Mecca before Islam: 2) Macoraba

The second in a series on pre-Islamic sources adduced for Mecca.
Previous: 1) Diodorus and the Kaaba.

“Macoraba: 73° 20′ 22°.”
— Ptolemy, Geography, §6.7.


Claudius Ptolemy was a Greek writer living in Alexandria in the second century CE. He was the author of several works, the most influential being his astronomical Almagest; but his Guide to Geography, completed between 141 and 147, would also prove highly influential in Europe and the Middle East. It consisted of a theoretical introduction, a list of notable places across the known world, and some accompanying maps. In order to preserve the maps’ accuracy when they were copied and recopied, Ptolemy gave coordinates for the place names, which he had calculated with some accuracy: a Ptolemaic map is distorted, but recognisable. His primary source was Marinus of Tyre, who wrote between 107 and 114.

Ptolemy has a section on the western and southern regions of Arabia, which is useful for those of us studying the region’s history: the peoples of ancient Arabia did not leave us many written sources, so foreign sources like Ptolemy help us track the rise and fall of towns, powers, religions and trade routes. Some of Ptolemy’s locations in Arabia are known to us, like Yathrib (Lathrippa), present-day Medina; some are unknown or disputed. Among the names is Macoraba – Ptolemy spells it Makoraba in Greek, but Latin translations prefer Machoraba and Macoraba – which Ptolemy puts in the west of the Arabian Peninsula.

There is a consensus in academic scholarship that Mecca is Macoraba. The coordinates put it roughly in the right place, and the name seems roughly correct. Several etymologies have been proposed, but the preferred solution is that it comes from an Old South Arabian word like *mikrāb, with the meaning ‘temple’. Macoraba was therefore a noteworthy centre of pre-Islamic religion as far back as the second century CE. When you encounter Macoraba in scholarly literature you are quite likely to find this etymology, and extremely likely to find the identification with Mecca.

Of course a consensus does not have to be unanimous, and there have been dissenting opinions. The most prominent so far was from Patricia Crone (d. 2015). In a brilliant, contentious book, Meccan Trade and the Rise of Islam (1987), Crone devotes a few pages to the evidence for Mecca in ancient literature, with negative conclusions: “The plain truth is that the name of Macoraba has nothing to do with that of Mecca, and that the location indicated by Ptolemy for Macoraba in no way dictates identification of the two.”

Meccan Trade had a strong impact on Early Islamic Studies, but Macoraba remains a staple of academic writing on ancient Arabia. The reason, I think, is not that our interpretation is particularly sound or explanatory, but that Macoraba has become so familiar that we don’t think to reexamine it. Macoraba has been part of our thought-world for a very long time. Crone responds to literature going back to the early twentieth century, but this blog will show that Macoraba-as-Mecca goes back to the seventeenth, by way of a fire in a Vatican apartment, an Israelite invasion of the Hijaz, and E.M. Forster quietly apologising for his grandfather’s visionary nonsense.

It is a wonderful story: the further I dug, the more I wanted to keep going. So many old Orientalist works have been digitised and freely distributed that I was able to follow new leads entirely online. There was never a citation I couldn’t follow: the trail stopped, but never went cold. Still, I hope there is more to discover. I will integrate appropriate feedback into this post, so I expect it to grow a little over time, and I think it deserves a definitive treatment in due course:  I’m planning to expand this preliminary work into an article for peer review.



Where was Macoraba?

If we suppose that Macoraba is Mecca, there is a slight problem with its coordinates. Ptolemy puts it southeast of Yathrib; Mecca is southwest. Even before 1800 Konrad Mannert noticed that Macoraba was too far from the coast and offered a solution: Ptolemy’s sources knew Mecca from the overland caravan route, and had never approached the town from the coast. Of course, we don’t know where Ptolemy’s information ultimately came from; but even this solution may be too elaborate, because in general it seems that Ptolemy had more trouble calculating longitude than latitude, meaning that his towns are more accurately positioned north-south than east-west.

This had decisive consequences for his geography of Arabia. Dūmat al-Jandal (Dumaitha) is indeed further north than Taymā’ (Thaima), which is further north than al-Ḥijr (Egra), Yathrib (Lathrippa), and Najrān (Nagara); but then Ptolemy puts Najrān way out east in the middle of the peninsula. The overall effect is to push towns away from the coast, crowding the heart of the peninsula and practically erasing the Empty Quarter of harsh desert in the southeast. Under these constraints, Macoraba’s location with respect to Mecca may be considered within a margin of error.

But we should be cautious. This margin of error is not itself evidence that Macoraba is Mecca; it merely opens the door for investigation. As this blog series will show, some of the ancient names that have been associated with Mecca are most likely in the region of the Sinai and the Gulf of Aqaba, or down by Oman and the Hadramawt. Macoraba has the virtue of at least being placed in the Hijaz, but imprecisely. We should heed Patricia Crone’s advice: “Naturally, Pliny’s longitudes and latitudes are inexact; but if they are inexact, one cannot identify places on the basis of them alone.”


From Hugh Kennedy (ed.), Historical Atlas of Islam, s.v. “Arabia According to Ptolemy”. Makoraba is in C3. Click to enlarge.


We are left with the problem that Macoraba is a different word from Mecca. It may seem odd to phrase it that way, but it needs saying and saying again: Macoraba is a different word from Mecca, and that is a problem if we want to identify them. It is not good enough to say that the first half of Macoraba sounds a bit like Mecca. We should either explain the difference or cut Macoraba loose. And there is a very long tradition of attempts to explain it.


