Jon McGinnis

3 S The Ultimate Why Question

Avicenna on Why God Is Absolutely Necessary

The question Why is there anything at all rather than absolutely nothing? was not a question medieval Arabic-speaking philosophers were
prone to raise, at least not in this exact wording. Instead, they were more
concerned with the related question, Why is there a world rather than
no world at all? or more exactly, Why does the world have the particular features that it has? Certainly in the classical and medieval periods
the standard answer to this latter question was simply, in one form or another, God. Plato invoked the need for a demiurge to explain the orderly
existence of our world; Aristotle argued that there must be an unmoved
mover to explain the manifest motion in the world; and Neoplatonists
later appealed to the One to explain the unified existence of the world.
What is common to all of these thinkers is that they began with what
might be called a physical fact, that is to say, some particular feature
about the way the world actually is, whether it be its order, motion, unification or the like, and then they invoked God as the required cause of
these physical facts. Since all these proofs for the existence of God begin
with what I am calling a physical fact about the world, one might call
them physical arguments for the existence of God.
The medieval Arabic philosopher Ibn Sn (9801037), the Latin Avicenna, found the use of physical arguments to prove the existence of
God wanting and complained that what was needed was a metaphysical proof for the existence of God.1 I understand his complaint to be that
1. See Commentary on Lambda, in Aris.t inda l-Arab, ed. A. Badawi (Cairo: Maktabat
an-nahda
. al-mi.srya, 1947), 2324; Talqt, ed. A. Badawi (Cairo: Maktabatal-al-Arabya,

65

66 Jon McGinnis
physical arguments for the existence of God prove only the conditional
necessity of God: since some physical fact exists, then God exists. If that
physical feature of the world had counterfactually not existed, and some
physically different world existed, then a necessary premise of the proof
would be lacking and so that particular argument would fail to prove
the existence of God. In contrast, a metaphysical argument, or so I contend, would prove the absolute necessity of God regardless of any physical facts or specific features about the way the world actually is, as such
a metaphysical argument would show that if anything exists, no matter
how it might exist, then God necessarily exists. For Avicenna such an
argument must begin from an analysis of existence itself, or being qua
being, and more precisely the irreducible modal structure of existence.2
In this study I want to consider Avicennas metaphysical argument
for the existence of God and the modal metaphysics that underpins it,
but I also want to consider how Avicennas modal metaphysics provided
him with the means to argue for another historically important philosophical thesis, namely, the eternity of the world. Avicennas argument
for the existence of God attempts to show that if anything exists, then
a Necessary Being, namely God, must exist; his proof for the eternity of
the world attempts to show that if it is even possible that the world exists, then the world must be eternal. What is of particular interest about
Avicennas proof for the eternity of the world, I shall argue, is that, when
it is coupled with his proof for the existence of God, the result is an even
stronger proof for the existence of God, namely, one that shows that if
anything whatsoever is simply possible, then necessarily God exists. In
other words, in response to the ultimate why question: Why is there
anything at all rather than absolutely nothing? Avicennas answer
comes down to Because something is possible.

1973), 62; and the distinction was implied though not explicitly made at al-Isharrt wa-ttanbht, ed. J. Forget (Leiden: E. J. Brill, 1892), nama.t 4, fa.sl 29, 14647. For a discussion of the
historical context for this distinction see Dimitri Gutas, Avicenna and the Aristotelian Tradition: Introduction to Reading Avicennas Philosophical Works (Leiden: E. J. Brill, 1988), 26165.
2. For an excellent study of the historical context for Avicennas doctrines of existence as
well as the necessary and possible existents see Robert Wisnovsky, Avicennas Metaphysics in
Context (Ithaca, NY: Cornell University Press, 2003), part II.

Avicenna 67

I
Let us then begin with Avicennas analysis of existence and his modal
metaphysics. Book I, chapter 5 of the Metaphysics of his Shif is dedicated to an indication that something exists (mawjd) and what the divisions of existence (wujd) are. That something exists and that there is existence, he begins, is the first thing impressed upon the soul and simply
cannot be doubted.3 Trying to demonstrate that there is existence or that
something exists, he argues, is a fools errand, since every demonstration
proceeds from things better known than and prior to the conclusion.4
Thus if one assumed that there were anything better known than and
prior to existence itself, one would be committed to the existence of that
thing itself, and so would have to assume its existence, the very thing that
one was attempting to demonstrate. In short, for Avicenna, any proof
that there is existence must be inherently circular.
Such a claim I take to be clear enough and in need of little defense.
Avicenna goes on, however, to add that since the necessary (wjib wujd)
and the possible (mumkin wujd) are the primary divisions of existence,
they too must have the same epistemic status as existence, namely, any
attempts to prove that or what the necessary and possible are will be
inherently circular.5 The claim that necessity and possibility have, as it
were, the same epistemic status as existence should give one reason to
pause; for it is not immediately obvious how necessity and possibility,
even if they are existences primary divisions, are on a par with existence
when it comes to being better known than and prior to anything else
whatsoever; for it certainly seems that while one is immediately aware of
existence, in some way or another, one is not immediately aware of necessity and possibility.
Avicenna freely concedes that certain physical facts about the world
3. Shif, al-Ilhyt (henceforth Metaphysics), I.5, 22.1112; references to the Metaphysics are to the Cairo edition found in The Metaphysics of The Healing, trans. (with Arabic edition) M. E. Marmura, Islamic Translation Series (Provo, UT: Brigham Young University Press,
2005). Also see Najt, Ilhyt, ed. M. T. Danishpazhuh (Tehran: Danishgah-i Tihran, 1985/6),
I.4, 396. All translations are my own.
4. For a discussion of Avicenna on the primacy of existence see Michael E. Marmura, Avicenna on Primary Concepts in the Metaphysics of the Shif, in Probing in Islamic Philosophy
(Binghamton, NY: Global Academic Publishing, 2005), 14969 (henceforth Probing).
5. Metaphysics, I.5, 27.3228.27.

