Arabic philosophy

whatsoever. The result is that interpretation (3) will fail to account

for all eternal but caused things.

Perhaps because of his frustration at not coming up with a satisfactory

way to justify his assertion that the necessary of existence

through another will be divisible into two modes or states

and therefore be caused, Avicenna changed tack entirely in his next

work, the Origin and Destination (al-Mabda’ wa-al-ma‘a¯d), written

Cambridge Companions Online © Cambridge University Press, 2006

Cambridge Companions Online © Cambridge University Press, 2006

Avicenna 125

around 1013, when he was 33 or so. In contrast to the Philosophy for

‘Aru¯ d. ı¯, where Avicenna begins with a distinction between the necessary

of existence in itself and the necessary of existence through

another, and then moves to identify the necessary of existence

through another with the possible of existence in itself, Avicenna

starts his discussion in the Origin and Destination by distinguishing

the necessary of existence and the possible of existence, and only

after doing that moves on to identify the possible of existence with

the necessary of existence through another. What advantage did this

new approach give to Avicenna? How did it provide a way for him

to unblock the logjam that had frustrated him in the Philosophy for

‘Aru¯ d. ı¯?

By 1013, when he wrote the Origin and Destination, Avicenna

apparently felt that the distinction between the necessary of existence

and the possible of existence was sufficiently intuitive that he

could simply assert that the necessary of existence is by definition

that whose nonexistence is inconceivable, whereas the possible of

existence is by definition that whose nonexistence is conceivable.

(The impossible of existence, which we need not worry about too

much, is by definition that whose existence is inconceivable.) Once

he had laid out this assumed basis for the distinction, Avicenna then

claimed that we can conceive of the existence of something which is

possible of existence only in the context of a relation which that possibly

existent thing has, namely, the relation it has with its cause. It

is in virtue of this relation with its cause, Avicenna asserts, that the

possible of existence is also necessary of existence through another;

in itself – that is, conceived of in isolation from its relation to its

cause – it remains only possible of existence.

The advantage of the Origin and Destination’s approach to the

distinction is that in it Avicenna felt under no obligation to argue

that the causedness of the necessary of existence through another

results from its being composed of two modes or states, and this

in turn freed him from having to explain precisely what those two

modes or statesmight be. In the Origin and Destination, the causedness

of the possible of existence stems from its relation to its cause,

a relation whose nonexistence is inconceivable once we posit the

actual existence of something which is possible of existence. On the

other hand, the Origin and Destination’s approach could be faulted

for assuming too much – for assuming, that is, that everyone would

Cambridge Companions Online © Cambridge University Press, 2006

Cambridge Companions Online © Cambridge University Press, 2006

126 robert wisnovsky

find its ontological or definitional approach to the distinction to be

so intuitive that the distinction could now bear all the cosmological

weight that would be piled on top of it. After all, the distinction did

serve as the axis of Avicenna’s analysis of all reality in the universe.

In his middle period, the period when he wrote his greatest philosophical

work, The Healing, Avicenna continued to appeal to the

Origin and Destination’s definitional, assumed basis for his distinction.

But in the Metaphysics of The Healing, a crucial thought seems

to have occurred to him. As discussed above, Avicenna argues in

Metaphysics 1.5 that thing and existent – and by implication, essence

and existence – are extensionally identical but intensionally distinct.

Every thing will also be an existent, and every existent will also be a

thing, but to be a thing and to be an existent have different meanings.

In the very next chapter, Metaphysics 1.6, Avicenna distinguishes

between the necessary of existence, the possible of existence, and

the impossible of existence, following the definitional approach he

took in the Origin and Destination, although the phrases he uses are

slightly different. Slightly later on in this chapter, Avicenna introduces

the term ma¯hiyya, essence, into the discussion: that whose

essence is insufficient for it to exist will be caused, whereas that

whose essence is sufficient for it to exist will be uncaused. The new

distinction Avicenna draws here is less important for what it lays

out than for what it represents, because it is here, I believe, that

it occurs to Avicenna to use his essence vs. existence distinction

to provide the conceptual divisibility – and hence compositeness,

and hence causedness – of any being that is necessary of existence

through another and possible of existence in itself, a conceptual divisibility

that he had groped for unsuccessfully twenty-five years earlier

in the Philosophy for ‘Aru¯ d. ı¯.