Macoraba and early Islam

Our story should begin with early Muslim literature. Since the medieval scholars had access to Ptolemy’s Geography and many subsequent works, and since they took a great interest in the history and geography of Arabia, we might expect them to comment on Macoraba. Apparently they didn’t. Earlier studies on Macoraba have not recruited medieval literature as evidence, and my own brief investigations have found nothing. I will publish corrections here if any are forthcoming.

The Arabic geographer Yaqut al-Hamawi (d. 1229) does quote Ptolemy on the location of Mecca, which should tell us whether or not he identifies it with Macoraba. Strangely, though, the coordinates he attributes to Ptolemy – 78° 23° – don’t line up with Macoraba, or anywhere else in Ptolemy’s Geography at a glance, and they seem to put Mecca even further east than Ptolemy puts Macoraba. Meanwhile Yaqut has no entry for a place called Macoraba. I understand that Arabic geographers developed their own coordinates; perhaps they converted Ptolemy’s to their new standard, but I’m not competent to investigate that. The difference may also have to do with the Geography’s transmission in the Islamic world in contrast with how it was transmitted in Europe.

Those who wish to identify Macoraba as Mecca might help their case if they start to address questions like this; until they do, we can put them aside.


Mecca the Great (Makkah *al-Rabbah)

From the Renaissance until the middle of the seventeenth century, European scholars knew the name Macoraba from Ptolemy, but they didn’t know what to do with it. It was included in lists of geographical data simply because they had it to hand and it was very, very old, but it had no apparent significance beyond that.

The earliest on record to identify Macoraba with Mecca is probably Samuel Bochart (d. 1667), a Protestant churchman in France who, like many early Orientalists, applied his knowledge of Near Eastern languages to the study of the Bible. He is best known for his Sacred Geography, the first volume of which, Phaleg (1651), is a pioneering study of how Noah’s people dispersed and resettled after the Tower of Babel fell. Bochart compiles heaps of data from ancient sources. It is in one of his geographical lists that we find a short but significant note: “Macoraba Ptol. Meccha rabba, that is, ‘great’.”

In Classical Arabic, Bochart’s hypothetical name should be Makkah *al-rabbah. This does not sound terribly different from Macoraba. The obvious problem here is that the adjective rabb(ah) does not exist in Classical Arabic; evidently Bochart has borrowed it from Hebrew and Aramaic. There were other Arabian languages in the Hijaz when Ptolemy was writing, and it is not unthinkable that one of these should have used rabb(ah) as an adjective in this way, but the burden of proof has not been met by supporters of Bochart’s hypothesis. As Patricia Crone observes, we do not even have a Classical Arabic equivalent like Makkah al-Kubrā to support the idea that Mecca was ever called ‘great Mecca’. Bochart’s hypothesis is speculative in the extreme.

Nevertheless, the identification of Macoraba was well received. Around 18 years later, Leiden University’s Jacob van Gool reproduced it in his commentary on al-Farghani’s Compendium of Astronomy (1669): Mecca. A town built in a valley, latitude 21° 40′, where Ptolemy locates Macoraba”. Van Gool’s latitude corresponds neither with Ptolemy’s Geography as we have it, nor with Yaqut, but it is one of the latitudes that medieval Muslim geographers calculated for Mecca (not Macoraba). Presumably he recruited Mecca’s latitude from another work and synthesised with Bochart’s claim that Mecca is Macoraba. He does not credit Bochart directly, but he clearly respects his work: elsewhere he describes him as “most learned” and Phaleg as “preeminent”. Van Gool’s comments on Macoraba were then cited by Caspar Calvör, a Lutheran theologian in Lower Saxony, in his study On the World’s Diverse Religions (1705).



Mecca the Direction of Prayer (Miḥrāb)

An early mihrab at Qasr al-Hallabat, Jordan.

In proposing a connection between Mecca and Macoraba, Bochart set an etymological challenge: why is Macoraba called Macoraba, when we know it as Mecca? For over 75 years the challenge went unanswered. The first to take it up was Giuseppe Simone Assemani (al-Sim‘ānī, d. 1768), a Lebanese priest and Orientalist whose extensive work with Near Eastern manuscripts was patronised by the Vatican. His magnum opus was a Library of the East, in which he hoped to publish many of these manuscripts for the first time; sadly much of his progress was destroyed by a fire in his home, and only four volumes ever came out. One of these was a book On the Syriac Nestorians (1728), in which Assemani mentions Macoraba.

Assemani first acknowledges Bochart’s opinion, then offers an alternative: “…Macoraba, that is, Mecca rabba or great Mecca as Bochart thinks; or perhaps محراب Machrab, temple, because there was a shrine of the Arabs there, inside which was the idol Beccha”.

The last remark alludes to the name Bakkah, which according to the Qur’an was the first ever temple. Early Muslim scholars identified Bakkah with Mecca, but came up with several hypotheses to explain the word, none conclusive. Early Orientalist scholarship on the other hand speculated that Bakkah was at first the name of an idol, perhaps the god Bacchus of Graeco-Roman mythology. In any case, Assemani agrees with the Arabic historical tradition that Mecca was a centre of pagan worship before Islam. Given this relationship between Mecca and worship, Assemani proposes that Macoraba is from the word mihrab (miḥrāb), which he transcribes rather unhelpfully as “Machrab”.

