The Existence of Arguments in Classical Islamic Thought: Reply to Hannah Erlwein

In recent years, there has been notable interest in Islamic philosophy and theology from an analytic and not merely historical perspective. One important area of research that has garnered a great deal of research is the arguments for the existence of God. Recent work by Hannah Erlwein seeks to argue that this research has been in vain, for there are no arguments for the existence of God in classical Islamic thought. This paper analyzes Erlwein’s strategies in justifying this position, revealing that her research ignores an enormous amount of evidence that runs contrary to her thesis, in addition to demonstrating many of the errors and shortcomings in her work. Most disturbingly, the book seeks to present the Islamic philosophical tradition as fideistic and unintellectual based on a series of contrived interpretations of rather clear texts. In response, this paper demonstrates the importance of independent inquiry in the Islamic tradition by looking at a wide range of different relevant texts.

As part of her strategy, Erlwein implies that atheists did not exist in Islamic history, which may help explain why no one felt the need to prove the existence of God. Conveniently, however, Erlwein insists that this matter is beyond the scope of her book (Erlwein, 164), but a cursory glance at the most well-known works of doxography such as Shahrastānī's (d.1153) al-Milal wa-l-niḥal demonstrates that among the early Arabs referred to in the Qur'an, a group of them denied the existence of God, but this is nowhere raised in the book 3 . At one point, when she cites Ibn Ḥazm's (d.1064) doxographical work al-Fiṣal (Erlwein, 87), she conveniently ignores the fact that the title of that section, on the very page she cites, in bold type, is "The Second Division: Those Who Claim that the World Has Always Existed and It Has No Creator." Ibn Ḥazm then proceeds for the next twenty pages or so adducing proofs that in fact, the world has not always existed and therefore does need a Creator. 4 Erlwein instead tries to lead readers into thinking that Ibn Ḥazm (and Ashʿarī in this context) were only arguing against people who merely "call into question the belief in an originatedness world," (sic) (87); she completely ignores the fact that they also denied the existence of God. This is just one of many clear cases where Erlwein deliberately omits parts of the text that undermine her thesis.
The rest of the book is a long contrived attempt to apply her thesis to the works of several representative figures from the Islamic tradition, including al-Qāsim b. Ibrāhīm (d.860), al-Fuwaṭī (d.825), al-Naẓẓām (d.845), al-Khayyāṭ (d.913), al-Kindī (d.873), al-Māturīdī (d.944), al-Ashʿarī (d.936), al-Bāqillānī (d.1013), Avicenna (d.1037), al-Ghazālī (d.1111, Averroes (d.1198), andFakhr al-Dīn al-Rāzī (d.1210). In each chapter, Erlwein attempts to interpret all the authors as arguing for something about God's attributes and not about God's existence; given that these figures hail from different philosophical perspectives, there is no one line of thinking that unifies the book. In each chapter, rather, there is always a unique and creative twist to her interpretations, and as such, the work is highly idiosyncratic. It would have been useful to see a discussion of Jewish or Christian thinkers writing in Arabic, such as Saadia (d.942), Maimonides (d.1204), Yaḥya b. ʿAdī (d.974), Ibn Suwār (d.1017), Abū l-Faraj b. al-Ṭayyib (d.1044), Ibn Kammūna (d.1284), or Bar Hebraeus (d.1286), but one gets the impression that this would undermine a fundamental distinction underwriting her book; that is, the idea of a free-thinking Judeo-Christian West and its unthinking fanatical counterpart in the East.
In general, Erlwein's book exhibits major gaps in the evidence adduced for her thesis, questionable translations that indicate either an ignorance of the technical language or an intentional manipulation of the texts that is clearly disingenuous. In what follows, I will illustrate a number of cases of the omissions and obfuscations that sustain the 260 pages of Erlwein's work. Then, I will point out 4 well-known theses from the kalām tradition that were entirely omitted from her study despite the fact that each one of them is sufficient to demonstrate the falsity of her thesis: (i) the condemnation of taqlīd, that is, belief without proof. Since believing without proof was nearly universally condemned by the theological schools of Islam, it was a moral obligation on every capable person to reflect rationally and provide a proof for the existence of God; (ii) the tradition divides knowledge into what is noninferential and what is inferential (ḍarūrī and naẓarī, respectively). Beliefs which were noninferential are not ones which can be sought by inference, and therefore, the obligation to reflect does not pertain to them. Rather, the obligation to reflect pertains to the doctrines of revelation which were all inferential. The most relevant one to Erlwein's thesis is the existence of God, which by unanimous agreement in the kalām tradition, is an inferential proposition. That is, it is not self-evident and it requires an argument; (iii) the kalām tradition explicitly divides the doctrines of revelation into what can be known only by reason, what can be known exclusively by revelation, and what can be known by both. The paradigm example of what can be known only by reason is, of course, the existence of God. 5 It is thus untenable to claim that they took this belief "for granted" from the Qur'an, especially not at the level of argument; (iv) lastly, contrary to Erlwein's claims about the Qur'an and its interpretations by Muslim authors, I have included a final section showing that it was commonplace in the exegetical tradition to interpret the Qur'an as addressing atheism by citing 14 different exegetes in addition to other sources.

