Ba’al, Hubal, and Allah
A Rebuttal to the Islamic Awareness Article Entitled “Is Hubal the Same as Allah?” by M.S.M. Saifullah and ‘Abdallah David
By Timothy W. Dunkin

The Value of the Islamic Traditions
Ba'al and Hubal - Linguistic Matters
The Coming of Hubal
So Who Was Hubal?
End Notes


There can be little doubt that one of the most contentious propositions that may be encountered across the broad spectrum of Muslim-Christian debate is the suggestion that, rather than being the omnipotent God of creation, the God of Abraham, the sole and all-powerful Ruler of the universe, Allah might merely be instead the evolutionary development of a native Arab god from being a high god in a previously polytheistic, or at best henotheistic, religious environment to being the monotheistic deity now worshipped by over a billion Muslims the world over. As a theological system, Islam has invested quite a lot of emotional and spiritual capital into the belief that it is the final revelation of Allah, the return to the true religion of the only God from the apostate departures which are represented by every other system on earth. Therefore, any suggestion that the god of Islam may have merely been elevated to his present exalted status from a previous position of being one among many in the pagan system found in the Jahiliya, the so-called “Times of Ignorance,” will naturally meet with a negative response from Muslims. The venerable Carleton Coon observed,

“Moslems are notoriously loath to preserve traditions of earlier paganism, and like to garble what pre-Islamic history they permit to survive in anachronistic terms.”1

So it is with the object of our present inquiry. In their article entitled “Is Hubal the Same as Allah?," Saifullah and David attempt to counter the charge that Allah’s origin lies in the pre-Islamic god Hubal, a deity who was worshipped in the Ka’bah in Mecca according to the traditions. As will be shown below, however, much of their argumentation is erroneous, and much more of it is simply irrelevant because it does not truly investigate the issue. I will present a refutation of their claims, and also provide what I hope to be some insights which will encourage further scholarly investigation into the subject of pre-Islamic religious history.

The Value of the Islamic Traditions

A good deal of Saifullah and David’s contentions rest upon arguments made from various stories that appear in the traditional Islamic historiographic material. So before I begin to address their specific arguments, I will first assess what value, if any, this traditional material has as far as presenting an accurate picture of the pre-Islamic world of the Arabs.

One of the most serious impediments to true learning that has plagued the study of the cultural and religious milieu from which Islam arose has been the excessive reliance upon the Islamic traditions and biographies as trustworthy historical accounts. While this error may be quite understandable on the part of Muslims themselves, the fact that many Western investigators also demonstrate a continued confidence in these materials exposes a systematic deficiency on the part of much of our scholarly activities with regards to Islam. This is not to say that this traditional historiography is of no value. On the contrary, the biographies and the sunnat and the ahadith and the histories show us a great deal about the views, beliefs, and attitudes of the early Muslims as Islam gradually developed into its present form. The stories about Mohammed were invented so as to present an idealized picture of what early Muslims thought a prophet should be. The traditions redacted into the historical accounts enlighten us as to the values and outlooks held by the various competing factions in the first two centuries of Islam.

But it is there that the real value of the traditional material ends. The various data presented in these works have, beginning with Goldhizer, been recognized as contradictory and synthetic. Goldhizer offered to the Western study of Islam the first real challenge to the heretofore universal belief that the Islamic traditions presented to us a priceless and unique insight into a completely documented set of historical events, what Ernest Renan called “the clear light of history,” surrounding the rise of Islam. It is not coincidental that he was also one of the first Western scholars to actually engage in a scientifically systematic study of the traditional material. Goldhizer observed the contradictory nature of the various traditions, and the obvious evidences for invention and embellishment in these works.2 Schacht neatly summarizes the necessary conclusions from such a study,

"I should like to present some ideas on what, I think, is a necessary revaluation of Islamic traditions in the light of our present knowledge; but am at a loss whether to call my conclusions something new and unprecedented, or something old and well known. No one could have been more surprised than I was by the results which the evidence of the texts has forced upon me during the last ten years or so; but looking back I cannot see what other result could possibly be consistent with the very foundations of our historical and critical study of the first two or three centuries of Islam. One of these foundations, I may take it for granted, is Goldhizer's discovery that the traditions from the Prophet and from his Companions do not contain more or less authentic information on the earliest period of Islam to which they claim to belong, but reflect opinions held during the first two and a half centuries after the Hijra.”3

Since then, other scholars have noted the ahistoricity of these traditional materials. G.H.A. Juynboll argued persuasively for the origins of the standardization and transmission system based upon the supposed authority of Companions of Mohammed and other early Muslims, the isnad, beginning near the end of the first Islamic century. He believed that it arose out of a recognizable need on the part of the early Muslims to establish a solid basis upon which to ground their traditional beliefs and to bring order to the very haphazard system of commandments, stories, personal examples, and doctrines jumbled together and each claiming authority.4 Going further, Crone states concerning the Sira of Ibn Ishaq, one of the primary sources for traditional historical information about early Islam and the pre-Islamic period,

“The work is late: written not by a grandchild, but by a great grandchild of the Prophet's generation, it gives us the view for which classical Islam had settled. And written by a member of the ulama, the scholars who had by then emerged as the classical bearers of the Islamic tradition, the picture which it offers is also one-sided: how the Umayyad caliphs remembered the Prophet we shall never know. That it is unhistorical is only what one would expect, but it has an extraordinary capacity to resist internal criticism...characteristic of the entire Islamic tradition, and most pronounced in the Koran: one can take the picture presented or one can leave it, but one cannot work with it."5