68 Jon McGinnis
draw ones attention to existence in a way that they do not draw ones attention to necessity and possibility.6 Part of the reason for this fact is that,
to use an Aristotelian distinction that Avicenna himself appropriates,
physical facts, namely, things perceived by the senses, might be better
known and prior relative to us, but they are not better known and prior absolutely or by nature to the necessary and the possible, or so Avicenna believed.7 In other words, we often are better acquainted with certain
effects while being wholly oblivious to the causes of those effects, even
though the causes are better known and prior by nature or absolutely.
The reason that Avicenna believes that the necessary and possible are
on a par with existence is that ones conceptualizing (ta.sawwur)and
so ones knowingnecessity and possibility are exactly like conceptualizing and knowing existence; for in all of these cases there is nothing
more basic by which one could explain or define the notion in question.
Thus the very status Avicenna observed about existence, he also would
observe about modal terms:
Virtually everything that has reached you from the ancients concerning how to
explain [the necessary, possible and impossible] requires [you do so] circularly. That is because ... when you want to define the possible, you take either
the necessary or the impossible in its definition, and there is no other way but
that. [Similarly] when you want to define the necessary you take either the possible or impossible in its definition.8

For Avicenna, then, any explanation of what necessity and possibility are
must be inherently circular, and it is just in this respect that necessity and
possibility are like existence. There is something ultimately basic about
them. There simply is nothing better known or prior to them by which one
could define or demonstrate these modalities. Moreover, because any attempt to define or demonstrate these modal concepts is ultimately circular, Avicenna maintains that they have the same status as existence with
respect to being primary, albeit as the primary divisions of existence itself.
Here a somewhat extended digression seems warranted; for Avicen6. Ibid., 28.2832.
7. Avicenna treats at length the distinction between better known and prior by nature
and better known and prior to us at Shif, al-Burhn, ed. A. Badawi (Cairo: Association of
Authorship, Translation and Publication Press, 1966), I.10, and again at Shif, a.t-Taby,
ed.
.
S. Zayed (Cairo: General Egyptian Book Organization, 1983), I.1 (henceforth Physics).
8. Metaphysics, I.5, 27.1928.3.

Avicenna 69
na rests the primacy of necessity and possibility on the basis that they
can be defined only circularly, and yet it is not clear that this is the case.
Certainly if one considers only the definitions of necessity and possibility found in book Delta of Aristotles Metaphysics these concepts
are explained in terms of one another; for there Aristotle does define the
possible (dunaton) as that which is not necessary (anangk),9 while he
defines the necessary as that which cannot be otherwise (m endechomenon),10 where can and possible are logically equivalent.11
Still this is not the only way that Aristotle defined these terms. Indeed, implicitly in his logical works and explicitly in the De caelo Aristotle explains necessity, possibility, and impossibility in terms of a
temporal frequency model of modalities.12 On this model, if something
exists for all time, it is necessary; if it exists at some time, but not at another time, it is possible; and if it exists at no time ever, it is impossible.
Now what is interesting about analyzing modalities in terms of temporal
frequency is that this model apparently reduces modalities to non-modal
elements, namely, temporal elements. Such an account of modalities obviously raises a problem for Avicennas circularity thesis. Moreover, not
only was the temporal frequency model of modalities arguably the means
of defining modal concepts best known by both Greek-speaking and early Arabic-speaking philosophers; it was also the very one that Avicenna
himself used to explain necessity and possibility in his logical works.13
Consequently, since Avicenna was aware of the temporal frequency model and even uses it himself, he owes his reader an account of how temporal
9. Aristotle, Metaphysics V, 12, 1019b2729.
10. Ibid., V 5, 1015a34.
11. Cf. De Interpretatione 13, 22a1516.
12. See De Interpretatione 9, 19a14, 3236; 23a23, and De caelo I 11, 281a16; I 12, 281a28
30. For a discussion of Aristotles modal concepts, see Sarah Waterlow, Passage and Possibility:
A Study of Aristotles Modal Concepts (Oxford: Clarendon Press, 1982).
13. Within the Arabic-speaking logical tradition prior to Avicenna see Yahy
. ibn Ad,
Establishing the Nature of the Possible, edited and translated by Carl Ehrig-Eggert, Yahy
.
ibn Ad: ber den Nachweis der Natur des Mglichen, in Zeitschrift fr Geschichte der arabisch-islamischen Wissenschaften 5 (1989): 28397, [Arabic] 6397, esp. 7880. Interestingly,
during his discussion of De Interpretatione 9, al-Frb suggests that in certain contexts modalities might be taken as self-evident without appeal to temporal frequencies. As for Avicenna, one can see his use of the temporal frequency model of modalities at Shif, al-Qiys, ed. S.
Zayed (Cairo: Wizrat-al marif, 1964), I.4, and Al-Ishrt wa-t-tanbiht, nahj 4; also see Tony
Street, Fakhraddin ar-Razis Critique of Avicennan Logic, in Logik und Theologie: Das Organon im arabischen und im lateinischen Mittelalter, ed. U. Rudolph and D. Perler (Leiden: E. J.
Brill, 2005), 2 for a discussion of temporal frequency in Avicennas logic.

