Ibn Sīnā presents his arguments for God’s existence in a number of his writings, mainly in Metaphysics of the Cure (Al-Ilāhiyāt al-Shifā‘), The Deliverance (Al-Najāh), and The Pointers and Reminders (Al-Ishārāt wa al-tanbīhāt). These works address many other topics as well, like logic, mathematics, and natural philosophy. Ibn Sīnā employs a range of different arguments in these writings: some of which argue for God’s existence from motion (inspired by Aristotle), others from causation, and even others from composition (which is an important corollary to arguments appealing to causation). In this chapter, I will briefly outline Ibn Sīnā’s arguments from these sources, referring to both the relevance and function of his arguments in relation to contemporary debates between atheists and theists. The aim of Chapters 1 and 2 is to demonstrate why this particular rendition of the argument is particularly valuable in these debates. The intention is to provide theists intellectual ammunition that is currently either under-developed or completely lacking in both academic literature found in Philosophy of Religion publications and the most famous public engagements between theists and atheists in the last century. When reading Ibn Sīnā, one must keep in mind that he often switches from informative to persuasive writing. Accordingly, one must be careful not to conflate Ibn Sīnā’s logical and metaphysical taxonomies (which are intended as matter-of-fact clarifications) and his formal arguments. As an example, the 6th chapter of Ibn Sīnā’s Metaphysics of the Cure has been misunderstood to be an argument, where Ibn Sīnā’s intention was most probably simply taxonomical. On this point, Daniel De Haan states:

…Avicenna’s analysis of necessary and possible existence in Ilāhiyyāt I.6– 7 could not be a formal demonstration for God’s existence since these chapters are oriented towards providing us with insights into the proper first principles of metaphysics.5

Other scholars of Ibn Sīnā, such as Fazlur Rahman6

and Parvis Morewedge7

regard this kind of classification as an argument in and of itself.8

In what is perhaps the most comprehensive, yet concise passage written by the Persian philosopher demonstrating God’s existence, we find Ibn Sīnā state in The Deliverance: There is no doubt that there is existence. Everything that exists is either contingent or necessary. If it is necessary, then the pursuit of the necessary existence is complete. If it is contingent, I will make clear that this contingent existence will ultimately return back to a necessary existence. Before this, I will present premises (to prove this thesis). Among these premises is [the assertion] that it is impossible for an infinite regress of causes to account for a contingent existence. That is because all contingent existences either all exist at once or do not exist at once. If they do not all exist at once and they are infinite, one preceding another, we will deal with this matter in another section of this book. If they are all together and there is no necessary existence in the set, [we find that] upon exhaustive analysis [and whether or not it is finite or infinite] the set [of contingent things] can either be composed of contingent things or necessary things. If, in the set, there is a necessary existence and all other things in the set are contingent it will be such that the necessary existence overrides all of the contingent existences. If the set is composed of only contingent existences, then it requires that which supplies existence. This can either be in the set of contingent existences or outside of the set. If it is inside the set, then this has already been elaborated upon. Or, [it could be] that it is a contingent existence such that it is the cause of the set, and the cause of the set is by extension the cause of all of its constituent parts. If it is inside the set [or is the set itself] – even though this is impossible – this line of argumentation could still be valid, so from one perspective then one will have fulfilled the object in making this conclusion (i.e. because this would have proven the existence of the necessary existence). This is because anything that is self-sufficient (i.e. the set in question) is a necessary existence. And the necessary existence cannot be this; it is also impossible that a contingent existence can exist outside the set, as the set is by definition a collation of all contingent existences. Therefore, it (the supplier of existence) must be outside of it as well as necessary in essence. Thus, the set of contingent existences has culminated in the need for a necessary existence outside of the set in order to explain it. It is not the case that every contingent effect has a contingent cause as a matter of infinite regress.9

The first stage of Ibn Sīnā’s discourse is to establish ‘existence’ as the most foundational, transcendental, and universal category of analysis. The first major postulation which Ibn Sīnā makes is to state that ‘it is not possible for there to be an infinite regress of causes for a contingent existence’.10 Ibn Sīnā employs a set theory type of reasoning for this, gathering together all members of a specific description in one set and analysing that set thereafter. On this point regarding ‘possible’ or ‘contingent’ existences, Ibn Sīnā states:

If they are all together and there is no necessary existence in the set, [we find that] upon exhaustive analysis [and whether or not it is finite or infinite], the set [of contingent things] can either be composed of contingent things or necessary things.11