The traditional material is recognized as being very late - over a century after the events which it purports to be describing from first-hand witnesses. Crone further notes the incongruity of the many and various statements from these writings,

“There is nothing, within the Islamic traditions, that one can do with Baladhuri's statement that the kiblah (direction of prayer) in the first Kufan mosque was to the west (opposite direction to Mecca): either it is false or else it is odd, but why it should be there and what it means God only knows. It is similarly odd that Umar (second caliph) is known as the Faruq (Redeemer), that there are so many Fatimas, that Ali (Muhammad's cousin) is sometimes Muhammed's brother, and that there is so much pointless information...It is a tradition in which information means nothing and leads nowhere; it just happens to be there and lends itself to little but arrangement by majority and minority opinion.”6

As such, we can see that the Muslim historiography and traditions are not trustworthy presentations of historical events as they really were. Instead, this material often presents the viewpoints of the factions in power and events are cast as they wanted them to be. Muslims will argue that the systematic organization of many of the traditional materials (all the while depending, as we saw, on majority and minority opinions, on isnad chains of transmissional authority which are often not as dependable as one would hope) is evidence for their veracity. Yet, Wansbrough provides evidence which shows that the supply of isnad for statements or examples attributed to Mohammed and his Companions is a formal innovation datable only to the very beginning of the third Islamic century (200 AH/815 AD).7 Further, as Cragg so succinctly observed, the methodical organization and scrupulous concern for transmitted authority may themselves simply be the result of later redactive meticulousness,

“This science being so meticulous that it is fair (even if somewhat paradoxical) to suspect that the more complete and formally satisfactory the attestation claimed to be, the more likely it was that the tradition was of late and deliberate origin. The developed requirements of acceptability that the tradition boasted simply did not exist in the early, more haphazard and spontaneous days.”8

Thus, we must understand that any investigation into a question of history and empirical evidence simply cannot rest upon an uncritical acceptance of the Muslim traditions as literal, historical documents. Instead, we must take the same approach to them that Joseph Schacht counseled - that until traditions about the Prophet (and by extension, I believe we can say, history before and during his purported lifetime) are demonstrated valid by evidence, they should not be taken as authentic, but rather as the “fictitious expression of a legal doctrine formulated at a later date.”9 While I believe that we can perhaps see a “kernel of truth” lying at the heart of some of the statements made in the traditional materials that pertain to our present study, we must also exercise enough critical faculty to strip away the chaff that surrounds the kernel.

Ba’al and Hubal - Linguistic Matters

Before addressing the relevant traditions, let us first engage the particular linguistic arguments employed by Saifullah and David to substantiate their argumentation. They attempt to dispute the equation of Hubal (hbl) with the more general Semitic deity Ba’al (b’l) by making recourse to a number of questionable historical and linguistic arguments. They begin by summarizing Noja’s thesis10 that the name “Hubal” originated from the elision of hn-ba’al to habal or hubal, ha-/hn- being a form of the definite article in some early dialects used by the ancient Arabians. The assimilation of the n and the disappearance of the guttural ayin were proposed in the process.

Against this, they first present a somewhat extraneous argument against the transformation of hn- to the more familiar ‘l- article, known to us today in Classical and Modern Arabic dialects. They attack this thesis by stating,

“The idea that the h- or hn- article found in Ancient North Arabian is the ancestor of Arabic ’l- has been suggested by scholars over a long period.[31] This view has come under criticism due to the lack of epigraphic evidence for the transformation of h- or hn- to Arabic ’l-.[32] Theoretically, it can be argued that it could have happened in a number of ways, the problem always come back to the lack of epigraphic evidence for the actual process.[33] Noja assumed a similar transformation from the Ancient North Arabian h- to Arabic ’l-.[34] Not surprisingly, he did not furnish any proof either.”

This is beside the point, since the name under discussion is “Hubal” (with a theoretical ha-/hn- article still present), not some hypothetical “’l/al-bal” proposed by Noja. Concerning the linguistics of Noja’s proposal for the origin of the name “Hubal” itself, his thesis is certainly quite feasible. Saifullah and David correctly point out that, in the main, the hn- article appears to be somewhat older than the ‘l- form, but that Herodotus’ use of the name Alilat to describe the Arabian goddess may indicate an earlier use of the ‘l- form several centuries before that form finds broad attestation in the epigraphic record. They appear to be trying to argue from this that Noja’s hn-ba’al argument is invalid, since the ‘l- form was certainly available to use for any name meaning “the lord.”

However, their argument seems to assume a cut-and-dried linguistic uniformity in the ancient Arab world that simply was not there. Beeston has pointed out that prior to the general dominance of the ‘l- form of the article in Arabic dialects which was finally established around the beginning of the 6th century AD, there was a “linguistic mosaic in the peninsula.”11 The ha/hn- form was just as widespread and just as ancient as the ‘l-, even if we consider the evidence from Herodotus as sound. Retsö notes that the ha/hn- form is attested as early as the latter half of the 5th century BC in inscriptions found in Arab-occupied areas east of the Nile Delta near Pelusium, which mention “Geshem the Arabian” and which are devoted to HN-’LT (the goddess).12 This alone concretely places the ha/hn- form as contemporaneous to Herodotus’ ‘l- form. Further, Livingstone has proposed that the hn- form of the article (as it would have appeared in the Arabian dialect) should be implicitly understood to have existed with certain Arab terms that were apparently carried over wholesale into the Akkadian of a triumphal inscription celebrating victories won by the Assyrian king Tiglath-Pileser III (r. 744-727 BC).13 While this reading is more tenuous, it may well push the epigraphic evidence for the hn- form in Arabian languages back another three centuries.