70 Jon McGinnis
frequencies presuppose modal concepts; yet in his Metaphysics there is
no mention of this model at all. Thus not only does it seem that Avicenna
was being disingenuous when he said that necessity and possibility can
be defined only circularly, but also his general argument for the primacy
of these concepts seems jeopardized.
Although one cannot help but feel that Avicenna is being less than
forthright at this point, he does provide sufficient material to piece together a general argument showing that defining modalities in terms of
temporal frequency ultimately does involve a circular definition, albeit
this material comes not from his Metaphysics but from his Physics. In
that work, Avicenna, following Aristotle, defines time in terms of motion, namely, as the number of motion when it is differentiated into prior and posterior.14 Unlike Aristotle, however, before Avicenna defines
what time is, he provides an explicit proof that time is.15 We can set aside
the details of his proof for the reality of time, since what is interesting
about it is that it is argued in terms of a varying possibility (imkn),
which belongs to motion, to traverse greater and lesser distances. Avicenna, then, ultimately identifies this possibility to traverse greater and
lesser distances with time.16 In other words, on Avicennas analysis of
time, time is understood first and primarily in terms of the modal notion of possibility, and then only secondarily in terms of the motion in
which that possibility resides. In short, if Avicennas analysis of time is
correct, one comes to know time only if one already knows what possibility is. Consequently, any attempts to define or explain possibility and
the other modalities in terms of temporal frequency, and so in terms of
time, would in fact, on Avicennas analysis of time, implicitly presuppose
a modal notion, again namely, possibility. Certainly, relative to ourselves,
we may come to know that there is time before we know that there is possibility and what possibility is, but considered absolutely or by nature, if
Avicennas analysis of time is correct, possibility is better known than
and prior to time. Even if this argument is not Avicennas own as it applies to the circularity of defining modal notions in terms of temporal
frequency, it certainly is Avicennan in spirit.
14. Physics, II.11, 157.6.
15. See my Avicenna (Oxford University Press: Oxford & New York, 2010),7175 for a discussion of Avicennas proof.
16. Physics, II.11, 155.4156.17.

Avicenna 71
Thus, to return to the original aim, for Avicenna necessity and possibility stand alongside existence as its primary divisions. They, like existence itself, simply cannot be defined or explained in terms of something
more basic. Still, Avicenna did believe that one could provide descriptions of the necessary and possible, or in other words, criteria for identifying them, even if these descriptions or criteria are in terms of things
less evident in themselves than necessity and possibility.17 Again, as we
have already seen, however, Avicenna could and happily did concede that
sensible things are better known and prior relative to us, even though
necessity and possibility are better known and prior absolutely or by
nature.
Thus Avicenna describes the necessary in itself as that mode of being that is in itself wholly determinate and as such is in no need of a cause
to explain its existing at all. In fact, should the necessary in itself exist,
there would be no conditions whatsoever under which it could fail to exist. Conversely, Avicenna describes the possible in itself as that mode
of being that is indeterminate between existence and non-existence, in
which case either the presence of a cause selectively determines that it
exists or the absence of a cause selectively determines that it does not exist. Again, these claims are not intended to be definitions of the modal
terms they describe, but just descriptions. As such they provide one simply with criteria for identifying and so determining what would count as
falling under one or the other of existences divisions.
For the sake of completeness, I should add that Avicenna made one
further division of existence, which, as it were, is a hybrid of the necessary in itself and the possible in itself. This further division is that which
is possible in itself, but necessary through another, namely, it exists necessarily as a result of a cause. In effect, this last division, namely, the possible in itself, but necessary through another, corresponds with any created things actual or determinate existence.18 With this final division
Avicennas analysis of existence is complete.
17. Avicennas criteria for identifying the necessary and the possible in itself, as well as
the possible in itself, but necessary through another, can be found at Metaphysics I.6 29.32
32.6 and Najt, Ilhyt, II.1. George F. Hourani, has collected together and translated a number of the most important passages in which Avicenna discusses descriptions of the necessary
and possible in Ibn Sn on Necessary and Possible Existence, Philosophical Forum 4 (1972):
7486.
18. Metaphysics I.6, 31.2732.6; Najt, Ilhyt, II.3, 54849.