Simply put, a set of one type of things excludes other types. For example, a set of chairs excludes tables. Likewise, a set of pens excludes pencils. Ultimately, this means that a set of contingent existences excludes necessary existences and vice versa. Ibn Sīnā states that if somehow a necessary existence can exist in the set of impossible existence (though this is logically impossible), this would be a counterintuitive proposition to make for anyone who aims to deny such a necessary existence. Ibn Sīnā goes on to suggest that just as it is as impossible for a necessary existence to be represented in a set of only contingent existences, it is also impossible for a contingent existence not to be represented by the set of all contingent existences. With this groundwork put in place, Ibn Sīnā concludes that ‘the set of contingent existences has culminated in the need for a necessary existence outside of the set in order to explain it’.12 Summarising the argument further, Herbert Davidson mentions Ibn Sīnā’s argument as pointing that ‘possibly existent beings are traceable to a necessary existence (by virtue of itself)…something exists which is a possible existence by virtue of itself; therefore something exists which is traceable to a necessary being by virtue of itself’.13

Notwithstanding this, it is perhaps a good place to start understanding the metaphysical compartmentalisations and how Ibn Sīnā categorises existence, as this effectively lays the groundwork for his argument. Regarding this matter, Ibn Sīnā starts the 6th chapter of Metaphysics of the Cure by defining what constitutes ‘necessary existence’ by stating:

So we say that the necessary existence is uncaused, whereas a ‘contingent existence’ is caused, and that the necessary existence is necessary in all possible ways and perspectives conceivable. Its existence cannot be a result of the existence of anything else. If that were so, it would be as if each of those two things (i.e. the supposed necessary existence in question and the thing which it results from) are equal in terms of existence and interdependent of each other. And it is not cogently possible for the necessary existence’s existence to be as a result of many things. The necessary existence cannot cogently be ‘the reality’ that has any common aspect of it (with that which is not necessary). Thus, in order to be classified as ‘necessary’, the necessary existence cannot be added upon, constructed, mutable, divisible, or one of multiple contributors to its own existence which is specific to it.14

In The Pointers and Reminders, Ibn Sīnā starts off with a classification of that which is possible or ‘contingent’, muḥāl or ‘impossible’ (like a square circle), and necessary.15 On a similar note, in his Metaphysics Ibn Sīnā moves on to speaking directly of the modal specialities of the necessary existence: As for the necessary existence, it has no cause. This is because, if it had a cause, its existence would be because of it. Anything which is considered essentially by itself and its existence is not necessary in itself cannot be a ‘necessary existence’…so it has been made clear that a necessary existence is uncaused. As a corollary to this, it is not possible that a thing is a necessary existence in and of itself and also because of another thing…Moreover, [regarding] anything that is a contingent existence in essence, both its existence or non-existence would come about because of a cause. This is because if the [contingent] thing can be located, then its existence can be discerned from nothingness (or lack of its existence), and if it is not in existence, then its non-existence would be discernible from its existence. So, an exhaustive rendering of the options would suggest that its existence or lack of existence would either be caused because of something else or not. If it is from something else, then that something else is the cause of it. And if it is not from something else, then it is itself the necessary existence.16

In The Pointers and Reminders, Ibn Sīnā talks about a jumlah (roughly translated to a sequence or even a set) consisting of an infinite number of contingent things, arguing ad absurdum that such a set cannot be the necessary existence. Applying this to arguments posited by New Atheists, this part of the argument is one of the most potent and useful proofs in arguing against an infinite multiverse thesis. This thesis is often employed as a substitute for a necessary existence or ‘God’, a point that will be expanded on in following chapters. In the Metaphysics, Ibn Sīnā is most explicit in making a formal argument in maqālah (section) 8, chapters 1-3. It merits consideration that Ibn Sīnā effectively ends his book with this argument, and in many ways earlier chapters seem to build up to this point. In this section, Ibn Sīnā argues ad absurdum for the impossibility of an infinite regress, doing so in two novel ways highlighted below: first, arguing for the impossibility of all ‘middle causes’ having the same modal status, and second, employing an argument from composition. As will be explained, these types of arguments are perhaps best placed to deal with atheistic interrogations which stipulate an infinite multiverse. After conceptualising a set of all existent things (the aforementioned jumlah), Ibn Sīnā posits that: [The] ‘ultimate cause’ cannot be the last in the set of causes, nor can it be in the middle. That is because the middle cause in the set can only cause one effect. In addition, the effect does not cause anything. And each thing of the three types at each extremity in the set [of existent things] has a specialised modal status. Thus, the specialised modal status of the thing at the end of the set is that it is not a cause for anything [by definition].17