So, we see that the hn- form is definitely ancient. It was also, however, relatively recent, being found in Arabian inscriptions until around 500 AD. This means that the form would certainly have still found much use, especially among Arabian dialects such as Lihyanic, Thamudic, and Safaitic (all used in the Arab regions east of Syro-Palestine and in the northern part of the Hijaz) in the general timeframe that the tradition about ‘Amr ibn Luhayy (which we deal with below in more detail) seems to indicate that he had brought the idol of Hubal to the Ka’bah.

The proposed assimilation of the n in hn-ba’al ---> haba’al is certainly possible linguistically. Southern and Vaughn demonstrate that the assimilation of an n before a consonant is fairly typical in North Semitic languages,14 and indeed they note that it is well-attested and not just theoretical.15 This same phenomenon is observed in Hebrew, for instance, where the terminal n in the preposition min (with) is assimilated with the doubling of the following consonant (except, of course, when before a guttural or a resh, in which case the prepositional vowel is lengthened along with the assimilation of the nun). Voigt further points out that old North Arabic forms show assimilation of the n to the following consonant,16 and do not seem to show a doubling of the consonant, as is found in some other North Semitic languages. Thus, the proposed elision by Noja is certainly possible on this count, as well.

What of Saifullah and David’s argument that Hubal cannot come from Ba’l because this transformation would require the dropping of the glottal consonant ayin? Again, their arguments are less than convincing. They state,

“Moreover, for the name b‘l to become bl with the loss of ‘ayn, it would have to have been transmitted through a language such as Akkadian or Punic in which the ‘ayn had disappeared. This would give in Akkadian Bel and in Punic Bol.”

However, the dropping of the ayin is not impossible. Drijvers certainly did not consider it to be, as he saw no difficulties in stating that Ba’al-Bel-Bol (together) was the original West Semitic form of the name.17 Beeston states that the “conversion of consonant into vowel” such as occurs in the Punic bol for ba’l, is “well-attested in Semitic languages.”18 More to the point, Voigt demonstrates that glottal stops in Arabian dialects can contract, using the example of the contraction of the hamza in the conversion bi-?al ---> bi-l.19 This same principle could certainly apply to the contraction of the similar ayin. As such, Noja’s argument, based as it is upon the disappearance of the ayin, is most certainly plausible.

Saifullah and David argue, for some reason, that the conversion of b’l ---> bl would not have taken place because, while the forms Bel and Bol were found in Palmyra, Palmyrene Aramaic did not use the ha/hn- form of the article. They then proceeded to closely paraphrase (including the co-opting of his endnotes) a large portion of text from Healey’s Religion of the Nabataeans before getting back to the point by arguing that “there is no Nabataean and Safaitic epigraphic evidence which shows that the name b‘l to becoming [sic] bl with the loss of ‘ayn, which in turn enabled hb‘l to become hbl.” Again, both of these arguments are beside the point. There is no reason why the lack of the ha/hn- article in Palmyrene would have any bearing on this discussion whatsoever. Nobody has proposed that the name Hubal came from Palmyrene, and there were certainly many other dialects, including those much closer to the Arab milieu such as Nabataean (in which the name appears as hblw) from which an entrance by Hubal into the Arab consciousness could have been made. Many of these dialects also used the ha/hn- form of the article.

Further, their arguments involving Ba’alshamen seem gratuitous. By their own admission, this deity was introduced into the Nabataean realm from Syria, where he was the “lord of heaven,” and therefore had no direct connection with indigenous Arab religion. Nor has anyone suggested a direct equation of this deity with Hubal. Rather, the more proper argument as we will see below would be one that suggests that Hubal was the result of a long process of evolution from the Ba’al deities of other lands (ones where b’l-form dialects predominated) - and could be considered to be the same deity in much the same way that Ba’al and Hadad were deemed one and the same by the Phoenicians and Aramaeans or how Ba’al and Zeus were assimilated in Hellenistic Syria. This association would have been based upon similarities of station and function held in common by these gods in each area. Hubal would not need to have directly developed from some hypothetical Huba’l - he need only have been syncretized with the ba’alim of these other regions. The evolution would not need to have been linguistic, but only conceptual. Further, there is no reason why we would expect to see any epigraphic evidence to show “b’l becoming bl.” The vector for entry of Hubal into the Ka’bah in Mecca is traditionally considered to be either from the Transjordan or from Hit in Mesopotamia, both of which were already settled by Arabs in the timeframe suggested by the ibn Luhayy tradition. But, this does not mean that the dialect of these Arabs was Classical Arabic (and indeed, in the 3rd century AD it almost certainly was not). The dialect of these Arabs would have been closer to that of the Nabataeans and the Mesopotamian Arabs, dialects in which the form bl was used. If Hubal were brought to Mecca from either of these regions, then his name hbl would certainly reflect those dialects and simply indicate the direct carryover of the name into the later b‘l-containing dialect of Classical Arabic, and the whole issue of converting b’l to bl is nullified.