72 Jon McGinnis

II
Bearing in mind the primacy of the necessary and the possible in Avicennas thought, as well as their descriptions, I want to turn to Avicennas argument for a Necessary Existent or God as he presents it in his
smaller encyclopedic work, the Najt, or the Latin Metaphysices compendium.19 Here, as in the Shif, Avicenna takes it as beyond doubt that
something exists and that the necessary and the possible are the primary divisions of existence. Using this division of existence as his starting
point, Avicenna argues that there must be a Necessary Existent as follows: if, on the one hand, the existence that immediately presents itself
to us is necessary, there is a Necessary Existent. On the other hand, if the
existence of which we are immediately aware is only possible existence,
then, given the description of the possible, it must have a cause. At this
point, rather than denying the impossibility of an infinite causal chain,
as virtually all earlier physical arguments for the existence of God had
done, Avicenna asks his reader to consider the totality (jumla) of all the
possible things that exist at some given moment, regardless of whether
that totality is finite or infinite.20 He next observes that the existence of
this totality itself must be either necessary in itself or possible in itself,
given again that these are the two primary divisions of existence and that
the totality itself exists.
On the one hand, the totality cannot be something necessary in itself,
19. Najt, Ilhyt, II.12, 56668; for a translation of the text see Jon McGinnis and David C. Reisman, Classical Arabic Philosophy: An Anthology of Sources (Indianapolis: Hackett
Publishing Co., 2007), 21415. As for the occurrence of this proof in the Shif, see Michael E.
Marmura, Avicennas Proof from Contingency for Gods Existence in the Metaphysics of the
Shif, in Probing, 13148.
20. It has been noted that one thing that sets apart Avicennas proof for the existence of a
Necessary Existent from other proofs for the existence of God is that it does not require the
premise that an actual infinite is impossible: see Herbert A. Davidson, Proofs for Eternity, Creation and the Existence of God in Medieval Islamic and Jewish Philosophy (New York: Oxford
University Press, 1987), 300302 (henceforth Proofs); and Toby Mayer, Ibn Sns Burhn alSiddqn, Journal of Islamic Studies 12 (2001): 1839, esp. 2535. Almost certainly one of the
primary motivations for this new type of argument is the fact that earlier in the Physics of the
Najt (a.t-Tabyt,
IV.7, 24452, esp. 246), Avicenna permits the possibility of non-positional
.
actual infinities, as for instance an actual infinite number of immortal human souls; see Michael E. Marmura, Avicenna and the Problem of the Infinite Number of Souls, in Probing,
17179. Consequently, if the present argument simply assumed that there could not be an actually infinite number of (non-linearly) essentially ordered causes, Avicenna would have been
guilty of special pleading.

Avicenna 73
argues Avicenna; for the existence of a totality subsists only through the
existence of its members, in which case that which is purportedly necessary in itself would be necessary through another, namely, its members,
but this is a contradiction.21 Moreover, it is impossible that there should be
any conditions under which the necessary in itself should fail to exist; for
this is one of the very criteria by which one identifies necessary existence.
If, however, all the members of the totality exist only possibly in themselves, and things that exist possibly in themselves can fail to exist, then, if
each of the members of the totality failed to exist, the totality itself would
fail to exist. In other words, there is at least one condition under which the
totality of all possible things could fail to exist, namely, if all of its members failed to exist. Again, however, there is no condition under which the
necessary in itself could fail to exist, should it exist. Consequently, the totality of things possible in themselves cannot exist necessarily in itself.
On the other hand, Avicenna continues, if the totality is something existing possibly in itself, then in order for the totality to exist it
needs something that provides existence [that is, causes it to exist], which
will be either external or internal to the totality.22 If it is something existing internal to the totality, it itself would be one of the members of the
totality and as such that member cannot exist necessarily, since the assumption is that the totality includes only things possible in themselves.
Thus if the cause of the totalitys existence were internal and necessary in
itself, there would be a contradiction.
Furthermore, the cause of the totalitys existence cannot be something internal and possible in itself; for since anything internal is part of
the totality itself, it would be the cause of its own existence. In that case,
one and the same thing would be both a cause and an effect in the very
same respect, which Avicenna himself takes to be absurd. Still, it would
seem that Avicenna was aware that the denial of self-causation with respect to existence is too physically robust a claim for his metaphysical
argument. Thus, per impossibile, he granted for the sake of argument
that something could cause its own existence; however, in that case one
must still concede that this member is sufficient to necessitate its own ex21. At Najt, Ilhyt, II.45, 54952 Avicenna argues extensively that the necessary in itself cannot have parts, since this would in fact make the necessary in itself necessary through
another.
22. Najt, Ilhyt, II.12, 56768.