Some Illustrations of Gross Misinterpretation
One stark example of the kind of obfuscation in this book is Erlwein's interpretation of al-Bāqillānī's statement that "To believe in His unity is: to accept that He is real (thābit) and existent (mawjūd), and that He is the One, Unique God, like whom there is nothing," 6 Erlwein suggests the following interpretation: I propose that al-Bāqillānī's statement should in fact be translated as "He is established and an existent," rather than simply "existent." For his concern seems to be to address the hotly debated issue of whether it is right to say of God that He is "an existent." [...] It is evident that in the present context al-Baqillani declaration that God is "mawjud" does not refer to the question of whether God actually is part of reality; rather, God's being part of reality is assumed and the debate is about whether it is correct to apply certain terms to Him considering that they are also used to refer to creation, and tashbih or likening God to creation is sought to be avoided. (Erlwein 103-4) (sic) For one thing, the context does not involve any debate at all, let alone a debate about what words can be used to describe God. The "context" of Bāqillānī's statement is simply the listing of what every human being is obligated to know and believe, and it has nothing to do with a discussion about whether God can be called a "thing" or whether He may or may not be called "an existent," or whether non-existent objects should be attributed with some kind of reality. But, Erlwein sends her readers down a rabbit hole of irrelevant references before concluding that this is about denying "resemblance to creation." This is even more obvious given that in the previous page of the abovementioned citation, Bāqillānī states that the very first obligation on humanity is to have knowledge of God, which, further, can only be acquired inferentially by argument. 7 But, if Erlwein is right, then human beings are only required to believe that "God is one," but not that He actually exists. Indeed, if "God is existent" does not mean that He is real, it is unclear to me what kind of expression would satisfy Erlwein's requirements.
Another illustrative example is when Erlwein interprets Ashʿarī's argument for the existence of God as an argument for the "principle of causation," which is a rather impressive acrobatic maneuvre: Be this as it may, al-Ashʿarī's proof that creation has a creator has, in the first place, the objective of affirming the "principle of causation." This means that he seeks to establish, first, that effects have causes and, second, that these causes are external to them. (Erlwein,79) Of course, Ashʿarī begins this discussion by saying "If someone asks: What is the proof that creation has a Creator that created them, and a controller that controls them?" and then proceeds to provide a proof that there does indeed exist a Creator for the world. The objective is clearly not to affirm the principle of causation; a principle which many in the Ashʿarī school took to be a rational principle not in need of an argument. But, reading Erlwein's book without checking the reference, one would never know what the chapter heading was. Again, if her theory were right, then one might reasonably expect that there is a "Qur'anic pronouncement" that has ordered the Muslims to provide an argument for the principle of causation. The book is replete with these fanciful interpretations, more of which I will refer to below.