The Coming of Hubal

In light of what has been said above, it is unfortunate that the majority of our information concerning the place of Hubal in the Jahiliya comes from the traditional Muslim histories. These traditions show an unfortunate tendency on the part of the early Muslim historiographers towards making the early histories conform to the orthodoxy of Islam once it had matured and crystallized into its present form. Attendant with this is a certain amount of artificiality and contradiction built into these stories, as the early historians sought to reconcile and organize a scattered and highly variant body of subject matter. Let us take for a relevant example that of the story of how Hubal even came to be in Mecca. It is this tradition, relating the introduction of Hubal’s idol into Mecca by Amr ibn Luhayy, that Saifullah and David first address in their article. They make much of the “missionary” (i.e. Gerhard Nehls) who argues that Hubal originated as “Ha-Baal” from Moab. To counter this argument, they point out that the origin of Hubal from Moab is uncertain, as some traditions relate that Luhayy brought him to Mecca from Hit, a city in central Mesopotamia. Saifullah and David state,

“There is no clear-cut position that can be adduced from the Islamic traditions on the issue of the place of origin of the Hubal idol at Makkah, although all of them are united on its foreign origin.”

But this just makes my point - the traditions themselves are untrustworthy as history, per se. Some of the Muslim authorities in the graphia say he came from Hit, while others say he came from Balqa’ or Moab in Syria. But, we can find the kernel of truth. All of the authorities are united, we must understand, only in affirming Hubal’s non-Meccan origin, an affirmation likely true due to the uniformity of its proposition. However, I believe it is incorrect, after a fashion, to state that his origins were “foreign.” This is because, in the timeframe in which this event is reputed to have taken place, sometime around the 3rd century AD, both of the regions suggested as origins for Hubal were dominated by Arab tribes. The regions east of the Jordan river (including Balqa' and Moab) had long been known as part of “Arabia.”20 Likewise, Hit was a city in an Arab-dominated region which had been settled by Arab tribes for at least a century prior to the time of the Luhayy story.21 Indeed, the Arab tribes of central Mesopotamia played an important geopolitical role as client states and buffer zones between the two superpowers of the time, Rome and the Parthian/Sassanid Empire. As such, if Hubal was brought from either of these areas, it was most likely by an Arab, and then it is not at all far-fetched to suggest that he might well have been a deity with whom the Arabs were already familiar. This would explain the apparent ready acceptance of him by the Meccans, to the extent that they set up his icon as the prominent idol in the Ka’bah precinct. Hence, this story seems to relay a reliable substratum of information to us, once we view the kernel of truth as telling us that at some point in the Jahiliya, probably at some point in the 3rd century, a deity most likely already known to the Arabs as a cultural group was specifically introduced into the haram of Mecca, and was apparently made its presiding deity.

This historical reconstruction is supported by the fact that the name for this god was “Hubal,” without the ayin. This would seem to indicate that his origin was from among a dialect group which used the bl-form, and which also used the ha/hn- article. Dialects like these found representation in the northern Hijaz and Syrian areas. Further, this introduction appears to have taken place prior to the establishment of the ‘l-form (whose most well-known representative is the Classical Arabic of the Qur’an and the other traditional writings) as the dominant dialect type (around the beginning of the 6th century AD), which is why we would not see Noja’s hypothetical ’l-bal form. The earlier attestation of Hubal in the Hijazi regions of the Nabataean kingdom, as well as in the Transjordan and Syria, suggest that the Transjordanian origin of Hubal is the correct choice between the two suggestions.

Now, if Hubal was a known quantity to the Arabs, then how does he relate to Allah? We must understand that a straight-forward reading of the traditional material, even with the later redactions, seems to indicate that Hubal was the Lord of the Ka’bah, a position also attributed to Allah, whose house the Ka’bah now is (bayt allah). Perhaps the premiere story in the traditions which bears on this question is that of Abd al-Muttalib, the grandfather of Mohammed, and his oracle from Hubal.22 In this tradition, which deals with Abd al-Muttalib’s efforts at getting around a vow that he had made to Allah to sacrifice one of his sons, it is twice mentioned that Abd al-Muttalib prayed to Allah while standing next to the statue of Hubal. In their apologetic, Saifullah and David more or less dismiss the notion out-of-hand that this would suggest that Hubal and Allah were connected,

“As to how standing next to the statue of Hubal and praying to Allah is equivalent to Hubal actually being Allah is a great mystery. By this "logic," a Christian standing next to the cross and praying to the Trinitarian deity makes him a cross-worshipper.”

This argument, of course, lacks much and the outright dismissal is irresponsible. Their attempt to draw a parallel between Abd al-Muttalib’s standing next to Hubal while praying to Allah with a hypothetical Christian standing next to a cross and praying to God is non-sequitur. A cross is not an idol fashioned in the likeness of a certain god, nor is divination made to a cross, while both of these most certainly do apply to the statue of Hubal . The statement that al-Muttalib was standing next to the idol of Hubal is recognized as an euphemistic statement made by later Muslim traditionalists who were squeamish about depicting the grandfather of Mohammed praying directly to an idol. But we must understand, the idol of Hubal is central to the entire story. It was through this idol that the cleromantic divinations took place, as the Arabs sought guidance from the god. The purpose for al-Muttalib’s worship was to take part in just this sort of divination, and he does so while praying specifically to Allah. As such, al-Muttalib was doing more than just “standing next to” the icon of Hubal. The story quite clearly demonstrates that al-Muttalib viewed Allah and Hubal to be one and the same, which is why he explicitly prays to the one for guidance while simultaneously engaging in the divination governed and controlled by the other. The story shows clearly, if indirectly, the equation of the two in the mind of Abd al-Muttalib.