In Islam a mihrab is a niche in the wall of a mosque signalling the direction of prayer; I do not follow Assemani’s reasoning in translating it ‘temple’. Moreover it is not clear that the mihrab has a history going back to the days of Ptolemy. As an architectural feature of the mosque it seems to have developed after the Muslim conquests, perhaps as late as the eighth century; and the word itself is used in early Arabic poetry and evidently in the Qur’an to mean quite different architectural features. Even if we had reason to think there was a mihrab in Mecca centuries before Islam, the idea that the town should have been named after a niche in a wall tests the imagination. Finally, it seems unlikely that Arabic should be represented by Greek k: in later evidence, Greek writers do not even try to approximate , instead for example representing the name ‘Abd al-Raḥmān as Abderaman. Assemani’s solution does not seem any more helpful than Bochart’s.


Macoraba goes mainstream

These arguments should not convince specialists today: we are in a position to recognise one as a chimera and the other as anachronistic. But in the rising sun of scientific history, they cast a long shadow.

In 1739, Makkah rabbah found its way into the first ‘modern’ encyclopaedia in German: the boldly titled Great Complete Encyclopaedia, published by Johann Heinrich Zedler. Zedler also collaborated in a Historical-Political-Geographical Atlas published by Johann Samuel Heinsius, and it is no surprise that his material on Makkah rabbah is copied verbatim in the new article on Mecca (1747). Both of these works were written by teams of scholars in collaboration, anonymously, so we don’t know who wrote the material on Makkah rabbah; a specialist in early German Orientalism might be able to guess. The Atlas is a translation and expansion of a French Geographical Dictionary by Antoine-Augustin Bruzen de la Martinière, whose entries on Mecca and Macoraba (1735) do not connect the two names.

That is to say, we can probably thank a nameless German scholar in the 1730s for introducing Macoraba-as-Mecca to the encyclopaedia genre. Whoever it was, they did not cite Bochart. They may have read him directly or encountered his ideas through an intermediary; perhaps Assemani, but since they neglect the latter’s mihrab argument, that seems unlikely. It is not impossible that the author invented Makkah rabbah independently, but we cannot know.

In 1799 the omnivorous Prussian historian Konrad Mannert (d. 1834) published an entry on Macoraba in his Geography of the Greeks and Romans, where he commented, “I don’t know whether the name arose from Mecca rabba (great Mecca) or Machrab (a temple)”. It is clear that Mannert depends on Assemani for this detail, since he reproduces the unhelpful Latin transcription of ‘mihrab’ and the implausible definition ‘temple’; elsewhere he does explicitly cite Assemani citing Bochart.

Mannert’s Geography turned out to be a handy reference for scholars working in German, and through them Macoraba reached a wider audience. In 1824 the historical geographer A.H.L. Heeren identified Macoraba as Mecca off-hand, and although he did not provide a citation for this, he did cite Mannert in many other places. It was through the English translation of Heeren’s studies that a Scottish Orientalist called William Muir acknowledged Macoraba as Mecca in his highly influential Life of Mahomet (1858). To take another example, the University of Berlin’s geographer Carl Ritter cited Mannert in 1846 as the authority for Macoraba, and it was in part through Ritter’s work that Reinhart Dozy, a Dutch Orientalist, approached the question; we shall come back to Dozy below. Mannert was also a major influence on Friedrich Sickler’s Handbook to Ancient Geography (1824), which equates Mecca with Macoraba. (Sickler’s purported source for this, a page in Joseph Eckhel’s catalogue of antiquities, is actually about Ptolemy’s Moka in the Sinai.)

Perhaps most significant was Leipzig University’s classicist Albert Forbiger (d. 1876), who wrote an encyclopaedia entry on Macoraba, which directly cited Mannert for his two etymologies. This too was published in 1846 as part of August Pauly’s Encyclopaedia of Classical Antiquity, a landmark in modern historical studies. Through Mannert’s Geography and then through Forbiger’s encyclopaedia entry, Macoraba-as-Mecca enjoyed wide circulation in academic scholarship by the mid-nineteenth century.


Mecca the Warlike (Muḥāribah)

Seemingly ignorant of this German trend was an English churchman called Charles Forster (d. 1871), author of the unappetising Mahometanism Unveiled (1829), which railed against Edward Gibbon’s secular historiography. He wrote a few quasi-historical studies of the Near East, including a Geography of Arabia (1844), where he discusses Macoraba. He doesn’t seem to know Mannert and he was slightly too early to have read Forbiger. Since he does make extensive use of Bochart’s Phaleg, it stands to reason that he owes his knowledge of Macoraba to Bochart. Nevertheless, Forster does not address Bochart’s Makkah rabbah, and he may not even have been aware of Assemani’s Mihrab. Instead he comes up with a new and highly subversive etymology.

Forster proves imaginative but incautious. His argument starts with Pliny the Elder (d. 79), who records the opinion that the Cerbani people of southwest Arabia “excel in arms”. On this basis, our good friend Samuel Bochart speculated that the name Cerbani derives from a Phoenician word for warlikeness; in Hebrew, which is related, the word for battle is qerab. Forster approves Bochart’s reasoning, but prefers a derivation from the Classical Arabic ḥarb, meaning ‘war’. Going a step further, he posits that the Cerbani are identical with the Arabian tribe of Ḥarb, who are themselves warlike (he says).