The Obligation to Reflect and the Rejection of Taqlīd
There are a number of central topics that are absent in Erlwein's analysis, such as the question of taqlīd, roughly, "imitation," which, in the context of theology, means to believe a proposition without proof. According to all of the major schools of Islamic theology (Ashʿarī, Māturīdī, Muʿtazilī, etc.), human beings are obligated to inquire into the doctrines of revelation and to believe them with proof. One of those beliefs, of course, is the belief that God exists. This means that everyone who meets the criteria for moral obligation (any sane adult capable of reasoning) must reflect on all of their beliefs and come to have knowledge of them, that is, to know why they are true. 8 Although Erlwein is aware of some secondary literature on the issue (206), the rejection of taqlīd and the obligation to reflect (naẓar) are absent from her book. She does not, for example, consider the consequence it has for her overall strategy, namely, that according to their own theology, Muslim theologians should have condemned themselves either for unbelief or sinning for taking God's existence for granted, or, perhaps more absurd, that Muslims are not required to believe that God exists at all, they are only required to believe that He has the "attribute of creator." Indeed, on the few occasions when Erlwein does mention taqlīd, she confuses the matter by translating it as "following others in their beliefs" (fn.659, 206) or "following authorities" (fn.148, 46), "blind imitation of authorities" (63), "blind acceptance of tenets from authorities" (182), and "blind following of authorities" (205). But, taqlīd is used in two contexts: one in contrast to ijtihād and one in contrast to taḥqīq. In the context of law (fiqh), it means to follow a legal authority, and in this case, most jurists held that it was obligatory to follow qualified legal authorities (for otherwise, one would end up in the chaotic situation where each individual would either follow unqualified legal authorities or make up laws for themselves). In the context of theology, however, taqlīd means belief without evidence or proof. It is immaterial whether this means simply believing whatever one's parents believe, or one's society, or the local jurist, or following mere fancy. The overwhelming majority of the kalām tradition holds that it is morally obligatory to reflect and inquire into all matters of belief, including the origination of the world and the existence of God. 9 An important consequence of this is also Erlwein's emphasis on "worship" as the final objective of all these arguments, without noticing that in Islamic law, reflecting on proofs for God's existence is an act of worship, and that is what it means to be "the first obligation" on humanity, as is well known to anyone who has read an introductory section of a kalām manual. 10 It is worth noting that Erlwein's use of the term "objective" in her book is also equivocal, and it is one of the ways she obfuscates the issue. In the introduction, she appears to mean that the objective of an argument is its conclusion. But since the conclusion to all the arguments in the tradition is that "God exists," she shifts her usage of the term to mean their overarching purpose and takes license to interpret them as meaning that God is unique, or incorporeal, or the Creator (and therefore, the only one worthy of worship) (217-220). But, this is no longer the same use of the term "objective," and it certainly creates problems for her main thesis, which likewise fluctuates for convenience, from saying that (1) arguments for God are absent, to saying (2) arguments are present, but they have different objectives.

Knowledge of God's Existence is Inferential
The second major topic that Erlwein omits is tied to the first, and it derives from the epistemological principles stated in the introduction to the overwhelming majority of theological and philosophical summae: the distinction between noninferential knowledge (ḍarūrī, badīhī) and inferential knowledge (naẓarī, istidlālī, or iktisābī). 11 The first category includes a priori truths, such as the principle of non-contradiction, sense perception, one's knowledge of one's own existence and internal states (such as anger, sadness), and widely transmitted reports ("so-and-so is the President of the USA"). Noninferential knowledge does not require reflection or inference; all one needs in this case is to be exposed or attentive to the relevant propositions or sensible objects in order to have knowledge of them. Inferential knowledge, on the other hand, requires thinking. Indeed, since all the tenets of revelation were inferential, the obligation for reflection mentioned above is invoked. If, however, the tenets of revelation were noninferential, it would be quite problematic for one to think that reflection was a legal obligation. This is especially critical because all theologians of the major schools of kalām and the Peripatetics held that knowledge of God is inferential; that is, the only way to have knowledge of God is by rational inference. 12 In one instance, Erlwein does mention Bāqillānī's crystal clear statement that God's existence can only be known inferentially, but she drops an important detail in an attempt to twist the text into meaning something different. Compare the following two translations: Erlwein's translation (100) Bāqillānī's statement (my translation) he has to know that the first thing God has made obligatory upon all humans is speculation about His signs, pondering over the things He has power over (maqdūrātihi), and reasoning towards Him based on the traces (āthār) of His power and the witnesses (shawāhid) to His rubūbiyya, for He is not known necessarily and not observable by the senses; His existence and being (wujūduhu wa-kawnuhu) are only known by the compelling proofs contained in His deeds.