In opposition to the equation of Hubal with Allah, first suggested by Wellhausen largely because of the prominence of Hubal in the “House of Allah,”23 Saifullah and David bring forward several quotations to serve as authorities on which to base their rejection. The citations from Peters and von Grunebaum will not be addressed here, as they really amount to no more than simple affirmations of the traditional viewpoint found within the larger body of those authors’ texts, and present no argumentation against which criticism needs to be made. The statements by Margoliouth and Crone are more interesting, each in their own way. Citing Margoliouth, Saifullah and David state,

“For example, over 100 years ago, Margoliouth had casted [sic] doubts on Wellhausen's identification of Hubal with Allah and dismissed it as a "hypothesis."

They then proceed to focus upon Margoliouth’s use of the term “suggested”24 and make it appear as if Margoliouth was rejecting Wellhausen’s suggestion. This is despite his statement which he made immediately previous to the sentence quoted by Saifullah and David, “Between Hubal, the god whose image was inside the Ka‘bah, and Allah ("the God"), of whom much will be heard, there was perhaps some connection.”25 Saifullah and David are simply reading their own preconceptions into Margoliouth’s words. He was merely being cautionary - as any good investigator in a field in which so much evidence remains to be uncovered must be. Margoliouth was affirming that the link between Hubal and Allah was hypothetical - but then again, that only means that it is a proposal not fully borne out yet but the proposition of which nevertheless is based upon evidences at hand, nothing more and nothing less. Saifullah and David are merely putting words into Margoliouth’s mouth, even though what Margoliouth really said in no wise “noted that Hubal and Allah can't be one and the same entity,” as they would have us to believe.

Concerning the citation from Patricia Crone, Saifullah and David have merely cited the last of several suggestions made by Crone as to the disposition of Hubal and Allah - the one which is based upon an acceptance of the Muslim traditions as essentially historical in nature. If one does not accept that proposition, as I do not for reasons outlined above, then the arguments from traditions in which people are asked to renounce Hubal in favor of Allah are of little diagnostic value. Indeed, the more reductionist argument that Crone suggests prior to the statement cited by Saifullah and David, made on the basis of historical and archaeological evidence, would seem to strike against their arguments. While discussing aspects of Arabian litholatry (the worship of a deity through a stone), she notes that this can easily apply to Allah as well, through the black stone housed in the Ka’bah,

“If we assume that bayt and ka’ba alike originally referred to the Meccan stone rather than the building around it, then the lord of the Meccan house was a pagan Allah worshipped in conjunction with a female consort such as al-’Uzza and/or other “daughters of God.” This would give us a genuinely pagan deity for Quraysh and at the same time explain their devotion to goddesses.

“But if Quraysh represent Allah, what was Hubal doing in their shrine? Indeed, what was the building doing? No sacrifices can be made over a stone immured in a wall, and the building accommodating Hubal makes no sense around a stone representing Allah. Naturally Quraysh were polytheists, but the deities of polytheist Arabia preferred to be housed separately. No pre-Islamic sanctuary, be it stone or building, is known to have accommodated more than one male god, as opposed to one male god and female consort. The Allah who is attested in an inscription of the late second century A.D. certainly was not forced to share his house with other deities. And the shrines of Islamic Arabia are similarly formed around the tomb of a single saint. If Allah was a pagan god like any other, Quraysh would not have allowed Hubal to share the sanctuary with him - not because they were proto-monotheists, but precisely because they were pagans.”26

It is from here that Crone continues on into the statement quoted by Saifullah and David - a statement which, in context, seems to be a hypothetical answer to her previous questions if one were a Muslim who did not accept that Allah was previously a pagan god. She is not, per se, arguing against the equation of Hubal and Allah - indeed, she does not directly address the question at all.

But, we see some interesting information presented. Arabian sanctuaries housed no more than one male god. So indeed, what was Hubal doing in Allah’s house? The most reasonable answer is simply that Hubal and Allah were not viewed by the pre-Islamic Arabs as being different deities. They were compatible. More than that, they were co-personal. This brings sense to the al-Muttalib story, and rejects the otherwise nonsensical suggestion that praying to one god while at the same time divining through the other somehow does not mean that the two gods were really the same.

What then of the traditions relied upon by Saifullah and David, most notably that of Abu Sufyan (the leader of the Quraysh in Mecca), which depict the followers of Hubal and those of Allah as being in opposition to one another? These traditions are simply untrustworthy, and most likely represent polemical inventions by later Muslims to serve as object illustrations of the victory of Allah over the Jahiliya pagan system. The story in which Abu Sufyan cries, “Be thou exalted, Hubal!” and Mohammed replies, “Be thou more exalted, Allah!” is programmatic in its polemical presentation. This is especially so when we consider the addendum to this story, also adduced by Saifullah and David, in which Abu Sufyan holds a meeting with Mohammed and realizes the error of his previous ways, and becomes a good Muslim. The traditional literature of Islam abounds with this sort of story, in which pagans and apostates realize their error and “revert” to Islam as the only and obviously true way.27 There is simply no good reason to rely upon the traditions about Abu Sufyan and his (and Hubal’s) opposition to Allah as any sort of truly historical set of events, especially in light of the rest of the opposing evidences.