74 Jon McGinnis
istence, in which case it exists necessarily through itself. Again, however,
the assumption is that the totality includes only things possible in themselves, and so again there would be a contradiction, namely, something is
both necessary in itself and possible in itself.
Thus one is left only with the option that the totality is possible in
itself and the cause of its existence is something external to the totality.
Again, this external thing cannot be something possible in itself; for everything existing possibly in itself is included within the totality. Consequently, this external cause of the existence of all things possible in
themselves must fall within existences other division, namely, the necessary in itself. Therefore, concludes Avicenna, this external cause is something that exists necessarily in itself, and so there must be a Necessary
Existent. Q.E.D.
The first thing to observe about this argument is that if one sets aside
the modal metaphysics underlying it, then the argument is extremely
modest in the premises it requires. Avicenna assumes something about
sets or totalities, namely that they subsist through their members, but
such a claim seems to be almost true by definition. At one point he takes
self-causation to be absurd, but, as he quickly notes, the argument does
not absolutely require this premise. Thus, as far as I can see, the most ontologically or physically robust claim assumed by the argument is simply
that something exists.23 It is irrelevant to Avicennas argument whether
23. This fact, in addition to the fact that the argument is driven by Avicennas modal metaphysics, has led to a debate in the secondary literature as to whether the argument should be
classified as an ontological argument, a cosmological argument, or some hybrid of the two.
Perhaps the strongest supporter of the ontological reading is Parvis Morewedge, A Third Version of the Ontological Argument in the Ibn Sinian Metaphysics, in Islamic Philosophical Theology, ed. P. Morewedge (Albany: State University of New York Press, 1979), 188222; on the
other side is Herbert A. Davidson, Proofs, 298308. Others who see both an ontological and a
cosmological element are S. A. Johnson, Ibn Sns Fourth Ontological Argument for Gods
Existence, Muslim World 74 (1984): 16171, and Toby Mayer, Ibn Sns Burhan al-Siddqn.
My own view is that, strictly speaking, Avicennas argument is not an ontological one. The hallmark of ontological-style arguments is that they assume only a priori premises, and so have no
recourse to empirical experience. Now although Avicenna thinks that existence is one of the
primary intelligibles, and so one would never remember a time when one did not know existence, it does not follow that such a conception is innate or a priori; rather, for Avicenna, existence is the very first thing we experience and so the very first thing impressed on the intellect. Thus it seems that the argument is not strictly speaking an ontological-type argument at
all, since nothing is taken as a priori. Of course, if one means by ontological merely that Avicennas argument refers solely to existence or being as such without making reference to any
physical facts about existence, then in this qualified sense his argument is an ontological one.

Avicenna 75
the existence is chaotic or ordered, as is assumed by Platos argument
for a demiurge; whether the existence is undergoing motion or not, as in
Aristotles argument for an unmoved mover; or whether the existence is
unified or wholly multiple, as certain Neoplatonists assumed when arguing for the One. In short, based on the existence of anything however it
might exist and his analysis of existence, Avicenna can show that there
must be a Necessary Existent. In other words, the world could have been
wholly different than it isindeed a world that we might not even be able
to conceivebut provided that something exists, there will necessarily
be God.

III
We have now considered Avicennas analysis of the modal structure of
existence and his metaphysical as opposed to physical proof for the existence of God. In the remainder of this study, I want to consider how
Avicenna uses his modal metaphysics to argue for the eternity of the
world.24 In its simplest form, the structure of Avicennas argument is that
if it is ever possible that the world exists, that is to say, if it is possible
that something other than God exists, then the world must have existed
eternally. Since the world obviously exists now, its existence was possible.
Thus the world must have existed eternally.
That the world exists now cannot be gainsaid. The inference from the
worlds existence to its prior possible existence requires only the weak
and intuitively plausible modal commitment that if x is now, then x was
possible. Avicenna further justifies this modal commitment by arguing
that if there had been no prior possibility of xs existence, then the present existence of x would have been initially impossible; but if x had been
impossible, then it would not exist right now.25
24. Herbert A. Davidson provides an exhaustive list of arguments for the eternity of the
world found among medieval Islamic and Jewish philosophers in Proofs, 948; proof (b), 1617,
is a condensed version of the argument that I shall present here. Also. that this argument had
wide currency during the Middle Ages is witnessed by the fact that both Moses Maimonides
(Guide of the Perplexed, book II, ch. 14) and Thomas Aquinas (Summa theologiae I, q. 46, a. 1,
obj. 1) considered it one of the strongest arguments for the worlds eternity. For a general discussion of Avicennas arguments for the eternity of the world (although interestingly not the
argument presented here), see Rahim Acar, Talking about God and Talking about Creation: Avicennas and Thomas Aquinas Positions (Leiden: E. J. Brill, 2005), pt. 2, esp. ch. 4.
25. Metaphysics, IV.2, 140.711; Najt, Ilhyt, I.17, 534; and Physics III.11, 232.15233.3.

76 Jon McGinnis
The premise that clearly is in need of justification is that if the worlds
existence is possible, then the world must have existed eternally. Avicennas defense of this premise is complex, but considering it in some detail, I believe, will be fruitful. The general strategy of his argument is to
identify, as it were, the ontological basis for the possibility of anything
that comes to be, namely, the possibility associated with those things
that at one time did not existsuch as myself prior to 1965and then
later do exist.26 Thus he begins: if possibility exists, and again this is beyond doubt for him, then it is either (1) a substance in its own right, that
is to say, it is some self-subsisting thing, or (2) it inheres in a substrate
and subsists through that substrate. Option (2), namely that possibility
inheres in a substrate, can be divided into two further options, namely
that the possible in itself might exist either (2a) in a material substrate or
(2b) in an immaterial substrate, such as either a created intellect or God
Himself.
Avicenna denies that possibility is a substance in its own right (option 1), for in that case, possibilities would not be able to stand in the
proper relation with other substances. Thus he writes:
If [possibility] (jawz) were subsisting in itself, neither in a substrate (mahall)
.
nor in a subject, then as such it would not be something related; however, as
possible existence it is related to something and intellectually understood by the
relation. So it is not a substance subsisting essentially. Perhaps instead it is certain relation or accident belonging to a substance. Now it cannot be a substance
that has a relation, because that relation is associated with something assumed
to be non-existent. Also it is impossible that that relation is simply a certain
association, however it might chance to be; rather, it is a determinate association, wherein the association is determinate only in that it is possible. Therefore,
possibility is itself the relation, not some substance to which the relation is attached, in which case the sum of the two would be possibility.27

26. Here it is worth noting that while I shall speak simply of the possible and the possible
in itself, I mean specifically the possibility of those things that come to exist after having not
existed. In this respect, the present account is limited in that it does not address the possibility of sempiternal being such as angels, for while according to Avicenna, there is no time that
angels have not existed, their existence is nonetheless wholly dependent upon God. Thus for
Avicenna angels are beings that are merely possible in themselves even thought they have always existed.
27. Physics III.11, 233.311.