And [the moral agent] must know that the first obligation placed on all humanity by God is to reflect on His signs, and consider the objects of His power, and infer His existence by the effects of His power and the witnesses to His godhood, for Hethe Exaltedis not known by necessity, nor is he perceived by the senses, but rather, His existence, and His being in the manner entailed by His acts, can only be known by powerful arguments and irrefutable proofs" 13 (wujūduhu wa kawnuhu ʿala mā taqtaḍīh af ʿāluhu bil-adilla al-qāhira, wal-barāhīn al-bāhira) Despite the unambiguous nature of Bāqillānī's text, Erlwein seeks to convince us otherwise. Notice that Erlwein elides an important clause in this sentence, namely "His being in the way entailed by His acts," writing instead "His existence and being are only known by the compelling proofs contained in His deeds." Whether it was intentional or not, this is a mistranslation of the text which renders the verb kāna ("to be") as a complete verb, as opposed to an incomplete verb which requires a predicate to complete the statement. She then attempts to exploit this mistranslation in her analysis of the text, as we will see below. Notice that Bāqillānī affirms two general points: (i) God's existence and (ii) God's being in a manner entailed by His acts. The first refers only to God's existence, and the second refers to the additional attributes that are implied and inferred from His acts; His acts, to be sure, refer to the created world, for in Ashʿarī terminology, God's acts are the world. 14 But, Erlwein attempts to convince her readers that this in fact does not mean that it is obligatory on human beings to prove God's existence at all; in fact, it does not even mean "existence": God's "existence" (wujūd) in the present context has a different meaning and refers to something else. Rather than referring to the dichotomy between existence and non-existence in the question whether God exists, it refers to the classification 12 , 28, 56, 64, 66, 91, 99, 230, 234. of the kind of existence that belongs to God. We need to bear in mind that many mutakallimun were eager to stress that God's existence is of an entirely different kind than the existence that charaterises created things. (Erlwein 101) (sic) According to Erlwein, then, the very first obligation on all of humanity is to reflect on the signs of the world in order to know that God is not a body (101). This seems like a strange first obligation to pick. According to her, the "present context" of Bāqillānī's statement is to be found in a section 13 pages later, where Bāqillānī is elaborating an entirely different point, namely, that God is entirely unlike other things; in this vein, she also cites Shahrastānī's kalām summa (who lives over a century after him) in support of this point without any justification, and where he is likewise discussing a wholly irrelevant point (101). This is another case of deliberately misleading the reader, for the actual context, as we saw above, refers to what is obligatory on humanity to know and believe. Furthermore, she does not consider the following objections: (i) even if we concede that it was about the kind of existence, there appears to be a conceptual confusion, for we can still ask: these "kinds of existence," do they oppose nonexistence or not? If they are "kinds of existence," one would hope that they exist, so this requires an explanation, and (ii) there is an explicit discussion in the kalām tradition on whether a corporealist can be said to believe in God's existence, and many authors held that corporealism does, in fact, entail atheism. 15 This position is not difficult to see, for the basic argument for God's existence is that the world is composed of physical objects and all physical objects need a cause to bring them into existence. But, bodies cannot create other bodies, so the only way to explain the existence of the physical world is to say that it came into existence randomly; alternatively, one can state that the cause is a body, but it would need to be caused by another prior to it, and so on, ad infinitum. In either case, nothing recognizable as God would exist.
Therefore, in order for a cosmological argument to succeed, it must give us the existence of a cause that is incorporeal, that has no contingent properties, that has causal power, and so forth. Furthermore, Erlwein does not address the following question: if the proofs for the existence of God are in fact arguments for His incorporeality, why do all of these authors have separate chapters to show that God is not a body and unlike creation? That is, in every single kalām manual, there is a chapter discussing the existence of God, usually from the origination of the world or its contingency-and then a separate chapter on God's negative attributes, such as being incorporeal, and so 15 For example, Ashʿarī states in the Mujarrad Maqālāt that "The belief of one that God is composed of many elements and parts and is not a single entity [is one who] does not know God, and it is impossible for one to be ignorant of God if they know that God exists, just as the case that one who believes that Zayd, whom he observes to be composed of parts, is the Eternal being who has always remained an existent being, then he certainly has no knowledge of God, and has disbelieved in Him, because God is not Zayd." Mujarrad Maqālāt, ed. Daniel Gimaret, 227; similarly, Muhammad b. Yūsuf al-Sanūsī, one of the definitive Ashʿarī theologians of his day writes that "among them are those who believed in this false proposition (i.e., that every existent entity is a body) and it led him to atheism, that is, negation of the existence of the God entirely (nafy wujūd al-ilāh aṣlan) and that the worlds came into existence by complete chance without a cause, because when it became clear (to them) that for all temporally originated entities, any causal agent among them must be a body, they analogized without warrant and said: if the worlds had a causal agent, it would necessarily be a body, but it is impossible for bodies to create other bodies and most other properties; therefore, the bodies that compose the world have come into existence without a cause," see Muḥammad b. Yūsuf al-Sanūsī (n.d.), Sharḥ sughra as-sughra, 18, first edition, (Cairo: Matba'at al-taqaddum al-ʿilmiyya, 1322AH/1905CE).
forth. Have they been aimlessly repeating themselves in two distinct chapters of their books for a period of 500 years? I think that is most unlikely.