So Who Was Hubal?

We have previously seen that the understanding of b’l = bl is certainly not improbable on linguistic grounds, within the Semitic environment that is the setting for this discussion. Indeed, we see that throughout the ancient Near East, gods bearing these names, with and without the ayin, appear to be equivalent. Drijvers’ ready link between Ba’al, Bel, and Bol was already noted above. In Palmyra, the older deities Yarhibol and Aglibol, each bearing the archaic form of the name, appear to have been gradually assimilated into a cult association with the more recent Mesopotamian import Bel,28 and could even be considered as hypostases of that deity. Brody likewise notes that one of the forms taken by Ba’l at Palmyra was ‘Aglibol (bearing the older and non-ayin containing form), meaning “calf of Bol.”29 Fahd notes that Bel is the Assyrian counterpart to Ba’l.30 There appears to be no problem in equating Bel/Bol with Ba’al on the part of specialists in the field of ancient Near Eastern history and religion. Saifullah and David’s argument that the two cannot be conjoined because of the lack of an ayin is spurious. The two forms are clearly understood to be cognate, and there is no reason why any development of one into the other has to be directly observed since ultimately, we are dealing with the use of this name across differing dialectical groups for which we would not expect to see direct epigraphic linguistic progression, even when we deal with evidence solely from Arabia (due to the “linguistic mosaic” found in the peninsula at the time).

The name Ba’al appears to have originally been titular and localized - it would denote “the lord” over a certain region. Examples of this sort of usage would include Baal-Peor, Baal-Zebub, and Baal-Shamiyn. However, by the middle of the 2nd millennium BC Ba’al had also become a god, with his own name, in his own right,31 as evidences from the El Amarna documents and Ugaritic texts indicate. Hence, b’l/bl evolved from a generic title to a specific name. The local ba’alim remained, however, and were most likely viewed by their worshippers as being personifications or manifestations of the high god Ba’al.

The name Hubal, then, begins to be comprehensible to us, seeing as there is no sound argument against understanding Hubal to be a ba‘al. Hubal appears late on the scene, relatively speaking. We do not see any real evidence for his existence until the time of the Nabataeans, and from there he goes wherever the Arabs go - to Palmyra, the Hijaz, and so forth. The name, itself, seems to suggest that it originally was a title or epithet of a high god. Hubal means “THE lord,” seeming almost as if to differentiate him from others who might conceivably be given that title. In this sense, its use would be much the same as that given to ilah/allah. Handy notes this, when he states that “both il and b’l may designate two distinct deities, but they are also used as the generic word for ‘god’ and the common noun ‘lord’ respectively.”32 Just as with Ba’al, the name Hubal most likely originated as a general term or title, later being applied as the name of one specific deity. Hubal would have went from being a title applied to local deities, to being the name for a high god, one viewed as more universal in his power. There is nothing strange about the notion (and indeed it should perhaps be expected) that a high god in a henotheistic system (and one which in Arabia seems to have gradually been evolving towards monotheism) would be referred to with universalist terminology such as “the lord” or “the god,” denoting his stature as the god par excellence.

An example of this sort of evolving conception was found with the Nabataeans and other northern Arabian tribes who referred to Dushara, their high god, with the term ’lh’, “the god.”33 The name Hubal “the lord” certainly fits this motif of a local high god being referred to as “the Ba’al of ____” Likewise, the term Allah (= al-ilah, the god) has the same sort of ring to it. We know that other deities in the Semitic Near East were referred to with the title/epithet of Ilah/Allah. In South Arabia, the goddess 'lhtn (containing the suffixed article -n, making this the South Arabian equivalent to al-ilaha, "the goddess") was a sun-goddess and was paired with the deity 'lhn.34 This 'lhn would be the South Arabian equivalent to the more northerly al-ilah - Allah - and his association with a female solar deity suggests that he fulfills the role of a lunar god, per the typical astral arrangement in the settled parts of Arabia. The Edomite deity Qos/Quash, clearly connected with moon worship through the use of the typical crescent moon and star symbology found throughout the ancient Near East,35 was carried over into the Nabataean realm with the name Qos-Allah.36 Guillaume noted that Ilah was a name applied to the moon god among some Pre-Islamic Arabian tribes.37

Hubal did have astral, and in many cases specifically lunar, characteristics, just as we have seen were connected with al-ilah. Hubal is noted for having originally had a stellar aspect to his nature, in addition to the cleromantic functions he acquired in the Ka’bah.38 Hommel also notes that among the Nabataeans, Hubal was a moon god, one of two along with Dushara.39 Occhigrosso flatly states that Hubal was a moon god whose worship was associated with the black stone at the Ka’bah, and that he was also associated with Manat (also the object of Arabian litholatry).40 That Hubal should have a lunar station should not necessarily be surprising. If the name were originally titular, then its descent from and connection with b’l/bl will also carry with it a legacy of astral religion. In later ancient Near East times, the various ba’alim developed astral overtones, which were primarily solar in nature,41 but which could also be lunar. Even in post-Hellenistic times, we see this phenomenon continue to take place. A votary inscription in Harran devoted to the moon god Sin calls that god the “Baal of Harran.”42 In Palmyra, Yarhibol and Aglibol were names for the solar and lunar deities respectively who were associated with Bel of the Mesopotamian immigrants.