To illustrate this argument, conceptualise a linear set from P to P10. Ibn Sīnā is simply stating that by logical necessity P5 can only be a cause for P6 in the set, and not for those things which come before it. P10 is not the cause for anything in this set. He continues this line of reasoning by stating: At the other extremity of the set (i.e. at the beginning of it) the cause (by definition) is the cause for all other things in that set. The specialised modal status of the thing which exists at the middle of the set is that it must be caused by one thing and be the cause for one thing.18

If the set is finite and linear, P5 must be caused by P4 and must be the cause of P6; this is the case even if the elements between P1-10 are multiple, for as Ibn Sīnā suggests: That is [the case] regardless of whether the thing(s) at the middle is one thing or more than one thing. If it is more than one thing(s), then it is either finite or infinite: if it is finite, then the set is between two extremities (i.e. the first cause and the final effect). Therefore, each of the things in between these two extremities will have a specialised status (which corresponds to the finitude of the set).19

Ibn Sīnā expands this argument of an infinite set in both The Pointers and The Deliverance by considering a circular finite set. Herbert Davidson addresses this point in his work on medieval Kalām (scholastic theology) entitled as Proofs for Eternity: …a self-contained circular regress is shown to be absurd by an argument applying only to it. In the circular regress x y z, x would be a distant cause of z, and z would be the immediate cause of x. x would consequently be a distant cause of itself, which Avicenna regards as absurd. By the same token, x would be a distant effect of itself, which is equally absurd. And the point can be made again in a slightly different way, as follows: x would be dependent for its existence upon something-z-whose existence is posterior to it. But “when the existence of something depends upon the existence of something else that is essentially posterior to the first, the existence of the first is impossible.” A self contained circular regress of causes cannot, therefore, exist.

Having dealt with a finite set, Ibn Sīnā moves on to address the infinite regress objection, stating: If the set, on the other hand, is infinite, then such extremities (the first cause and final effect) will not have been realised, and all things in the set will have the [equivalent] status of the things between the two sets (i.e. they will not be first causes or final effects).21

On this rendering – and returning back to our P1-P10 example – P1, P5, and P10 are meaningless in an infinite set. All Ps have the same modal specification, and they cannot possibly be considered ‘initiators’, ‘middle points’, or ‘final effects’. Ibn Sīnā elaborates on this point by stating: [Regarding] this infinite set, if you add or subtract from it, its status as ‘infinite’ will remain the same. In this case [of the infinite set], it is not possible for the set of causes to exist without having a cause that is itself uncaused and originating. Accordingly, all of the things in the set will have the status of the middle things in the set (i.e. being neither final effect or first cause), and this is logically impossible.22

Ibn Sīnā, like Aristotle, was an eternalist. In other words, he actually believed that the universe was pre-eternal. This startling fact puts him at odds with many Islamic thinkers who opposed such notions (most notably al-Ghazālī). Despite believing in an eternal universe, Ibn Sīnā, like Aristotle, argued quite clearly for the impossibility of an infinite regress of causes.23 In addition to the argument from middle causes, Ibn Sīnā makes an argument from composition in the following manner: So we say [that] the necessary existence cannot be considered a ‘composite construction’ such that there is a certain ‘whatness’ attribute [attributed to it], and that such ‘whatness’ is itself the quality of necessity…it would be logically impossible that in such a ‘whatness’ there is an actual reality, for if it had a reality and that it was differentiated from the necessary existence then it would indicate that something unnecessary had brought about something necessary. And this would indicate that the thing in question is not, in fact, necessary.24