Avicenna 77
Avicennas general point, as I understand it, is that if possibility were a
substance in its own right, then it could never properly be related to other substances in the requisite way that possibilities are so associated with
substances. Neither here nor in other parallel passages does he explain
exactly what possibilitys proper relation to substances is;28 however, my
suspicion is that, for Avicenna, what it is to be a certain substance is in a
very real sense determined precisely by the range of possibilities that belong to that substance. In this respect, then, the range of possibilities that
belong to a substance must be related to it essentially, whereas any two
distinct substances can only ever be related to each other accidentally. So
if possibility were a substance, the range of possibilities that belong to a
substance would be accidental to it, just as the clothes I am wearing (one
substance) are accidental to me (another substance). It certainly seems
false that the range of things I can do, for instance, are accidental to me;
for in a real sense what defines me as the sort of substance I am, namely,
human, is determined by what I as a human can do.29
Thus it would seem that possibility must exist in a substrate. Now if
that substrate is immaterial, it might be either a created intellect, such
as the human soul or an angel, or a wholly uncreated immaterial being,
namely, God. The substrate for the possibility of what comes to be cannot
be a created, immaterial substrate; for insofar as such a substrate is created, the possibility of its existence is ontologically prior to (even though
not necessarily temporally prior to) the actually existing created thing;
however, if this created immaterial substrate is to provide the ontological
basis for possibility, its actually existing would be ontologically prior to
possibility itself, in which case, points out Avicenna, there is a contradictionthe created immaterial substrate is both ontologically prior to and
ontologically posterior to the possible.30
As for the suggestion that a wholly uncreated, immaterial being,
namely, God, is the ultimate ontological basis for the possibility of what
comes to be, there is for Avicenna a sense in which this is correct and a
sense in which it is incorrect. At the end of this paper I shall return to
the sense in which this claim is correct, but for now it is enough to recall
28. Metaphysics, IV.2, 140.1114 and Najt, Ilhyt, I.17, 53536.
29. This point is somewhat suggested when Avicenna writes: those [created] essences are
that which in themselves are what is possible of existence (Metaphysics, VIII.4, 276.2630).
30. Cf. Physics, II.12, 160.1316.

78 Jon McGinnis
that for Avicenna the Necessary Existent is the ultimate explanation and
cause for the actual and determinate existence of anything that is merely
possible in itself.
As for the present context, namely, whether God can be the substrate
in which the possibility inheres and has its subsistence, Avicenna finds
the claim wholly untenable.31 He offers a number of arguments against
this thesis, each depending upon one of the various ways one might consider that God is the ultimate basis or explanation of the possible in itself. Two such arguments that I shall consider correspond with two common accounts of how possibility might be referred back to God. One is
that possibilia refer to ideas in the mind of God, in which case these ideas
represent the exemplars of all that can possibly be; the other is that the
possibility of somethings existing resides in Gods power to create that
thing, and so all things falling within the range of Gods omnipotent
power are possible and otherwise they are impossible. In general terms,
Avicenna would argue that populating the divine mind with possibilia or
multiple ideas jeopardizes the divine simplicity, whereas referring possibility back to Gods power without some independent account of possibility strips the notion of omnipotence of any significance.
Against the first suggestion, namely, that the possible in itself is referred back to the divine understanding, Avicennas critique begins by
observing that Gods understanding is simple and not made up of multiple ideas.
[W]hen it is said that understanding (aql) belongs to the First [Cause, that is,
God], it is said according to simple intention (which I explained in the Psychology) and that there is not in it a variety of various ordered forms ... ; for it belongs to that [Cause, that is, God] to understand things all at once as one, without either being multiplied by them in His substance or conceptualizing them
by their forms being in His essence.32

In other words, for Avicenna there simply are not multiple ideas in the
mind of God by which he knows things; rather, God knows directly only
31. Although Avicenna addressed, on an ad hoc basis, the suggestion that God might be a
substrate of possibility Moses Maimonides, clearly working within the falsafa tradition itself,
took any position that makes God a substratum to be manifestly absurd, since it jeopardizes divine simplicity, for God would then be a composite of the substratum and that which inheres
in the substratum; see Guide for the Perplexed, book I, ch. 52.
32. Metaphysics, VIII.7, 291.69; also see Najt, Ilhyt, II.1819, 59399.