In fact it is only toward the end of the Metaphysics of The Healing

that Avicenna actually appeals to the essence vs. existence distinction

in this way. In Metaphysics 8.4, a chapter devoted to proving

God’s existence as well as his uniqueness, Avicenna argues that God

is the only being that is necessary of existence in itself because only

God is not divisible into essence and existence, unlike all other

beings, which are composites of essence and existence, and hence

caused, and hence merely necessary of existence through another

and possible of existence in themselves. In God, Avicenna asserts,

essence and existence are intensionally as well as extensionally

identical: God’s essence refers to nothing other than his existence.

Cambridge Companions Online © Cambridge University Press, 2006

Cambridge Companions Online © Cambridge University Press, 2006

Avicenna 127

Although this appears on the surface to be a crucial move, and despite

the fact that later Muslim thinkers were intensely preoccupied with

determining the extent to which essence and existence really were

identical in God, Avicenna’s new argument, at least in the context

of the Metaphysics of The Healing, is not an indispensable component

of his general discussion there of the distinction between

the necessary of existence in itself and the possible of existence in

itself. This is because, as I already mentioned, that general discussion

is still framed in the terms Avicenna first set out in the Origin

and Destination, namely, that the basis for the distinction was

definitional.

In his last major work, the Pointers and Reminders, the two trends

inAvicenna’s articulation, justification, and use of the distinction all

come together in a terse synthesis. As he had in the Metaphysics of

The Healing, Avicenna discusses the distinction between necessary

and possible existence very soon after he discusses the distinction

between essence and existence. But in the Metaphysics of The Healing,

Avicenna’s general discussions of essence and existence (1.5)

and of necessary, possible, and impossible existence (1.6) were separated

by seven long sections from his proofs of God’s existence and

of God’s uniqueness as the only being that is necessary of existence

in itself (8.4). In the Pointers and Reminders, by contrast, Avicenna

moves straight from his discussions of essence and existence and

of necessary, possible, and impossible existence (interspersed with

some comments on the different categories of causes) to his proofs of

God’s existence and of God’s uniqueness. As a result the distinction

between essence and existence is pressed into immediate service,

now playing a fundamental role in the proofs of God’s existence and

of his uniqueness as the only being that is necessary of existence in

itself.

Legacy

As I mentioned above, Avicenna’s distinction between essence and

existence is more underdetermined, more variously interpretable,

than his distinction between the necessary of existence in itself and

the necessary of existence through another – possible of existence

in itself. This is not to say that the latter distinction was passed

on to subsequent thinkers in a fully crystallized form. On the contrary,

it created a number of philosophical problems, opportunities,

Cambridge Companions Online © Cambridge University Press, 2006

Cambridge Companions Online © Cambridge University Press, 2006

128 robert wisnovsky

and challenges to subsequent mutakallimu¯ n. One problem was that

God’s being necessary of existencemight imply a conceptual duality

in him, since there is a sense in which “necessary of existence” can

be understood as a species (albeit a species with only one individual

member) whose genus is “existent” (or “substance”) and whose

specific difference is “necessary.” Just as the species “human” is

notionally divisible into – and hence a composite of – the genus

“animal” and the specific difference “rational,” so too the species

“necessary of existence” will be notionally divisible – and hence a

composite of – the genus “existent” (or “substance”) and the specific

difference “necessary.”27 But once he was understood as a composite,

God no longer had any basis – any non-definitional basis, that is –

to claim sole occupancy of the category “uncaused being.” A second

problem is, should God’s existence be understood as an attribute

(s.

ifa) which is somehow additional to his self or essence, in the way

that Sunnı¯ mutakallimu¯ n understood God’s attributes of knowledge

(‘ilm), power (qudra) and life (h.

aya¯ t)? Should God’s existence be seen

instead as an accident of – and thus caused by – his essence? Or should

God’s essence and existence be taken to be wholly identical?28

Athird loose end is thatAvicenna seems to take for granted that (B)

“necessary of existence through another” and (C) “possible of existence

in itself” are convertible: that every existent that is (B) will

also be (C), and vice versa. The convertibility of (B) and (C) is certainly

plausible, even intuitive, in the case of concrete, extramental

existents (mawju¯ da¯ t fı¯ al-a‘ya¯n). But what about mental existents

(mawju¯ da¯ t fı¯ al-dhihn)? The answer I give will depend on other

interpretive commitments I have made. If, for example, my overriding

concern is to uphold Avicenna’s position (I) on essence and

existence – that essence and existence are intensionally distinct but

extensionally identical – then I shall tend to maintain that the existence

of an essence in the mind deserves to be called “existence”