The argument from composition suggests that anything which is composed is generated and is therefore dependent. If there is a quality within something that makes it ‘necessary’ and capable of being extracted from while also being characterised with differentiation, then it would suggest that the entity in question is in need of it. As such, this would be enough proof to indicate the contingency of that entity. Taking Ibn Sīnā’s premise that any composed entity that exists by its parts (and not by virtue of itself) is composite and therefore any composite is dependent,25 the prospect of an infinite number of contingent things in existence becomes an impossible prospect to maintain. Hebert Davidson summarises this argument in the following way: Hence, on this alternative, “whether the group is finite or infinite,” it stands in need of a factor that will continually “provide it with existence.” The factor, Avicenna assumes, must be either (β1) within the group or (β1) outside it. Assuming that the whole group is (β1) ultimately maintained by one of its own members would, however, be tantamount to assuming that the member in question is a cause of itself. For to be a cause of the existence of a group is “primarily” to be the cause of the individual members; and inasmuch as the supposed cause is itself one of the members, it would perforce be a cause of itself. Yet the supposed cause has already been assumed, as one of the members of the group, to be possibly existent; and the possibly existent is precisely what does not exist by reason of itself. Therefore, it could not be the cause of the collection of which it is one member.26

This point is particularly relevant to the discussion on the New Atheism movement and its interrogations of theism. In his three-part specialised podcast on Ibn Sīnā, Peter Adamson mentions the relevance of this argument to modern day discussions between theists and atheists. When discussing the part of Ibn Sīnā’s argument in which he examines the implausibility of the entire set of existence being necessary if its parts are contingent, Adamson refers to Bertrand Russell’s famous fallacy of composition objection. The latter will be revisited in greater detail in the chapter on objections. Adamson provides the example of a big clock made up of small parts and mentions that just because there are causes within the universe, it does not necessarily mean that the universe itself has a cause. Adamson explains that this: …sounds like the sort of thing a modern-day atheist may say. Avicenna is relaxed on this point. He sees that an opponent might raise this objection and, as if shrugging his shoulders, says that in that case, the opponent would just be giving him what he wants…after all, he is out to prove that there is a necessary existence. The opponent has actually admitted that – it’s just that the opponent thinks that the necessary existence is the universe itself – the objection is no objection at all, but a capitulation.27

Having said this, Adamson explains that according to Ibn Sīnā, the universe (or the set of units) cannot be the necessary existence because: If a single necessary existent had parts, then something would need to distinguish those parts from one another. But then, by the same reasoning, we just used, the parts would wind up being different from one another, and then they would not be necessary – but how can a necessary existent have contingent parts?28

When using Ibn Sīnā’s arguments pastorally, it is important to note the threshold at which an interlocutor has shifted away from atheistic explanations and moved towards explanations which are somewhat more commensurate with deism or classical theism. The purpose of these logical arguments is not to convince the interlocutor of all of the attributes of God in accordance with scripture by arguing (as Ibn Sīnā does above) from first principles. Rather, the prime objective is to demonstrate to an atheistic detractor that a worldview which does not acknowledge the status of an originator of existence is deficient and can be easily repudiated. To summarise, Ibn Sīnā’s Burhān argument can be categorised as a cause-based argument from contingency. The thrust of the argument is that existence cannot contain only contingent actualities, as contingent existence cannot bring rise to itself. A collection of contingent existences –whether finite or infinite – is not self-sufficient or necessary, as such a collection is made up of many component parts. If there is anything to differentiate Part A from Part B, such a thing would indicate the non necessity of Part A or Part B, as well as the ‘whole’ which is being described. For this reason, the necessary existent cannot be composed of component parts. In addition, because of this reason, a set of contingent existences – whether finite or infinite – cannot be a necessary existence. Therefore, there is the requirement for a necessary existence to subsist outside of the set of contingent existence, which ultimately causes this set, as well as all of its members. Such a necessary existence which is outside the set of contingent existences must be self-sufficient; it requires nothing to generate or cause it, as the failure to be such would necessarily indicate its contingency.

Richard, who I have described in the introduction, may reject the proposition of a necessary existence in exchange for an infinite multiverse. The question that may be posed to Richard is whether such a multi-verse is self-sufficient or dependent. If Richard replies that it is dependent, then the question is: is it dependent on something which is dependent or something which is independent (self-sufficient). If independent, then there is agreement at least on the point that there exists an ‘independent entity’. If dependent, then is there an infinite regress of dependent things? If so, what does such infinite regress depend on? Richard may say that it depends upon nothing. If this is so, Richard is admitting the infinite multiverse is independent, again a point of agreement as Richard would be agreeing to the existence of something self-sufficient. The major point of disagreement relates to what the independent being is. One can argue that anything composed of detachable/attachable parts is dependent, and since the universe is composed in this way, therefore the universe is dependent. The same thing can be said about an infinite multiverse. This will be fleshed out in the chapters to come.