Forster believes that ancient Mecca belonged to the Ḥarb, though the evidence is circumstantial at best. He agrees with Bochart that Pliny’s Cerbani are identical with the Carbae mentioned in Agatharchides’ description of the Red Sea coast; but from context both may be easier to situate down in the Yemen, near the Sabaeans, rather than up in the environs of Mecca. Meanwhile the Ḥarb only enter the historical record with Islam, centuries after Pliny describes the Cerbani. They have had a strong presence in the Hijaz since then, but there is nothing tying them to ancient Mecca. Nevertheless Forster assumes that Mecca was their town; indeed their capital. He also assumes that Mecca is Ptolemy’s Macoraba. Consequently he derives it from the same root as Ḥarb: Macoraba is muḥārib(ah), warlike. ‘Mecca’ is then “an idiomatic abbreviation” of this original name.

As we have seen, the representation of Arabic by Greek seems unlikely. But that is the least of Forster’s problems in his spindly thread of inference. His even grander assertion that the Cerbani/Ḥarb are the biblical Qedar is equally groundless. His grandson, the novelist E.M. Forster, would modestly say the old rector “had the disadvantage of resting on imperfect research.” We might instead say that Forster’s book is a tribute to works like Samuel Bochart’s Phaleg, weaving theology and history with great imagination but not much rigour.


Mecca the Great Battlefield (Makkah Rabbah)

While Forster dallied with Bochart, the Orientalist community was already starting to forget him. Some twenty years later when the radical Dutch scholar Reinhart Dozy (d. 1883) approached Macoraba, he did so through recent geographies in German, including Carl Ritter’s, both of which – thanks to Mannert – took for granted that it was an ancient name for Mecca. What these geographies lacked was any attempt at etymology. Dozy came up with one for The Israelites at Mecca (1864), his outré application of biblical criticism to the study of Islam’s origins.

Dozy considers the possible Arabic roots for Makkah and quickly rejects them: the name should therefore come from outside Arabic. He relates that Orientalists have sometimes rejected Macoraba-as-Mecca because the Muslim historical tradition leads them to think the town of Mecca was founded too late for Ptolemy to have known it. Dozy agrees, but with a spin: the name Makkah existed before the town. Macoraba is the Hebrew makkah rabbah, ‘great slaughter’, as in 2 Chronicles 13:17 and Numbers 11:33. Mecca was built on the site of a great battlefield. “No wonder the Arabs, who didn’t know Hebrew, couldn’t explain the name!”

For Dozy, that battlefield is the one in the Tanakh where the Simeonite tribe of Israel defeated the Canaanites. The Tanakh reports that they “destroyed” (yaḥarimu) the Canaanites and their town, which is punningly called Ḥormah. The town is unidentified, but Dozy relates it to the Arabic ḥaram, the sanctuary at Mecca. He argues therefore that the Simeonites conquered the land where Mecca would eventually grow. Through this narrative he is able to explain certain parallels between Muslim ritual and ancient Israelite practice as the residues of an Israelite conquest.

Needless to say, Dozy’s book did not convince his Orientalist colleagues, though it did scandalise many Jewish readers. It is curious that Dozy returns to Makkah rabbah, so redolent of Bochart’s, but through such a different schema. Though perhaps this should not surprise us: Bochart and Dozy both worked with Hebrew, using the Tanakh as a foundational text for their wider researches into Near Eastern history. Independently, it seems, they both saw rabbah in Macoraba because their training had primed them to think Hebraically.


Mecca the Capital (Makkah Rabbat Banī Malik)

Dozy’s conclusions may have been derided, but his comments on Macoraba caught the attention of Aloys Sprenger (d. 1893), whose career as a researcher, translator and principal took him between Europe and northern India. In his Ancient Arabian Geography (1875), Sprenger summarises Dozy’s Makkah rabbah and gives a rejoinder: the element –raba has the sense of ‘capital’, as in Rabbat-Moab, the ancient capital of Moab. We might add a second example, Rabbat-Ammon for the Ammonites.

But in both of these, the formula is Rabbat-X, capital of the so-and-so. That does not seem to be true for Macoraba, which is simply ‘Mecca the capital’. Sprenger anticipates this: Mecca belonged to the Banu Malik, and the town’s full formal name is recorded in Pliny’s Natural History (§6.157) as Mariaba Baramalacum, *Makkah-rabba(t?)-banī-Malik.

It is worth pointing out, first of all, that Sprenger has changed the pattern he is adducing from Rabbat-X to Town-Rabbat-X in order to fit Mecca into Macoraba and Mariaba. This is on top of a weakness in his source: Mariaba Baramalacum suffers from inconsistent spelling across manuscripts of Pliny. The Loeb edition has Maribba Paramalacum, which the editor moreover interprets as two separate names; likewise an English translation independently yields “Marippa, …belonging to the Palamaces”. Sprenger is inferring from an unstable text. And even if we take Pliny for granted, his description takes us down the Red Sea coast past the Minaeans before reaching Mariaba Baramalacum; since we know the Minaeans were in northern Yemen, that would put Mariaba Baramalacum well south of Mecca and Macoraba.

Sprenger does not compensate for the likely pronunciation of the feminine marker in the construct Rabbat-X. He does not explain in this section why Arabic banī should be represented by Greek bara-, but in a later section on the Banu Malik he explains that this is the Aramaic bar, meaning ‘son of’. Surely, however, the plural of bar should be bnay, so that would be no explanation, even if we had reason to place an Aramaic name in Mecca. In short, Sprenger adduces a Jordanian pattern which does not quite fit a Hijazi name, so he changes it to suit one uncertain spelling of an ungrammatical foreign name in the Yemen. Such is the allure of Macoraba.