Saifullah and David’s arguments against Hubal as a moon god are simply wrong. Contrary to their claims, the view of Hubal as having lunar provenance is attested by others besides Winckler and Brockelmann. Likewise, while it is true that Nielsen’s particular theory about astral triads in Arabian religion was overstated and has rightly been rejected, this does not mean that there was no astral, and especially lunar, character to pre-Islamic Arabian religion, as Saifullah and David appear to be arguing. Indeed, the evidences from archaeology and history, tell us that astral religion made up a goodly share of pre-Islamic Arabian devotion. It was to Tayma in Arabia that the Babylonian king Nabonidas went to buttress his devotion to the moon god, and the presence of lunar temples all across the peninsula and the appearance of lunar gods in the pantheons of the various tribes of pre-Islamic Arabia show that moon worship played a significant role in the religious life of the people of Arabia prior to the rise of Islam.

And it is here that we see that two seminal claims advanced by Saifullah and David - the rejection of Allah being the same deity as Hubal, and the dismissal of the characterization of Hubal/Allah as a lunar deity - fall apart. Clearly “Allah,” both as a title and as a proper name, was applied to lunar deities in the ancient Near East. Allah also shares many direct similarities with Ba’al/Hubal. We know that at various times in pre-Islamic Arab regions, Hubal was linked to the same deities with whom Allah was connected. Hoyland informs us that Hubal was worshipped jointly with Manat in the Hijaz portion of the Nabataean kingdom,43 and that he was served by a priestly office jointly with Dushara and Manat at Hegra, also in the northern Hijaz.44 Indeed, the earliest inscription to bear Hubal’s name shows him to be associated with Manawat, a cognate name of Manat, in the Nabataean kingdom.45 Also among the Nabataean remains have been found references to Ba’l along with Manat and al-Uzza.46 All in all, despite the claims of Saifullah and David to the contrary, Hubal does indeed seem to have been “integrated into the divine family” of Allah.

This is even more enlightening when we consider that the evidence of the much earlier Ras Shamra texts tell us that Ba’al was a god who had three daughters, just like Allah.47 It is not at all improbable that Ba’al with his three daughters passed, with some modifications and evolution due to the passage of time, to being Hubal with three daughters - Hubal (the lord) known also by the name Allah (the god, al-ilah). It then becomes explicable why the Qur’an would condemn the worship of the daughters of Allah as shirk (association of other deities with Allah), while remaining strangely silent about the worship of Hubal. The worship of Hubal was the worship of Allah - the error of the particular idolatry in question lay solely in associating daughters with Hubal/Allah. Allah, as a title,48 was applied to Hubal, the god’s name, so the writers of the Qur’an did not see a need to raise a row about Hubal. It is likely that only later, when the absolute monotheism of Islam became more crystallized and reference to the names of pre-Islamic deities in conjunction with Allah became discouraged, do we see the traditions arising in which Hubal is opposed to and ultimately defeated by Allah.


The identification of Hubal with earlier b’l gods has been shown to be linguistically feasible, but paradoxically this linguistic possibility is not necessary to make a case for the connection. The traditions which deal with Hubal, while showing a great amount of redaction by later Muslims, nevertheless still contain a core of information that helps to show us that Hubal was understood to be the Lord of the Ka’bah. Hubal demonstrates the characteristics of having been a high god, and as was seen, his presence in the Ka’bah is not merely incidental, but is most logically understood to have been as “the Lord of the House.” The suggestion that the terms ba’l and ilah, both general terms, can refer to this “lord of the house” interchangeably is by no means out of bounds. Despite claims to the contrary, Hubal appears to have had astral characteristics among his repertoire, and he was associated with goddesses with whom Allah was also associated. Further, Allah was a name applied elsewhere to moon gods, in Yemen and in Nabataea. The conclusion that can be drawn from all of this is that Hubal, his position as a major deity perhaps affirmed by calling him “THE lord,” and who carried a legacy of lunar provenance, was the ba’l of the haram precinct in Mecca. Further, he was the deity raised to strict monotheistic status during the early development and solidification of the Islamic religion and known henceforth as Allah.