Avicenna 79
his simple essence and everything else only inasmuch as he knows that
essence.33 Avicennas reason is that if the Necessary in Itself, that is God,
were to understand multiple things, and not merely Itself alone, then
the divine understanding, which in a very real sense is identical with
the Necessary in Itself, would be constituted by or subsist through these
multiple ideas. Now inasmuch as a unity made up of multiple elements
subsists only through those elements as its cause, so likewise the Necessary in Itself would be caused, which is absurd. The same point, stated
slightly differently, is that for Avicenna whatever has multiple parts requires a cause in order to make those parts a unified existent. Thus, since
there can be no cause for the existence of the divine mind, it cannot have
multiple parts, in this case, multiple ideas.34
Therefore, on the one hand, the plurality of possibilities that exist in
the world cannot refer back to a plurality of ideas or exemplars in the
mind of the Necessary in Itself; for there absolutely is no plurality in the
Necessary in Itself. On the other hand, if the possible in itself were referred back to the absolutely simple understanding of the Necessary in
Itself, and that understanding is to remain simple, then the possible in
itself would have to be identical with that understanding; but that would
be to make the Necessary in Itself possible in itself, which is a contradiction. In short, for Avicenna, since the Necessary in Itself is absolutely
simple, the possible in itself cannot be referred back to the divine understanding.
Avicenna likewise argues against locating the foundation for possibility solely in divine causal power. The suggestion would be that something is possible only because it is within Gods power to cause it to be.35
33. For differing approaches to Avicennas conception of divine knowledge, and particularly how the Necessary Existent knows things other than himself, see the following: Michael E.
Marmura, Some Aspects of Avicennas Theory of Gods Knowledge of Particulars, in Probing,
7195; Rahim Acar, Reconsidering Avicennas Position on Gods Knowledge of Particulars, in
Interpreting Avicenna: Science and Philosophy in Medieval Islam, ed. J. McGinnis (Leiden: E. J.
Brill, 2004), 14256; and Peter Adamson, On Knowledge of Particulars, Proceedings of the Aristotelian Society 105 (2005): 27394.
34. Metaphysics, VIII.4, 27378, provides an extended discussion of why the Necessary Existent cannot be a composite, whether of quantitative parts or merely conceptual parts, such
as ideas constituting the divine mind, which in fact is essentially the Necessary Existent; for
these parts would be causes of the Necessary Existent, but the Necessary Existent is in no way
caused.
35. The following authors all attribute this conception of possibility to Aquinas: Gerard
Smith, Avicenna and the Possibles, New Scholasticism 17 (1943): 34057; Beatrice Zedler,

80 Jon McGinnis
Avicenna rejects this suggestion since, if possibility is understood to be
through an agents power rather than through some independent account of possibility, then both the terms power (qudra) and possibility
(imkn) become vacuous.
[Possibility] is not the power of the one who has power over it, otherwise when
by replacement it is said that it is not an object of power because it is not possible in itself, it would have been said that it is not an object of power because it
is not an object of power, or it is not possible in itself because it is not possible
in itself. Clearly, this possibility is different than the one who has power over
it ....36

In other words, if possibility referred to nothing more than the power of

the agent to do or make something, then to say, Something is possible
if it falls within the agents power simply means, Something falls within the agents power, if it falls within the agents power. Consequently,
should one say that God is omnipotent and so can do whatever is possible, one actually would be saying that God can do whatever God can
do. Clearly, however, everything can do what it can do, and so everything would be omnipotent on this account of possibility. For Avicenna,
then, what is needed is an independent understanding and grounding of
the possible in itself, if Gods power over all possible things is not to be
trivial.
Before completing Avicennas proof for the eternity of the world, let
us quickly assess the argument to this point. Again, Avicennas primary
aim is to justify the claim that if the existence of the world was ever possible, then the world is eternal. In order to prove this claim, he has one
consider the ways in which possibility might be said to exist. Avicenna
notes that possibility might be understood to exist either as a substance
in its own right or in a substrate. He quickly denies that it exists as a
substance, and so concludes that possibility must exist in a substrate. He
goes on to catalogue the various types of substrates in which it might
Another Look at Avicenna, New Scholasticism 50 (1976): 50421 and Why Are the Possibles
Possible, New Scholasticism 55 (1981): 11330; and James Ross, Aquinass Exemplarism; Aquinass Voluntarism, American Catholic Philosophical Quarterly 64 (1990): 17198. For reasons
quite similar to Avicennas, Lawrence Dewan (St. Thomas and the Possibles, New Scholasticism 53 [1979]: 7685) rejects this conception of possibility in Aquinas in favor of concept compatibility.
36. Al-Ishrt wa-t-tanbht, nama.t 5, fa.sl 6, 152; also see Metaphysics, IV.2, 139.11140.12,
and Najt, Ilhyt, I.17, 53435.

Avicenna 81
inhere. These could be either material or immaterial; if possibility were
to inhere in an immaterial substrate, that substrate might either be created, such as the human intellect or one of the separate intellects, or not
be created, in which case it would be God. We have just seen a series of
arguments that Avicenna provided against the inhering of the possible in
itself in an immaterial substrate.
Avicenna concludes this extended argument by asserting that the ultimate substrate for the possibility of whatever comes to exist after having not existed must be matter.
We ourselves call the possibility of existence the potentiality of existence, and
we call what bears (hmil)
the potentiality of existence in which there is the po.
tentiality of the existence of the thing subject, prime matter, matter and the
like, on account of many different considerations.37 Thus, whatever comes into
existence is preceded by matter.38