Mecca the Temple (*Mikrāb)

So far speculation over Macoraba has worked with Hebrew, Arabic and Aramaic, all of which were available to Orientalists before the mid-nineteenth century. New possibilities opened in 1841, when two German scholars independently published their research deciphering another ancient language: Old South Arabian. This language, or rather a small family of closely related languages, was spoken in the Yemen, and survives mostly in inscriptions.

The first to apply Old South Arabian to Macoraba was probably  Eduard Glaser (d. 1908), a Bohemian archaeologist whose work in the Yemen was pathbreaking. He wrote an Outline of Arabian History and Geography (1890) where, with easy confidence, he asserts that Mecca is Macoraba, and that Macoraba is derived from a word like *mikrāb, meaning ‘temple’. In fact Glaser goes further: he relates Macoraba to Mochorbae, a port mentioned by Pliny. He proposes that Mochorbae is modern Jedda, and was named after Macoraba, which it served as a port.

We can immediately dismiss the second claim, because Pliny’s Mochorbae is not in the Hijaz, but in the southeast between Oman and Hadramawt: another warning not to put our etymological fancies ahead of the text. Still, Glaser’s etymology for Macoraba deserves a response. His *mikrāb is obviously not Classical Arabic, but an Old South Arabian word for ‘temple’, MKRB(N) in the inscriptions; compare Ethiopic mekuerāb, also ‘temple’. To that extent – if we assume that Macoraba is Mecca and if we assume that Mecca was a holy site going back centuries – the name might maybe fit. The trouble is then to explain how a name with an Old South Arabian derivation got planted in the Hijaz.

Such an explanation was attempted by Martin Hartmann (d. 1918), who taught Arabic and wrote on Islamic Studies in Berlin after a career in the diplomatic service. One of his books was The Arabic Question (1909), which his biographer has called “a great grab-bag of archaeological, philological, and historical ruminations on Arabia”. Hartmann observes that, according to early Muslim lore, tribes periodically came up from the Yemen and established themselves elsewhere in the peninsula, including Mecca. The implication is that Macoraba-as-Mecca was once controlled by people who spoke Old South Arabian.

Further research has shown that the migrations are more than legendary, but they began in the third century, long after Ptolemy. Moreover, from their inscriptions in the peninsula and from their names recorded by outsiders, it seems that the migrants did not speak Old South Arabian, but something nearer to Classical Arabic. If we want an Old South Arabian derivation for Macoraba, we should want to see evidence – more than inference – that the language was actually spoken in the region.

Another objection has been posed quite recently by Mikhail D. Bukharin of the Russian Academy of Sciences. He says that Macoraba is unlikely to be South Arabian *mikrāb, because the k of Semitic languages is rarely represented by the letter kappa (k) in Greek. If I understand correctly, the Semitic was perceived by Greek speakers as their aspirated letter chi (kh), so that for example the Arabic name Malik would be transcribed in Greek as Melekh, not Melek. This would be another reason to second-guess our assumption that Mecca is Makoraba: why not *Makhoraba? At the very least it should serve as a warning to consult the experts in historical and comparative linguistics when we go hunting for etymologies. Bukharin’s own suggestion that Macoraba is derived from the Arabic for ‘West’ (maghrib) remains speculative, but it also reminds us how many possibilities of derivation could lie beyond Mecca, if only we could think outside the box that Bochart built.

Interestingly, Glaser’s idea has caused some confusion, because his *mikrāb temple looks so similar to the “Machrab” temple proposed by Assemani and popularised by Konrad Mannert. Even such an erudite commentator as Adolf Grohmann (1928) conflates them. As early as 1911 Martin Hartmann was goaded into spelling out the difference in a short letter to an Orientalist journal. He correctly intuits that “Machrab” was an attempt to transcribe miḥrāb, and correctly dismisses it. By comparison, he says, the virtue of Glaser’s *mikrāis that it is based on South Arabian inscriptions, which Mannert – and then Forbiger – could not have had access to. The letter is early evidence of Mannert’s lasting influence on the wider Orientalist community, and of the growing attention paid to Macoraba: no longer an antiquarian curiosity but a site of philological debate.

Thanks to Glaser, Hartmann and other commentators at the turn of the century, *mikrāb became the favoured etymology for Macoraba. It seemed to bind Mecca’s antiquity, the success of its temple, the name given by Ptolemy and (by abbreviation) the name that Muslims knew. Scholars may also have enjoyed recruiting Old South Arabian to the study of early Islam, just as some scholars in our day recruit Syriac, without fully appreciating the limits of such an approach.


Mecca Brings us Closer to the Gods (Muqarribah)

Glaser’s *mikrāb remains the most popular derivation, but it was not the last. Over sixty years later the Iraqi scholar Jawad Ali (d. 1987) offered a variation on the Old South Arabian theme in his seminal Abridged History of the Arabs before Islam (1951–3). This etymology seems to be the most recent of note, and it is the first to have been proposed in Arabic.

Ali’s argument begins in the Yemen. We know from inscriptions that a major official was the MKRB, maybe pronounced *mukarrib. Scholars writing after Ali have suggested that he was a high king; Ali seems to think he was an arbitrator of the sort we find in the Hijaz before Islam. Either way, Ali speculates that the *mukarrib’s justice was the justice of the gods, and he brought the people closer to the gods. In Classical Arabic, a person who brings-things-closer is a muqarrib. Ali posits that this is the meaning of Old South Arabian *mukarrib.  He then turns his attention to Mecca. Like the *mukarrib, Mecca brought people closer to the gods: it was a religious sanctuary before Islam. Consequently, he says, Mecca gained the epithet al-Muqarribah, the place that brings us closer to the gods.

