mosques are aligned with the Kaaba in Mecca by means of astronomical horizon phenomena, namely, winter sunrise in Egypt and winter sunset in Iraq. The first generation of Muslims knew what they were doing with regard to mosque orientations and later generations over many centuries developed remarkable and more sophisticated means for finding the sacred direction. We moderns just have to learn how they dealt with the need to align mosques in the sacred direction toward the sacred Kaaba in Mecca. It is not something one can imitate or investigate with an iPhone, and no Google maps are going to help much.

Accurate mosque orientations towards Petra

To give credence to his Petra theory Gibson needs to rewrite the history of science, a subject about which he is singularly uninformed. He wants us to accept that when the first generation of Muslims expanded out of Petra (!) they knew all about astrolabes (!) and spherical trigonometry (!) and the like. When they wanted to build mosques around the world from al-Andalus to China facing the Kaaba in Petra they used these advanced mathematical techniques to calculate the pibla (my word) toward Petra and they were able to do this to within a degree or two. In fact, the ‘real’ Muslims used simple astronomical alignments to find the direction of the Kaaba, and there was no need for any mathematical system. (However, as part of the Graeco-Roman world, the Nabataeans long before the advent of Islam did have such devices as sundials.)

Mosque orientation before Gibson

Gibson’s claim about Petra deliberately ignores everything that modern scholarship has uncovered about the ways Muslims over the centuries have determined the sacred direction. His first book Qur’ânic Geography (2011) had not a single reference to any serious book or article on the qibla. His later works have been padded with a few references to my works but they deliberately omit any reference to five articles which presented an overview of what was known before Gibson appeared on the scene:

  • “On the astronomical orientation of the Kaaba” (with Gerald S. Hawkins) (1982);
  • “Astronomical alignments in medieval Islamic religious architecture” (1982);
  • “The orientation of medieval Islamic religious architecture and cities” (1995);
  • “The earliest Islamic mathematical methods and tables for finding the direction of Mecca” (1996); and
  • “The sacred geography of Islam” (2005).

For myself, I am fairly confident that Islam started in Mecca and Medina, and that all early mosques were deliberately aligned to face the astronomically- aligned Kaaba in Mecca. These orientations were achieved by the early Muslims with a considerable amount of success within the limits of their capabilities, mainly using astronomical alignments or building on earlier foundations that were inevitably also astronomically aligned. Later mosques were aligned either in qiblas calculated from the available geographical data using mathematical procedures, although the old procedures continued to be used.

In each major centre in the medieval Islamic world there was a palette of several qibla-directions accepted by one interest group or another. There might be a qiblat al-ṣaḥāba, a direction chosen by the first generation of Muslims who settled in that locality, usually an astronomically-defined direction, and favoured thereafter; there might be different directions favoured by the individual legal schools; there might be a different astronomically-defined direction that was favoured by some; and there could be two mathematically-determined qibla-directions, one based on approximate methods and the other based on an exact procedure. The modern qibla, based on accurate geographical data and derived by exact mathematical methods, is irrelevant to the investigation of the motivation behind the orientation of any historical mosque.

I consider it necessary to respond to Dan Gibson’s latest pronouncements for three main reasons:

  • People seem to forget that the sacred direction in Islam is not toward Mecca but toward the Kaaba in Mecca. There is a significant difference between facing an edifice that one cannot see but which one knows is astronomically aligned and facing a distant city. People need to be reminded of this, because what was obvious to a medieval mind is not obvious to us moderns. All of Gibson’s mosques are aligned toward the Kaaba in one way or another. Since the 9th century, when mathematical geography and mathematical methods became available, mosques have generally been aligned toward Mecca, usually, but not always, using mathematical methods. In major centres there was sometimes a palette of qibla-directions – covering as much as a quadrant of the horizon – used by different interest groups.
The Petra Fallacy

Figure 3. A brilliant geometrical construction for finding the qibla proposed by Habash al-Hasib, the leading astronomer of 9th-century Baghdad. The complicated modern formula can be derived directly from Habash’s diagram.

Without knowing this, it is somewhat precarious to try to explain an early mosque orientation.