Of course the reason why, for Avicenna, whatever comes to exist is preceded by matter is precisely because, for him, the existence of anything
that comes to exist is preceded by the possibility of its existence, and, as
has just been argued, there would be no possibility if there were no matter qua subject in which that possibility inheres.39
37. At Physics I.2, 14.1415.5 Avicenna lists the various considerations: This matter, inasmuch as it potentially receives a form or forms, is called its prime matter; and, inasmuch as it
actually bears a form, it is called in this [book] its subject. (The sense of subject here is not the
sense of subject we used in Logic, namely, as part of the description of substance; for prime
matter [15] is not a subject in that sense at all.) Next, in as much as it is common to all forms,
it is called matter and stuff [lit. clay]. It is also called an element because it is resolved into
[elements] through a process of analysis, and so it is the simple part receptive of the form as
part of the whole composite, and likewise for whatever is analogous. It is also called a constituent because the composition begins from it in this very sense, and likewise for whatever is
analogous. It is as though when one begins from it, it is called a constituent, whereas when one
begins from the composite and ends at it, then it is called an element, since the element is the
simplest part of the composite.
38. Metaphysics, IV.2, 140.1517; also see Najt, Ilhyt, I.17, 536.
39. For treatments of Avicennas theory of matter see Arthur Hyman, Aristotles First
Matter and Avicennas and Averroes Corporeal Form, in Essays in Medieval Jewish and Islamic Philosophy, ed. A. Hyman (New York: KTAV Publishing House, 1977), 335406; Elisabeth
Buschmann, Untersuchungen zum Problem der Materie bei Avicenna (Frankfurt am Main: Peter Lang, 1979); Abraham D. Stone, Simplicius and Avicenna on the Essential Corporeity of
Material Substance, in Aspects of Avicenna, ed. R. Wisnovsky (Princeton, NJ: Markus Wiener
Publishers, 2001), 73130; and Jon McGinnis, The Avicennan Sources for Aquinas on Being:
Supplemental Remarks to Brian Davies Kenny on Aquinas on Being, Modern Schoolman 82
(2005): 13142, esp. 13537.

82 Jon McGinnis
Consequently, for Avicenna it is simply absurd to claim that God created matter, and so the world, at some first moment in time before which
it did not exist. This is because if only what is possible can be created,
and matter is supposed to be created after not existing, then the possibility of creating the matter must have existed prior to the matter; but
again for Avicenna, this possibility requires the existence of matter as
the substrate in which it inheres. In short, to claim that the possibility of
creating matter existed prior to the matter is tantamount to saying that
the possibility of creating possibility existed prior to possibility, which
indeed is absurd. In summary, given that the existence of the world has
always been possible, matter must have existed eternally, and thus something other than God exists eternally.
We have just seen that God cannot be the ultimate ontological basis
of the possible in itself, inasmuch as one might consider God to be a substrate for the possible in itself or one might think that in some way God
creates possibilities ex nihilo; for in a very real sense, for Avicenna possibilities are simply given. Still, there is a sense in which, for Avicenna,
God is the ultimate cause for the actual existence of the possible in itself,
namely, inasmuch as God ultimately actualizes and makes determinate
the existence of anything possible in itself.
One of the more interesting features of Avicennas claim here is that
if matter were not eternally being actualized and so made determinate,
matter, indeed the possibility for whatever comes to be, would have not
existed, and as such the very existence of the world would have been impossible. The reason is that, for Avicenna, matter is wholly indeterminate
with respect to existence, where again the indeterminacy of existence
is one of Avicennas preferred descriptions of possibility. Now Avicenna did not believe that anything ever actually exists as wholly indeterminate, as for example existing all at once indeterminately as a possible
platypus, plankton, platinum and everything else.40 Matter actually exists only insofar as it is specified to a particular kind by a cause, and we
have already seen that for Avicenna the Necessary in Itself is the ultimate
cause for any determinate or actual existence. Consequently, if the Necessary in Itself did not ultimately cause the determinate actual existence
40. See Jon McGinnis, The Avicennan Sources for Aquinas on Being and A Penetrating
Question in the History of Ideas: Space, Dimensionality and Interpenetration in the Thought
of Avicenna, Arabic Sciences and Philosophy 16 (2006): 4769, esp. 5962.

Avicenna 83
of the possible in itself, then, since nothing exists indeterminately, nothing would have been possible in itself and everything, other than God,
would have been impossible. Simply stated, for Avicenna, either the existence of the world is eternal or its existence would have been impossible;
nevertheless, the actual existence of the world wholly is dependent upon
God as its cause.
I conclude, now, with one brief final observation about the relation
between Avicennas arguments for the Necessary in Itself and the eternity of the world. I noted about Avicennas argument for the Necessary in
Itself that its only existential commitment was that something actually
exists. What we have seen in the course of Avicennas argument for the
eternity of the world is that if anything that comes to be is even possible,
then matter must exist as the substrate in which that possibility inheres.
If we now couple these two conclusions, we see that Avicennas two arguments jointly show that the mere possible existence of the world entails
that God necessarily exist; for if there is that possibility, then the matter
in which the possibility inheres must exist; but according to Avicennas
argument for the Necessary in Itself, if anything exists, then God necessarily exists. In short, if anything, whatever it might be, is possible, God
must exist.
Returning to the central question of this volume, Why is there anything at all rather than absolutely nothing? we can state that for Avicenna the answer is simply: it simply is, because something is possible.