He points out that holy sites do attract epithets: in Arabic Jerusalem is known as ‘The Holy’ and ‘The Holy Sanctuary’. That much is obvious, but there are grave problems with the rest of Ali’s argument. For one thing, even if we grant that Macoraba is Muqarribah, the shift in vowels is so stark that we should want an explanation. For another, the Arabic root for bringing-things-closer is the same in Old South Arabian: QRB, not KRB. It is uncertain what the word *mukarrib literally meant, but the root KRB suggests it had something to do with holy rites. The word muqarribah is most unlikely to be relevant.

Oddly enough, the Classical Arabic lexicographers tell us that their root KRB can indeed signify approach and closeness. This does not vindicate Ali’s argument, because it remains entirely speculative. He assumes but does not demonstrate that Macoraba should be identified with Mecca. He does not try to show that Mecca, or any other holy site, was ever known as a place that brings-things-closerAli starts with the assumption that Macoraba is Mecca, and then fashions an explanation ad hoc to bridge those two different names.


Mecca and the Holy Hill (Makrūb)

So much for etymologies; but it has been suggested that the medieval Muslim historical tradition bears the dim recollection of a place called Macoraba. This was proposed by Jan Retsö, an Arabist in Swedish academia, in his learned book The Arabs in Antiquity (2003). He points to the ninth-century Meccan historian al-Azraqi and his Tales of Mecca, which includes stories about the Kaaba in ancient times. Al-Azraqi relates:

“The site of the Kaaba vanished and perished in the flood between Noah and Abraham. Its site was a hill of red clay which the deluge did not submerge. Yet the people knew that the temple was located there, even though it was not confirmed. Those who were oppressed and sought protection came from all over the world and called out to it ‘al-makrūb’; there were few who called out like this whom it did not answer. The people made pilgrimage to the site of the temple until God gave its location to Abraham when he wanted the temple to be rebuilt and its true religion and laws to be revived. Since God sent Adam down to Earth, His temple has never ceased to be glorified and sanctified by nation after nation, religion after religion; and before Adam the angels used to make pilgrimage to it.”

Retsö infers that al-Azraqi did not understand the meaning of al-makrūb in this passage. I agree. Usually a makrūb is someone anxious and troubled; the root has many unrelated applications according to the medieval lexicographers, but none can be said to obviously fit this text. For al-Azraqi, the name that people called out to the empty site of the Kaaba was arcane. Retsö’s implication is that al-makrūb could be the ancient name Macoraba, passed down through legend.

What is attractive about Retsö’s idea, in my opinion, is that he roots it in a local text: this is ostensibly a Meccan legend reported by a Meccan writer, which lends the evidence far more weight than airy speculations about Old South Arabian deserve. Nevertheless I have insurmountable doubts as to whether we can relate al-makrūb to Macoraba. The morphology of makrūb is quite different from that of Macoraba, and we have seen that Semitic is relatively unlikely to be rendered as Greek k. Most pressingly, I am not convinced that what poses as an ancient story about the Meccan sanctuary has much historical value.

Those who have benefited from Crone’s Meccan Trade and other studies will know that the medieval stories about Arabia before Islam have been weathered by literary forces for so long and with such intensity that even the matter-of-factest details are brought into question. Characters are swapped in and out, details are spun off to write law and exegesis, and stories are relocated to new environments. This activity is most energetic where the storytellers had the most interest: in the development of the Qur’an, the life of Muhammad, and the history of Mecca.

This is not the place to explain how academic scholarship is theorising the medieval historical tradition. For now, all we can say is that an opaque word in a very late sacred history may be too ephemeral to link up with a geographical name attested seven hundred years earlier. Those who are sanguine about the Arabic tradition’s historicity would do well to consider this anecdote in their treatments of Macoraba; I have to be cautious.


Conclusion: Macoraba, Meccaless

Macoraba’s etymologies have often been adduced to show that Mecca was a prominent holy site long before Islam. In practice, scholars have assumed that Mecca was an ancient holy site, they have assumed that Mecca is Macoraba, and they have gone looking for etymologies to cohere with those assumptions. It is telling that after 350 years of experimentation we have a handful of incompatible etymologies, none of which quite fits. It is also telling that five very different interpretations have resorted to languages that flourished outside ancient Mecca: Aramaic from Sprenger, Hebrew from Bochart and Dozy, and Old South Arabian from Glaser and Ali. Why, then, is there still a consensus that Macoraba is Mecca?

Unless I am mistaken, the idea was first noted in the seventeenth century; in the eighteenth century it found its way into encyclopaedias; and in the nineteenth, it was fairly common knowledge among Orientalists. Konrad Mannert helped to popularise the idea, but it was probably Albert Forbiger’s entry in Pauly’s Encyclopaedia that assured its place in the canon of Orientalist thought. The Encyclopaedia was so respected that a completely new edition in Pauly’s name was launched in 1893 and completed in 1978; yet another was launched in 1996, which is still producing volumes. Both have had entries on Macoraba; both declare it Mecca.