just as much as the existence of an essence in the concrete, extramental

world does. (This is because I want to avoid holding the

Mu‘tazilite position that essences in the mind can be construed as

non-existent). And this tendency will further lead me in the direction

of assuming that Avicenna’s distinction between (A) “necessary

of existence in itself” and (B–C) “necessary of existence through

another” – “possible of existence in itself” will be just as applicable

to mental beings as it is to concrete existence. An important question

arises, however: what is the cause that makes a mental being

Cambridge Companions Online © Cambridge University Press, 2006

Cambridge Companions Online © Cambridge University Press, 2006

Avicenna 129

“necessary of existence through another”? If the mental being is a

universal such as “Horse” or “Cat,” then the active intellect seems

the likeliest candidate to serve in the role of cause, since it is the

source of all forms in the sublunary world: it is the source of substantial

forms in the case of concrete beings such as the individuals

this horse and that cat, and it is the source of intelligible forms in the

case of mental beings such as the universals “Horse” and “Cat.” But

what about “Unicorn” and “Phoenix,” which are universals that it

is possible to conceive of, and which thus possess mental existence,

yet which are unrealizable as concrete, extramental beings – as the

individuals this unicorn or that phoenix? A further step leads us to

the problematic status of “Pentagonal House” (to use an example

Avicenna cites in a different context), which is a universal that it is

possible to conceive of, and which thus possesses mental existence,

but which (unlike “Unicorn” or “Phoenix”) might well exist someday

as a concrete, extramental being – as the individual this pentagonal

house – even though up to now (at least as far as Avicenna was

concerned) such a house has never existed concretely.

In the case of mental beings that are either unrealizable or as yet

unrealized in the concrete world – mental beings such as “Unicorn,”

“Phoenix,” or “Pentagonal House” – it is harder to determine which

cause makes them “necessary of existence through another.” Perhaps

the cause is still the active intellect; if that is so, the individuals

this unicorn and that phoenix will be unrealizable as concrete

beings (and this pentagonal house is as yet unrealized as a

concrete being) simply because the sublunary world happens to contain

no matter that is suitably disposed to receive those forms from

the active intellect. However, in the case of the intelligible forms

“Unicorn,” “Phoenix,” or “Pentagonal House,” as opposed to the

substantial forms that would inhere in this unicorn or that phoenix

or this pentagonal house, there are suitably disposed receptacles –

human intellects – that can and do receive what issues from the

Active Intellect. Alternatively, one might reasonably claim that in

the case of “Unicorn,” “Phoenix,” and “Pentagonal House,” it is

merely one of my soul’s internal senses that serves as the cause that

makes them “necessary of existence through another” – my faculty

of “rational imagination”, say, which can separate, juggle, and combine

different pieces of abstracted sense data.29 Of course, if I am

not so committed to promoting Avicenna’s position (I) on essence

and existence, I shall be tempted to avoid all the foregoing problems

Cambridge Companions Online © Cambridge University Press, 2006

Cambridge Companions Online © Cambridge University Press, 2006

130 robert wisnovsky

by clearly restricting the distinction between (A) and (B–C) to concrete,

extramental beings.30 But then I am still left with the original

problem: (B) and (C) will not be convertible, since there will be as

yet unrealized entities – “Pentagonal House,” for example – that are

possible of existence in themselves but that are not (yet) necessary

of existence through another.

Despite this problem, Avicenna’s analysis of God as the necessary

of existence presented a golden opportunity to Sunnı¯ mutakallimu¯ n

struggling with the troublesome consequences of their theory that

God’s attributes were, in some real sense, distinct from God’s self.

Before Avicenna came along, Sunnı¯ mutakallimu¯ n had treated eternality

(qidam) as the most important meta-attribute – the kind

of attribute that is predicable both of God’s self and of God’s

attributes. (This position emerged because the Sunn¯ıs, in contrast to

the Mu‘tazil¯ıs, were committed to the traditionalist notion that the

Qur’ ¯an, conceived of as God’s attribute of speech, was not created,

and hence was eternal.) But if all – or even some – of God’s attributes

are eternal, then a risk arose of seeming to allow for a multiplicity

of uncreated, eternal things, which in turn infringed upon God’s

uniqueness as the sole possessor of eternality, that ultimate criterion

of divinity which set him apart from all other beings, which are

temporally originated (muh. dath).