  • The concept of the qibla is not just about legal scholars splitting hairs or mathematicians performing calculations or architects building mosques, it is about the millions and millions of faithful Muslims who for well over a millennium over a large part of this planet have exercised their utmost to pray towards the physical focus of their religion, a symbol of the presence of their God. This they do or have done in their mosques, but also in their homes and at work and whilst travelling. Also, in death the faithful are laid to rest in the same direction in which they have been praying during their lives. No Muslim needs some ill-informed Besserwisser to announce to them that they and their forefathers have been praying in the wrong direction for over a millennium and that they should have been praying towards a city in Jordan that has absolutely nothing to do with early Islam.
  • There are very few people – Muslims, non-Muslims and independents – who know anything about historical qibla determinations and even fewer who would be able to counter Gibson’s ‘new’, basically absurd theories which appear to rely on ‘scientific evidence’.
  • I am well aware of the potential damage Gibson has done / can do to our field. But more seriously, Gibson’s writings are guaranteed to contribute to Islamophobia amongst those who have no idea about the one and only civilization which really took orientations seriously for over 1,400 years.

Critiques of critiques

Most people are either numerate, which means that they like numbers and know how to handle them, or innumerate, in the sense that they don’t like numbers and shy away from them. Such people shudder when confronted with a direction such as 292°, because they have no idea that modern usage measures directions from 0° clockwise to 360° = 0°; these people might prefer to read 22° N of E. Now Gibson’s book is all about numbers, some real (measurements of mosques) and some irrelevant (MODERN directions of Petra and Mecca). Alas, most reviews of Gibson’s qibla extravaganza have been made by people not well versed in numbers.

In the acknowledgements to his Early Islamic qiblas Gibson thanks two scholars Rick Oakes and Ahmed Amine whom we shall mention below. (He also thanks one of the leading archaeoastronomers of the Near East, and of Petra, my colleague Juan Antonio Belmonte, who was even more surprised than I was to find his name in Gibson’s acknowledgements, for Gibson never mentions ethno- or archaeoastronomy.)

It is important to consider Gibson’s approach to mosque orientations in light of his methodology. For he uses MODERN geographical coordinates to calculate directions of buildings to Petra or Mecca or Jerusalem when those who erected these buildings did not have access to such coordinates. Nor did they have EXACT mathematical procedures for calculating directions of one place to another. So when Gibson writes that a given mosque faces (the MODERN direction of) Petra, not (the MODERN direction of) Mecca, this is not to be taken seriously. If I were to say this or that mosque faces Mecca not Petra, that might be equally absurd. If either of us says that a given mosque faces exactly Petra or Mecca so that those who built it must have had the geographical and mathematical knowledge to determine the pibla / qibla accurately, this would be nonsense. For mosques in the earliest period were laid out in directions that were not calculated at all.

In my first critique of Gibson’s Petra thesis I deliberately stated that I would not demonstrate his error for all of the mosques he had misinterpreted but would present enough examples to demonstrate that not only are his interpretations erroneous, but also that the whole idea of assessing the “errors” of medieval orientations by comparing them with MODERN directions is flawed. Some later commentators didn’t understand this.

Rick Oakes is an American scholar of theology concerned with the history of the Qur’ān and of early Islam. He has posted his evaluation of my critique of Early Islamic Qiblas on the blog of the International Qur’anic Studies Association (IQSA), an outfit based in Atlanta claiming to be “devoted to the study of the Qur’an from a variety of academic disciplines”. Oakes’ focus here is not on the science, mathematics, or astronomy that was (or, rather, was not) available to early Muslims, nor is it with how they could have pointed any of their earliest mosques in any particular direction. But rather, he naïvely focusses on the 17 mosques that Gibson says face (the MODERN direction of) Petra. He does not argue whether or not they were pointed toward (the MODERN direction of) Petra intentionally. He does not argue that Gibson’s mosque orientation measurements are accurate, but that these Gibson’s conclusions based on these orientations deserve confirmation or refutation. He overlooks my refutation of all of them, so he repeats this appeal from his non-critical review of Gibson’s first book.

Oakes begins by omitting that I first published my review of Gibson on my own website and later on the Muslim Heritage site. He writes that I “revised” my review after a petty response by Gibson, when, in fact, I just removed a comment about his missionary connection. Oakes identifies five mosques whose orientations I did not even mention: the Masjid al-qiblatayn in Medina and four other very minor mosques I had never heard of. He seems so convinced about Gibson’s finding that 17 early mosques point toward (the MODERN direction of) Petra that he challenges other scholars to offer better explanations than that this was deliberate. It all becomes a game: who gets it right and who gets it wrong. Oakes correctly observes that my explanations of why the mosques in Amman, Fustat, Jericho, and Khirbat al-Minya (only these!) are preferable to Gibson’s explanation that they point toward (the MODERN direction of) Petra. While he is correct in mentioning that I wrote that the Sanaa Mosque points toward (the MODERN direction of) Petra, he missed the fact that this does not mean that it was deliberately laid out to face Petra: I also said that the axis of the Mosque was ‘parallel’ to the main axis of the Kaaba, so that the qibla-wall is ‘parallel’ to the SE wall of the Kaaba.