Through the Paulys we can also trace the changing attitudes to Macoraba’s derivation. While Forbiger made do with Mannert’s etymologies, Adolf Grohmann’s entry (1928) is a dense survey of hypotheses since Mannert, which comes down in favour of *mikrāb. Some eighty years later Isabel Toral-Niehoff’s contribution cuts right to the consensus:

According to Ptol. 6,7,32, city in north-western Arabia Felix, already at an early time equated with Mecca. Based on the southern Semitic root mkrb (‘temple’, ‘sanctuary’ but also ‘altar’). In pre-Islamic Mecca there was a temple to the moon god Hubal, who was worshipped by the tribes in the neighbourhood.

It is a marvellous illustration of *mikrāb’s triumph: what no-one could have imagined in the first Pauly was then presented as the strongest hypothesis in the second and undisputed fact in the third. As we have seen, the historiography of Macoraba is an embarrassment of riches; the latest Pauly staves off doubt by narrowing our vision to a consensus viewpoint. I am not implying that Toral-Niehoff did wrong: her role was to state the consensus for the record, which she did. But when a consensus endures by inertia, not merit, encyclopaedias can help interpretations to outlive their usefulness.

When the Orientalists came to write their own central reference, the Encyclopaedia of Islam (1913–1938), Macoraba-as-Mecca was reproduced by Henri Lammens and A.J. Wensinck in their entries on Mecca and the Kaaba; this was reiterated in the second generation of the Encyclopaedia (1960–2005) by Wensinck and by W. Montgomery Watt. The third edition has been underway since 2007, with the same high standards as its predecessors (see for example the entry on the “Ḥājib”, forthcoming). The new entries on Mecca and the Kaaba have not yet been published, but we can reasonably expect them to uphold the traditional equivalence of Mecca with Macoraba, sustaining a very old and very shaky hypothesis through the 21st century.

It is not impossible that Mecca is Macoraba, but right now, given the state of the field, there is no rigorous way to bridge the two names. It is easier to conclude that they are different places. Patricia Crone seems to be vindicated, and hopefully this blog post will be a second small step toward overturning the consensus.



Jawād ‘ALĪ, al-Mufaṣṣal fī Ta’rīkh al-‘Arab qabl al-Islām, vol. 4, 2nd ed. (1993), 9–10.

Giuseppe Simone ASSEMANI, Bibliotheca Orientalis, vol. 3 part 2, De Syris Nestorianis (1728), 561, cf. 583.

Muḥammad al-AZRAQĪ, ed. Ferdinand Wüstenfeld, Chroniken der Stadt Mekka [Akhbār Makkah] (1858), 20.

Samuel BOCHART, Geographia Sacra, vol. 1, Phaleg (1651), 237.

Mikhail D. BUKHARIN, “Mecca on the Caravan Routes”, 115–34 in Angelika Neuwirth et al. (eds.), The Qurʾān in Context (2009), 112.

Caspar CALVÖR, De Variis Orbis Religionibus (1705), 1094.

Patricia CRONE, Meccan Trade and the Rise of Islam (1987), 134–6.

Reinhart DOZY, De Israelieten te Mekka (1864), 80–1, 94–5.

Joseph ECKHEL, Doctrina Numorum Veterum, part 1 vol. 3 (1794), 503.

Albert FORBIGER, “Makoraba”, in Real-Encyclopädie, vol, 4 (1846), 1346.

Charles FORSTER, Historical Geography of Arabia, vol. 1 (1844), 265–6.

Eduard GLASER, Skizze der Geschichte und Geographie Arabiens, vol. 2 (1890), 87, 235.

Jacob van GOOL, Alfraganus (1669), 98, cf. 235.

Adolf GROHMANN, “Makoraba”, in Paulys Realencyclopädie, vol. 14:1, “Lysimachos–Mantike” (1928), 807–8.

Martin HARTMANN, Der Islamische Orient, vol. 2, Die Arabische Frage (1909), 121–2.

—“Makoraba. Eine Abwehr und eine Warnung”, in Orientalistische Literaturzeitung 6 (1911), 281–2.

A.H.L. HEEREN, Historische Werke, vol. 11 (1824) 111.

Johann Samuel HEINSIUS (pub.), Historisch-Politisch-Geographischer Atlas, vol. 7 (1747), 1006–7.

Henri LAMMENS and A.J. WENSINCK, “Mecca”, in Encyclopaedia of Islam.

Konrad MANNERT, Geographie der Griechen und Römer, vol. 6 part 1 (1799), 113.

Antoine-Augustin Bruzen de la MARTINIÈRE, Grand Dictionnaire Géographique, vol. 5 part 2 (1735), 15235–7.

William MUIR, Life of Mahomet (1858), 131 n.

Jan RETSÖ, The Arabs in Antiquity (2003), 438 and n. 61 on 450.

Carl RITTER, Erdkunde, vol. 12 (1846) 15, 231.

Friedrich C.L. Sickler, Handbuch der alten Geographie (1824), 768.

Isabel TORAL-NIEHOFF, “Macoraba”, in Brill’s New Pauly (1996—).

Aloys SPRENGER, Die alte Geographie Arabiens (1875), 155210.

W. Montgomery WATT, “Makka”, in Encyclopaedia of Islam, Second Edition.

A.J. WENSINCK, “Ka‘ba”, in Encyclopaedia of Islam.

A.J. WENSINCK and Jacques JOMIER, “Ka‘ba”, in Encyclopaedia of Islam, Second Edition.

Hermann von WISSMANN, “Makoraba”, Paulys Realencyclopädie, Supplement, vol. 12 (1970), 792.

YĀQŪT al-Ḥamawī, ed. Ferdinand Wüstenfeld, Mu‘jam al-Buldān, vol. 4 (1869), 616.