Erlwein even suggests that R. J. McCarthy, the editor of Bāqillānī's Tamhīd, was "misleading" his readers by entitling the section "On the Existence of God and His attributes" (89). If Erlwein is right, however, I think we would be more in the right to accuse the Islamic tradition of "misleading" readers and themselves into thinking they have arguments for the existence of God, not least every single academic who has worked on the question.
In another instance, Erlwein tries to discount an explicit statement from Fakhr al-Dīn al-Rāzī that contradicts her thesis: "Since knowledge of [God's] existence is not immediate (ḍarūrī), rather it is based on reasoning (istidlālī), [God] reports here what proves His existence" (Erlwein, 217). She argues that perhaps this is just a contradiction (217) or Rāzī is interested in rejecting corporeality, rather than proving the existence of God. 16 These are hopeless explanations of Rāzī's statements, not least because he has authored, elsewhere, an entire work on the refutation of corporealism, Asās al-Taqdīs, which does not, in fact, rely on the arguments for the existence of God. 17 Unsurprisingly, despite the notoriety of this work, it does not even show up in Erlwein's bibliography.
Indeed, Ashʿarī presents four pages of arguments for why the existence of God is an inferential proposition and therefore requires proof, some of which is worth citing here: If [knowledge of God] were noninferential, it would not be possible for fleeting doubts to come to mind, or for something to incline one to its denial, because what we know noninferentially is such [that it cannot actually be doubted]. Since we have seen that doubts regarding knowledge of God do in fact occur and inclines one to denial, we conclude that [knowledge of God] is not noninferential.
[In contrast], since man's knowledge of himself must be noninferential, it is not possible for something to incline him to its denial, and it is not possible for doubts regarding his existence to occur to him such that he comes to believe that he does not exist. And since it is possible for man to go from belief in God to disbelief in God, and to abandon knowledge of God, it becomes known that [knowledge of God] is not noninferential. 18 Again, even though Erlwein is aware of this work, these sections are simply not discussed. Either she did not read the books she cites, or, she read them, and decided to ignore excerpts that undeniably refute her thesis.
In the chapter on Ibn Sīnā, Erlwein resorts to similar tactics. Despite acknowledging with great difficulty that Ibn Sīnā says the existence of God is not self-evident, but 16 Erlwein does not seem to consider the quite evident fact that the argument from the temporal origination of the world gives you the existence of a cause, who must also be necessary, eternal, incorporeal, and so forth. There is no conflict between a concern with existence and a concern with incorporeality. Indeed, for the argument for existence to succeed, the cause must be incorporeal. Furthermore, it is quite evident that all kalam manuals clearly distinguish the arguments for the existence of God from the arguments for God's transcendence. requires proof (112-113), she makes an unjustified segue into the subject matter of physics, stating that: …unless physics can take it for granted that at least a part of reality is of corporeal existence, it would not be able to claim a subject matter for itself [...] In analogy to the way Ibn Sina speaks of the existence of the subject matter of physics (that is, "the kind of existence which is characteristic of it"), I suggest he speaks of "wujud al-ilah" which has to be established by metaphysics (115-116) (sic). Erlwein here equivocates between (1) the investigation into the "kind of existence" that bodies have in metaphysics and (2) the investigation into whether God exists in metaphysics. Somehow, since physics can take its subject for granted (perhaps following "the pronouncements of the Qur'an"), then metaphysics, which investigates God's necessity, likewise takes His existence for granted. Erlwein does not consider that perhaps the reason why physics can take the existence of its subject matter "for granted" is because the existence of its subject is self-evident: it is plainly clear, by sense perception, that bodies exist all around us. Ibn Sīnā's point is that God's existence, in contrast, is not self-evident, and therefore, God cannot be the subject matter of metaphysics, because the subject of every science must either be self-evident or proven by a higher science. Since there is (1) no higher science than metaphysics and (2) God's existence is not self-evident, it follows that God cannot be the subject of metaphysics.