End Notes

(1) - C. Coon, "Southern Arabia: A Problem for the Future," The Annual Report of the Smithsonian Institute, 1944, p. 398
(2) - See especially the statements in I. Goldhizer, Muhammedanische Studien, Vol. 2, pp. 18-19
(3) - J. Schacht, "A Revaluation of Islamic Traditions," Journal of the Royal Asiatic Society, (1949), p. 143
(4) - See G.H.A. Juynboll, Muslim Tradition, p. 5
(5) - P. Crone, Slaves on Horses, p. 4
(6) - Ibid., p. 12
(7) - J. Wansbrough, Quranic Studies, p. 179
(8) - A.K. Cragg, Encyclopaedia Britannica: Macropaedia, 15th Ed. (1998), Vol. 22, p. 11
(9) - See J. Schacht, Origins of Muhammadan Jurisprudence, p. 149
(10) - S. Noja, “Hubal = Allah,” Reconditi: Instituto Lombardo Di Scienze E Lettere, Vol. 28 (1994), pp. 283-295
(11) - A.F.L. Beeston, "Languages of Pre-Islamic Arabia," Arabica, Vol. 28 (1981), p. 183
(12) - J. Retsö, The Arabs in Antiquity: Their History from the Assyrians to the Umayyads, p. 250
(13) - A. Livingstone, “An Early Attestation of the Arabic Definite Article,” Journal of Semitic Studies, Vol. 42(1997), p. 261
(14) - M. Southern and A.G. Vaughn, "Where Have All the Nasals Gone? nC > CC in North Semitic," Journal of Semitic Studies, 42(1997), p. 282
(15) - Ibid., p. 263
(16) - R. Voigt, “Der Artikel in Semitischen,” Journal of Semitic Studies, Vol. 42 (1997), p. 225
(17) - H.J.W. Drijvers, The Religion of Palmyra, p. 10
(18) - A.F.L. Beeston, Bulletin of the School of Oriental and African Studies: London University, Vol. 46, no. 3(1983), p. 552, in a review of Bruce Ingham’s North East Arabian Dialects
(19) - Voigt, op. cit., p. 225
(20) - A term, in fact, which originally encompassed only the Sinai, the deserts east of Syria, and the northern parts of the Hijaz around Midian and al-’Ula (Dedan).
(21) - See especially Retsö, op. cit., chaps. 15-16
(22) - Related in its fullest form by Ibn Ishaq in his biography of Mohammed, see The Life of Muhammad: A Translation of Ishaq's Sirat Rasul Allah, trans. A. Guillaume, pp. 66-68
(23) - J. Wellhausen, Reste Arabischen Heidenthums, p. 75
(24) - “...yet the identification of the two suggested by Wellhausen is not yet more than an hypothesis.” - D. S. Margoliouth, Mohammed And The Rise Of Islam, p. 19
(25) - loc. cit.
(26) - P. Crone, Meccan Trade and the Rise of Islam, pp. 192-193
(27) - For instance, there is the story, related by Ibn Sa’d and Ibn ‘Asakir, which details the conversion of Hind bint ‘Utba, the wife of Abu Sufyan and the mother of the future Caliph Mu’awiya. In this story, Hind dreams for three successive nights. The first night, she is in pitch black darkness and Mohammed appears to her in a beam of light. The second night, she dreams she is on a road, and Hubal and Isaf (another idol) are on either side of the path, calling to her to leave the path, while Mohammed is in front of her showing her the right path. The third night, she is standing at the brink of hell, and Hubal calls upon her to enter, while Mohammed seizes her clothing from behind to draw her back. When she wakes the next morning, she strikes the idol in her house with an adze and says to it, “You have misled me for a long time!,” after which she converts to Islam and pledges her allegiance to Mohammed. See M. Lecker, “Was Arabian Idol Worship Declining on the Eve of Islam?,” pp. 4-5, in People, Tribes and Society in Arabia Around the Time of Muhammad, Ch. III. He cites this story from ‘Ali ibn al-Hasan ibn ‘Asakir, Ta’rikh madinat Dimashq, 70:177 and Muhammad ibn Sa’d, al-Tabaqat al-kubra, 8:237. In this particular article, Lecker presents a number of similar stories from the early Muslim historiographers which contain this programmatic theme of dramatic conversion to Islam, often accompanied by magical or supernatural circumstances.
(28) - J. Teixidor, The Pantheon of Palmyra, pp. 2-3
(29) - A.J. Brody, “Each Man Cried Out to His God”: The Specialized Religion of Canaanite and Phoenician Seafarers, p. 56, note #95
(30) - T. Fahd, Le Pantheon de l’Arabia Centrale a la Veille de l’Hegira, p. 53, note #8
(31) - H. Ringgren, Religions of the Ancient Near East, p. 131
(32) - L.K. Handy, Among the Host of Heaven: The Syro-Palestinian Pantheon as Bureaucracy, p. 25
(33) - J.F. Healey, The Religion of the Nabataeans: A Conspectus, p. 92
(34) - M. Maraqtan, "An Inscribed Amulet from Shabwa," Arabian Archaeology and Epigraphy, Vol. 7 (1996), p. 91
(35) - See I. Browning, Petra, p. 28
(36) - See N. Gleuck, Deities and Dolphins, p. 516
(37) - A. Guillaume, Islam, p. 7
(38) - T. Fahd, The Encyclopedia of Islam, eds. B. Lewis, V.L. Menage, C. Pellat, and J. Schacht, Vol. 3, p. 536
(39) - F. Hommel, First Encyclopedia of Islam, eds. M.T. Houtsma, T.W. Arnold, R. Basset, and R. Hartmann, Vol. 1, pp. 379-380
(40) - P. Occhigrosso, The Joy of Sects, p. 397
(41) - See e.g. F.M. Cross, Canaanite Myth and Hebrew Epic, n. 13, p. 7, who notes the equivalence of Baal Shamen with Zeus Helios, a solar deity, in Nabataean inscriptions.
(42) - Teixidor, op. cit., p. 43
(43) - See R. Hoyland, Arabia and the Arabs: From the Bronze Age to the Coming of Islam, p. 142
(44) - Ibid., p. 159
(45) - Fahd, The Encyclopedia of Islam, eds. B. Lewis, V.L. Menage, C. Pellat, and J. Schacht, loc. cit.
(46) - A. Negev, Nabataean Archaeology Today, pp. 10,14-15
(47) - See A.S. Kapelrud, Baal in the Ras Shamra Texts, pp. 80-82
(48) - The Arabic sources relied upon by Wellhausen to say that Allah was always used as a proper name are, as seen above, necessarily suspect, and probably are the result of later redaction by Muslim theologians of a later day.