Johann Heinrich ZEDLER (pub.), Grosses vollständiges Universal-Lexicon, vol. 20 (1739), 8.


Consulted editions of Ptolemy’s Geography


—ed. Karl Friedrich August Nobbe (1845), 105.


—Valencia BH Ms. 0693, Italy (1477), 44r.
—Nancy Ms. 354, France (15th century), 118v.
ed. Nicolaus Germanus (1482), n.p.
ed. Henricus Petrus (1540), 115.
—ed. Vincenzo Valgrisi (1562), 222.


Edward Luther Stevenson, Geography (1932), 139. (This translation is however widely discredited.)


Ian D. Morris

Ian D. Morris

Historian and occasional human, powered by tea and cat videos.


  1. Same cause products the same effects as with Diodorus. There’s no Mecca before Islam. Then the consequences have to be drawn : none Mecca means logically none Muhammad ; he is bonded with Mecca by all Muslim discourse so much that the non-existence of the first leads to the non-existence of the second. And put this figure elsewhere won’t resolve any question about the existence of the Quran.
    Because there lies the real topic of the Mecca and the Muhammad existence : where comes from the Quran ? The theory of Islam about that is the traditional account. Wrong, because of the non existence of Mecca. There is none B plan here. Therefore, the explication of the existence of the Quran should be search elsewhere.

  2. is there no evidence that ancient arabs living in the south would interchange the letters ‘m’akka and ‘b’akka? that is the standard muslim explanation. y is that rejected?

    • That is quite an old idea – I think I’ve seen Philip Hitti raise it – but no, as far as I can tell, it’s weak. A linguist tells me there is no real correspondence between the radicals /m/ in Arabic and /b/ in Ancient (or Old) South Arabian. Even if there were, that would constitute a regular difference between languages, by no means a choice within a single language. Furthermore, /m/ does exist in ASA, so if the name Mecca were borrowed into ASA, there’s no obvious reason the pronunciation should be forced from /m/ to /b/ (nor vice versa). These are purely linguistic problems that seem to falsify the claim, and that’s before we ask the historical question of what Yemeni ASA has to do with a Hijazi toponym in an Arabic text.

  3. I think you’ve been unfair to Al-Sim’ani here. It’s not difficult to see Al-Sim’ani’s connection between Mihrab and Mecca. From Lane’s Lexicon, mihrab can also mean “the chief, or most honourable, sitting-place…the place where kings and chiefs and great men sit; a high place.. the highest chamber in a house…” The Kaaba itself supposedly does sit on a hill.

    The Quran uses it to mean “place of worship”, which ties it directly with Mecca as a place of worship. Lane’s Lexicon quotes Quran 29:12: maharibu bani Isra’il – the places of worship of the Children of Israel

    The other uses in the Quran clearly refer to it as a place of worship. See 3:37 and 34:13. It didn’t come to mean a prayer niche until later. The standard tafsirs for 34:13 for example, like Al-Tabari’s tafsir, propose the meaning of mosques, architectural niches, general buildings of worship, or even residential buildings.

    As for the use of South Arabian languages in Arabia proper, there’s no way to definitively rule anything out. Inscriptions in rocks all over Arabia show that the language situation on the ground was extremely complex, with many of these Arabic-like languages having seemingly died out. And South Arabian was definitely used in what’s today Saudi Arabia. Northern Saudi contained odd amalgams of Aramaic and Arabic. I would have no problem believing that there was a semitic pre-Arabic language spoken in the area somewhere that referred to Mecca as Meccaraba or something similar. Retso’s argument makes the most sense, and both the Semitic root mkrb and the mihrab arguments are believable.

    Frankly, I find Patricia Crone’s contributions worse than useless. She seems to have successfully made a career for herself by simply challenging any and all traditionally accepted narratives [I’ve omitted the rest of this comment, which is needlessly unpleasant about Prof. Crone, a respected colleague. It would be easier to reject this comment entirely, but there are valid issues to address here, and I’m grateful to Tariq for raising them. —IDM]

    • Well spotted. The problem is that the maps you’ve consulted are out of date; their creators did not have access to the range of manuscripts, including Greek, that we have today. What they’ve read as Negran in Latin is apparently a corruption of Egra in Greek, which is ancient al-Hijr, modern Mada’in Salih. Najran would seem to be Ptolemy’s Nagara. See for comparison Kennedy’s recent map, above.

      • There are few problems with Kennedy’s map. First, Egra, according to Strabo, is a port city in the land of Obodas, which doesn’t mix well with Kennedy’s location. Ptolemy didn’t mention Egra at all. Ptolemy’s Nagara, however, is in Oman, which is I think is not mentioned in Strabo. Lastly, Negran mentioned in Strabo is described as ‘peacable and fertile’ place which matches better with a near Asir/Yemeni location.

        • I wasn’t clear: what is written Negran in medieval Latin translation is written Egra in the Greek version. Ptolemy’s Egra is roughly where al-Hijr was historically, so its identification is reasonably solid. Negran therefore seems to be a corruption. More up-to-date resources like Kennedy’s map have taken this into account.

          Obodas refers to the land of the Nabataeans, which al-Hijr did belong to; I don’t know why Strabo puts Egra on the sea, but that is Strabo’s mistake, not Ptolemy’s.

          As I’ve said above, Ptolemy’s east-west calculations are weaker than his north-south calculations. As a result Nagara is correctly positioned with respect to other identified towns north-south but not east-west.

Leave a Reply

Your email address will not be published. Required fields are marked *