Reason and Revelation
Another important omission in Erlwein's work is any discussion on the epistemic relationship between rational knowledge and revealed knowledge. Ashʿarī writes the following: It is necessary in inquiry that one begins their inquiry into rational matters (al-ʿaqliyyāt), because they are the foundation, and revealed matters (al-samʿiyyāt) are the branch; the foundation is first, and the branch is second, and it is necessary to establish the first before the second. Then, if one has completed (their inquiry) into what is required by reason, they move on to what occurs by revelation. 19 In keeping with this approach, all of the major schools of kalām divide the propositions of theology into three according to their epistemic priority: (1) what can only be known by reason, (2) what can only be known by revelation, and (3) what can be known by both. 20 The most common example cited by these authors regarding what can only be known by reason is the existence of God. 21 Despite this division being mentioned in many manuals of kalām and uṣūl al-fiqh (the science of legal principles), by the very authors discussed in her book, this tripartite division is nowhere mentioned. Of course, the main motivation behind this division is to avoid circular reasoning. That is, it is incoherent to accept the truths of 19 1Ibn Fūrak (1987), Mujarrad maqālāt al-Ashʿarī, 319, lines 21-23. 20 See, for example, al-Bāqillānī (1998), al-Taqrīb wal-Irshād, vol.1, 228;al-Juwaynī (1978), al-Burhān fi uṣūl al-fiqh, vol.1, 136-7;Fakhr al-Dīn al-Rāzī (2015), Nihāyat al-ʿuqūl, vol.1, 142. Rāzī in facts says it is impossible to know God without inference-so if we assume that he did not have an inference, then he must not have known God and, thus, is an atheist as well. 21 Ibid. revelation unless one has first ascertained that something such as revelation is possible, which requires, first and foremost, that God exists, that he is living, powerful, knowing, willing, and so forth. 22 Thus, the pivot for this tripartite distinction is the affirmation of prophethood. That is, the truth of revelation can only be established if the authenticity of prophethood is established, but for the prophethood to be established, one must establish first the existence of God, His attributes, and what is possible with regard to His acts. So, not only is it false that these authors took the existence of God for granted because they "followed the Qur'an's pronouncements" per Erlwein, they were explicit in holding that one cannot accept any data from the Qur'an, let alone God's existence, until one has first proved that God exists, has power and will, and that the specific human being in question who relayed the Qur'an, was in fact truthful. Only after prophethood is established can one then take propositions from revelation.

Qur'anic Exegesis
Recall that Erlwein claims that the authors did not address atheism, because they followed the Qu'ran, which presumably did not address atheism. It is important to note that the majority of the Islamic tradition did in fact hold that the Qur'an addresses the question of atheism. Shahrastānī (d.1153) (who is only briefly mentioned in the book) mentions that the inhabitants of Arabia before Islam were divided into three groups: (1) those who denied the existence of the Creator and denied resurrection; (2) those who only denied resurrection; (3) those who only denied prophecy. 23 According to Shahrastānī, the first category "denied the creator and resurrection, they believed that nature gave life, and [the passage of] time negated it, and they are those who the Majestic Qur'an described: 'And they said: There is nothing but our life in this world, we die and we live' (Q. 45:24) indicating (that there is nothing but) the sensible natures in the lower world, and that life and death is limited to their composition and decomposition. So, what composes is nature, and what destroys is [the passage of] time, "And nothing destroys us but the passage of time; but they have no certainty regarding that [which they say], they are but speculating" (Q. 45:24). So, [God] argued against them with self-evident truths and intuitive Qur'anic signs in a number of verses and chapters..." 24 Shahrastānī is not unique in this affair either. Another major author Erlwein discusses is Māturīdi, who discusses the dahriyya as those who deny the existence of God in his eighteenvolume commentary on the Qur'an at least twenty-two times, and at several instances explicitly acknowledges that some of the Arabs of Mecca did not accept the existence of God. 25 Take the following example from Māturīdī's commentary on the Quran (4:150-1): His statementthe Exalted and Transcendent -'Verily those who disbelieve in God and His messengers and seek to separate between God and His Messengers' may 22 Ghazālī presents the same progression from the pure reasoning of the mutakallim until he arrives at knowledge of the truth of prophecy and can finally accept revelation, see his al-Mustaṣfa fi uṣūl al-fiqh, vol.1, 13-4. 