In brief, Oakes has unfortunately overlooked what I wrote about the absurdity of using MODERN directions to investigate orientations of buildings that were built well over 1,200 years ago and the folly of ignoring cardinal and solstitial directions in interpreting orientations that were laid out toward astronomical horizon phenomena or on pre-Islamic foundations that were cardinally aligned. He is apparently ready to believe Gibson’s claims about Petra if somebody can confirm them.

Suggestions for future research

Fortunately, nowadays one would not have to travel the length and breadth of the Muslim world to have a new look at mosque orientations. These are not “theories” about early mosque orientations, these are simply suggestions for future research. What concerned investigators might want to do in the future with the major mosques of the medieval period (7th-15th centuries) is the following:

(1)  determine which mosques were built on the authority of the Prophet or his Companions;

(2)  determine which mosques were built on the foundations of, or in line with pre-Islamic religious architecture which happened to be cardinally aligned (such as in Jerusalem and Damascus);

(3)  determine which mosques were built according to the street-plans of pre- Islamic cities which happened to be solstitially aligned (such as Córdoba, Tlemcen, Tunis, Kairouan);

(4)  determine which mosques were built toward winter sunrise (taken as one qibla-direction from Egypt to al-Andalus), and toward winter sunset (taken as one qibla-direction from Iraq to Central Asia), or toward some other astronomical horizon phenomenon;

(5)  determine which mosques face more or less due south in Jordan and Syria;

(6)  determine which mosques face due west in India and due east in N. Africa; and

(7)  determine which mosques more or less due north in Yemen and E. Africa.

Mosques which do not conform to these norms can possibly be explained by means of information on the local qibla in treatises on folk astronomy and sacred geography (astronomically-defined directions) or treatises on mathematical astronomy (qiblas calculated from available medieval geographical data using exact or approximate mathematical methods). Local topography or hydrography may also have played a role. In all such investigations, no conclusions should be drawn based on qibla-directions calculated from MODERN geographical data using some kind of EXACT mathematical procedures. Also, measurements and calculations to the nearest degree are adequate for investigative purposes; any attempt at greater ‘accuracy’ is unrealistic.

To any interested parties, I would recommend looking at the five articles which I mentioned above, not least my article on the earliest mathematical methods and tables for finding the qibla. I am confident that such simple approximate methods had far more influence in mosque alignment than any complicated exact methods and tables. But one cannot use any of these without knowing what geographical coordinates were available over the centuries. The complexity of Islamic geographical tables giving longitudes and latitudes, and the basic reference work by E. S. & M. H. Kennedy, Geographical coordinates of localities from Islamic sources (Frankfurt, 1987), presents 14,000 sets of longitudes and latitudes from some 80 Arabic and Persian astronomical and geographical sources.

In investigating the orientation of a historical mosque it is important to take into consideration the original surrounding street-plan and the various qibla– directions that were favoured in that region at the time. Without such information it is not a little arrogant to suppose that one can make any sensible pronouncement regarding the reason behind the orientation of an edifice that was built over a millennium ago. Woe betides anyone who claims to explain any medieval mosque orientation without realizing how complicated is the subject of orientations.

Notes added in September 2020:

If Dan Gibson had claimed that his investigations revealed that the earliest mosques faced precisely toward Mecca, I would have immediately pronounced that this was absurd, not least because he would have been claiming that they face the modern direction of Mecca for each locality. Besides the sacred direction in Islam is toward the Kaaba, not toward Mecca. Modern directions from one place to another were available only from the 19th century onwards. 

No precision was to be expected in the 7th and 8th centuries, precision of the kind we moderns take for granted. But the first generations of Muslims had all the technical knowledge they needed to face the Kaaba, because the Kaaba is astronomically aligned, that is, the major and minor axes of its rectangular base face significant astronomical horizon phenomena, and its corners roughly face the cardinal directions. So to face the Kaaba in any locality, they had only to face the direction in which they were standing right in front of that part of the Kaaba which was associated with that locality. Modern people would be perplexed if they would be asked to face a building that they cannot see; the first generations did not see this as a problem.

As it is, Gibson’s data show that these mosques face precisely towards Petra, a very nice place which has, however, nothing to do with the history of early Islam. Gibson’s Nabataean Muslims could never have accurately determined the pibla toward Petra from places between al-Andalus and China. We have to look elsewhere in order to investigate the orientations of these mosques. In brief, they do not face Mecca, and they do not face Petra, but they do face the Kaaba, within the limitations of what the Muslims knew in the 7th and 8th century. The orientations of their first mosques are not “careless” or “inaccurate”, as has been claimed by many a historian of Islamic architecture. But they have a lot to teach us, including that they do not face anywhere specific.