The post-Avicennian Sunnı¯ mutakallimu¯ n realized that necessity

of existence could also be conceived of as a meta-attribute, and one

with fewer problematic implications than eternality. God’s self is

necessary of existence, as are God’s attributes; yet the attributes

are not necessary of existence in themselves, since they are merely

attributes and thus, strictly speaking, not “selves” but only predicated

of selves. Instead, God’s attributes are necessary of existence

in (or with – the Arabic preposition bi-means both) his self.31 Other

Sunnı¯ mutakallimu¯ n realized that necessity of existence could be

construed as a mode of the copula that bound the attribute (understood

as a predicate) to God’s self (understood as subject).32

Any mutakallim, Sunn¯ı or Sh¯ı‘¯ı, interested in taking up

Avicenna’s new distinction still faced one crucial challenge. The

Qur’ ¯an describes God as possessing not only the attribute of causal

power (qudra), but also the attribute of will (ira¯da). One theme that

runs through much of the first part of Ghaz¯ al¯ı’s Incoherence of

the Philosophers is that Avicenna’s conception of the relationship

Cambridge Companions Online © Cambridge University Press, 2006

Cambridge Companions Online © Cambridge University Press, 2006

Avicenna 131

between God and the world – a relationship between a being which

is necessary of existence in itself, and all other beings, which are

necessary of existence through another – robbed God of any true

agency. It is not enough, according to al-Ghaz¯ al¯ı, to conceive of God

merely as a cause; we must also conceive of God as an agent and

thereby make some room for God’s will. Al-Ghaz¯ al¯ı reckoned that

Avicenna expected too much from Muslim intellectuals, by forcing

them in effect to reduce all the names and acts of God so clearly

and powerfully described in the Qur’ ¯an to one single, simple name –

the necessary of existence in itself – and to one single, simple act –

self-intellection. What would be left of Islamic doctrine after such a

radical reduction?

In particular, Avicenna’s conception of the relationship between

God and the world entailed the denial of God’s most important

act, namely, his having created the world at some moment in the

past. This is because Avicenna’s distinction between the necessary

of existence in itself and the necessary of existence through another

was expressly designed not to map onto the mutakallimu¯ n’s distinction

between the eternal (qad¯ım) and the temporally originated

(muh. dath). Remember that one of Avicenna’s objectives was that his

new distinction would be able to differentiate between an uncaused

being – God – and all other, caused beings, eternal or temporally

originated.

The balancing act that preoccupied post-Avicennian mutakallimu

¯ n, both Sunnı¯ and Shı¯‘ı¯, was to make good use of Avicenna’s

compelling and innovative analysis of God as the only being which

is necessary of existence in itself, while at the same time ensuring

that they had not thereby committed themselves to the idea that God

necessitated the world’s existence. That is to say, the Sunn¯ı and Sh¯ı‘¯ı

mutakallimu¯ n were entirely comfortable appropriating and naturalizing

Avicenna’s analysis of God as the necessary of existence in

itself, as long as that formula was understood in an intransitive and

not a transitive way: it should point to God’s transcendence of and

separateness from the world, but not indicate the manner in which

God was causally involved with the world. The mutakallimu¯ n felt

that Avicenna’s distinction was useful because it provided the basis

for an excellent proof of God’s existence, and more generally because

it provided what appeared to be a watertight method of differentiating

between God’s existence and the world’s existence. But as an analysis

Cambridge Companions Online © Cambridge University Press, 2006

Cambridge Companions Online © Cambridge University Press, 2006

132 robert wisnovsky

of theway God causes the world’s existence, it required serious qualification,

since it could be taken to imply that God’s causation of the

world was no more than an involuntary act of necessitation, rather

than a voluntary act of agency.

At first glance, then, one of the great challenges of naturalization

facing the Sunnı¯ and Shı¯‘ı¯ mutakallimu¯ n of the post-Avicennian era

was to find some way to avoid throwing out the baby – the idea that

God is the only being that is necessary of existence in itself – when

they threw out the bathwater – the idea that God eternally necessitates

the world’s (eternal) existence. In fact, this was done with relative

ease. The Sunn¯ı-M¯atur¯ıd¯ı mutakallim al-Nasaf¯ı (d. 1114–15),

for example, incorporated Avicenna’s distinction into a section of

his Manifesto of the Proofs in which the compositeness of all beings

other than God – their being composed of motion and rest, or of

substance and accident – point to their not being necessary of existence

in themselves.33 And the Sh¯ı‘¯ı mutakallim¯ un H.

ill¯ı (d. 1326)

and La¯hı¯jı¯ (d. 1661) in their comments on al-T. u¯ sı¯’sOutline ofDogma

(Tajrı¯d al-i‘tiqa¯d, also known asOutline ofKala¯m– Tajrı¯d al-kala¯m),

both insist that asserting that God is necessary of existence in itself

does not entail God’s necessitation (ı¯ja¯b) of the world’s existence.34

T.

u¯ sı¯ himself had earlier tried to interpret Avicenna’s God as only an

efficient cause and not a final cause as well, probably because the