23 : vol.4/94; vol.5/8, 26, 158-9, 390; vol.8: 36, 238-9; vol.10: 28, 58, 291-2, 399; vol.11: 405; vol.12: 80, 142, 363; vol.15: 81; vol.16: 253; vol.17: 401. accommodate two interpretations. It may be that the [conjunctive wāw] means "or", as if God had said: Verily those who disbelieved in God and His Messengers or seek to separate between God and His Messengers,' and therefore, His statement 'disbelieve in God' refers to the atheists (al-dahriyya), because they deny God and do not believe in Him, and they believe in the eternity of the world, and this part of the verse refers to them. 26 It is untenable for Erlwein to deny texts like these or to make readers think that these are negative terms used to slander their opponents who were also Muslims; for even if we accepted this kind of explanation in a specific use of the term dahriyya, it still requires that Erlwein concedes that the term does primarily mean atheist; otherwise, it could not even be used as a term of abuse between Muslim interlocutors. In any case, for the most part, it is quite clear that the dahriyya are atheists. For example, Māturīdī makes the following remarks on Qur'an (15:44), which mentions 7 gates of Hell: And if that is the case, then the seven gates mentioned here are for the people of Unbelief (ahl al-Kufr) while the sinners do not enter them (ahl al-kabāʾir). It may be that one gate is for the sophists who deny the existence of the world: what is sensible and what is not, they affirm nothing; and another gate for the atheists (aldahriyya) who are those who deny the existence of the Creator. 27 Indeed, we can turn to just one verse on the topic (Q. 52:35 "Or were they created by nothing, or were they the creators?") 28 and see that almost every major commentary considers this an argument for the existence of God against the Arabs who, as atheists, denied it. 29 Similarly, al-Juwaynī pits himself explicitly against atheists (al-mulḥida and al-dahriyya) when he states: 26 Ibid,vol.4,94. 27 Ibid,vol.8/36. 28 One can examine available commentaries for other similar verses: (Q. 14:10: "Is there doubt in God, the Creator of Heavens and Earth?"); (45:24: "And they say it is but our worldly life, we live, we die, and nothing destroys us but the passage of time"); Commenting on verses (26:23-31 "And what is the Lord of the Worlds?") Rāzī explicitly considers the possibility that Pharaoh did not believe that there was a God at all, and that the world was necessary of existence, and the motions of the planets were the causes of everything, see: Rāzī, al-Tafsīr al-kabīr, vol.24, (Beirut: Dār al-Fikr, 1981), 127-128. Despite this part of Rāzī's exegesis being cited by Erlwein, these particular passages are again deliberately omitted. The second division of their objections connects to their attempts to deny a Creator, and in that, they have two ways: one is to argue that to affirm the existence of an entity which is self-sufficient and transcends having contingent properties, as held by the people of Islam, is unintelligible. And the other ways are related to justice and oppression, for they may state: an action which does not contain harm or benefit for the agent, nor does it prevent harm from them, is an act of frivolity and stupidity, and so on. 30 We know he is referring to atheism here because of the arguments that they are presenting. The first is to say that the notion of a self-sufficient entity that creates the World and is yet utterly transcendent is unintelligible. The second argument roughly amounts to an argument from evil by inferring from the existence of evil or trivial things in the world to a contradiction in how Muslims have understood God. None of these arguments can be attributed to any of the Islamic sects.
By itself, the presence of atheists in Arabia or the Islamic world during the classical period does not prove anything about whether the Islamic tradition adduced arguments against it or not. Nevertheless, showing that they did exist sheds light on some other biases held by Erlwein, namely, that even doubt and atheism is a strictly western or Christian possession. This leads to another important point. Although Erlwein cautions care in how we define atheism when we study Islamic philosophy, she goes on to use, without much justification, a very specific modern understanding of atheism in order to settle the question and declares that using any other definition would be "fallacious" (13). So, the only person who counts as an atheist, it would seem, is someone who affirms that there is no God. But, the Islamic tradition sees things differently, for there is more than one way one can be a non-believer in the existence of God. For example, one could simply have absolutely no knowledge of God, let alone explicitly affirm His existence or deny it, and would therefore be a non-believer. Similarly, one could be unsure if there was a God, an agnostic, in which case, they would be a non-believer as well. Indeed, it would be more relevant to seek that definition in the tradition: how would they understand one who meets the Islamic legal conditions of who counts as someone who does not believe in God? Since all schools of Islamic thought required that moral agents must have certain knowledge (ʿilm) of God's existence, which is normally not possible without a proof (especially if that person was not socialized into believing it), it follows that (i) even one who opines (ẓann) that God exists is considered a non-believer. Therefore, (ii) it is obligatory upon them to find proof for God's existence that gives them certainty and eliminates mere opinion. Thus, the key concepts of "atheism" and "worship," upon which Erlwein's strategy hinges, have no meaning for the authors she discusses. 31

Conclusion
Given how bizarre Erlwein's thesis is, it is worth asking about its origins. She makes it appear as though this conclusion slowly revealed itself to her after years of careful study. 30 Al-Juwaynī (1969), al-Shāmil fī uṣūl al-Dīn, ed. Nashshār, 223. 31 Similar problems arise for Erlwein in her use of the terms rabb (god, lord, master, caretaker, etc.) and ilāh (divinity, deity, god, etc.), where she does not actually ask the authors themselves what these terms mean in any rigorous manner.