Dan Gibson is not to be held back from his claims, to the extent that he has even included mosques built on pre-Islamic foundations and finds that they face Petra too. I am reminded of a certain fellow who measured the orientations of numerous medieval cathedrals in Europe and found that they were facing Mecca and concluded that they must have been built as mosques. On the subject of churches, it is an absurd claim that medieval churches face east or face Jerusalem. If one measures the orientations of French cathedrals, for example, one comes up with a span of 100° on the eastern horizon.

The claims of Dan Gibson have been greeted with enthusiasm by those who seek to denigrate Islam and distort Islamic history. Serious reactions to them from the scholarly world have led to laughable rhetoric on the part of a fundamentalist outfit concerned to attract Muslims away from Islam.

The interested reader may care to consult the following works to gain some control over the subject and to better understand the folly of the claims of Gibson and the perverseness of his advocates. Most of the nonsense they attribute to the present author results from their inability to understand what he wrote to counter Gibson’s claims and their readiness to distort it.

(a) “On the orientation of the Kaaba” (1982). Without an understanding of the layout of this sacred edifice, one cannot begin to explain the orientation of any early historical mosques. For this and other reasons, Gibson does not mention the orientation of the Kaaba ( A new book on many aspects of the historical Kaaba by a well-informed Western scholar, Simon O’Meara, has just been published (

(b) Astronomical orientations were widely used in ancient times, including in Arabia and especially by the Nabataeans in Petra. See J. A. Belmonte et al., “Equinox in Petra: Land‑ and skyscape in the Nabataean capital” (2020 – ( Gibson does not like astronomical orientations, even though they were used by his favourite Nabataeans (long before the advent of Islam). His claims about Petra as the cradle of Islam are not taken seriously by scholars of Nabataean Studies.

(c) “Finding the qibla by the sun and stars – Islamic sacred geography” (2019), a survey of some 50 medieval sources, documenting some 20 different schemes of sectors of the world about the Kaaba together with their qiblas as the same astronomical directions one would be facing if one were standing directly in front of the Kaaba ( Gibson claims that “dumb” people used these astronomical orientations to find “sloppy” orientations without realizing that they were the means used by the first generations of Muslims to carefully find the direction of the astronomically-oriented Kaaba. In fact, they were smart to use these directions.

(d) “The wind-catchers of medieval Cairo and their secrets” (2020), where the ‘secrets’ relate to the layout of the astronomically-aligned Fatimid city alongside the Roman Nile-Red Sea Canal as well as the various qiblas used in the city from the 7th century onwards ( – text) / (– illustrations). Gibson does not realize that there are entire Muslim cities facing the Kaaba, but that this can only be understood if one knows what people thought was the direction of the Kaaba, which was not the modern direction of Mecca, and most certainly not the modern direction of Petra!

(e) “The enigmatic orientation of the Grand Mosque of Córdoba” (2019), on the way in which the Mosque was laid out according to an astronomically-aligned Roman suburban street-plan, the admittedly curious orientation of the Mosque being later confirmed by the astronomical directions proposed in various medieval schemes of sacred geography, that is, divisions of the world around the astronomically-oriented Kaaba ( All major mosques of the Maghrib face the same direction for the same reason ( Gibson in one of his most imaginative moments proposed that these mosques are all parallel to an imaginary line between Petra and Mecca.

(f) “al-Bazdawī on the qibla in early Islamic Transoxania” (1983/2012) on the orientations of mosques in Samarqand by a well-known 11th-century judge ( Contrary to what Gibson thinks, this judge knew what he was talking about.

(g) “Bibliography of books, articles and websites on historical qibla determinations” (2018), listing some 150 items ( Gibson’s first writings on the qibla referred to none of this material. Later he consulted King’s article “Qibla” in the Encyclopaedia of Islam but not King’s article “Mekka as centre of the world”, in which Islamic sacred geography is presented for the very first time (

It is not unfair to say that Gibson has no idea about historical qibla determinations, but he is considerably better informed than the hapless souls who seek in a series of some 18 and more videos to promote his nutty ideas about Petra and to demolish his detractors (

~ Click here for the full article ~

by David A. King,
Professor of the History of Science,
Johann Wolfgang Goethe University, Frankfurt am Main

3.2 / 5. Votes 9

No votes so far! Be the first to rate this post.


Copy link
Powered by Social Snap