More likely, however, is that Erlwein adopts the position from Ibn Taymiyya, whose quarrel was primarily with Fakhr al-Dīn al-Rāzī. Following the standard view of most schools of kalām, al-Rāzī argues that reason must always be given epistemic priority over revelation, because the veracity of revelation can only be known by reason; so if we allow revelation to undo reason, then we will also undo revelation itself (as implied from Reason and Revelation above). 32 In contrast, Ibn Taymiyya argues that this juxtaposition is false but only on his narrow reading of reason. In direct connection to this is his unique position that there is no argument for the existence of God in the Qur'an, indeed, we do not need any arguments, because the natural disposition of man (fiṭra) already inclines to the truth of the proposition. 33 Here we find the one lone author that fits Erlwein's reading of the Islamic tradition, yet despite how similar Ibn Taymiyya's position is and how widely it has been discussed in the secondary literature, he is conspicuously missing from the entire book. 34 But, Erlwein is clearly aware of the terms of Taymiyyan theology, especially his distinction between tawḥīd al-rubūbiyya and tawḥīd al-ulūhiyya, as found in Ibn ʿAbd al-Wahhāb (d.1792) (Erlwein, 49). 35 It is highly unlikely that the only noteworthy figure in the Islamic tradition who holds this view has escaped Erlwein's attention, especially given that her doctoral adviser, Mustafa Shah, has a long paper on him, in addition to editing a volume which included Griffel's article cited above. More likely, Erlwein took Ibn Taymiyya's thesis and suppressed mention of him. Indeed, mentioning Ibn Taymiyya would cause a lot of trouble for her thesis, because he explicitly acknowledges that the kalām tradition did have arguments for the existence of God, and the whole point of his project was to show that arguments for the existence of God were not needed, and indeed, harmful.
In conclusion, Erlwein supports her thesis not with arguments so much as with a strategy of omission, mistranslation, and contrived interpretation. If this will not convince the thinking reader, it may at least baffle them. The only discernible takeaway from this book is that Muslim thinkers were fideist, anti-rational, and unusually concerned with incorporealism. She does not pretend to give us an analysis of one thinker or the other: she has set out to characterize the totality of Islamic thought. She would have us think that Muslim thinkers, no matter their school or approach, are simply zealous followers of the Qur'an and that they lack the true philosophical sensibilities found in their Christian and western counterparts. Knowledge of the tradition tells a very different story: we find one of the longest concerted efforts at working at these proofs in human history, to the degree that it led one eighteenth century Ottoman commentator, Mehmed Saçaklızade (d.1732), to express his irritation with just how concerned scholars were about it. I end with his statement: And among the works that discuss some questions of kalām is the treatise 'On Demonstrating the Existence of the Necessary Being' by al- Dawānī (d.1502). And it has a commentary (sharḥ), and a gloss (ḥāshiya), and some students spend a year studying this text, and it only contains one question, that is, that the world has a God who necessarily exists, with long-winded and weak proofs, and excessive amounts of argumentation that result in nothing but the undermining of belief and the conjuring of deadly doubts, and who has doubt in God the Exalted?: "Is there doubt in God, the Creator of heavens and earth?" (Q 14:10), so what chance is there for one to achieve certainty from such a treatise? Rather, studying such texts results in doubt for people of certainty, and increases the doubts of those who already doubted. 36 Funding Open Access funding enabled and organized by Projekt DEAL.
Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/.