Metaphysicians, you might think, are interested in everything that exists. But actually this would be to underestimate them, because they are interested in things that don’t exist too. Things that are not yet existent, like tomorrow’s breakfast; things that might have existed but don’t, like Peter Adamson’s sister or Sherlock Holmes; and things that could not exist, like round squares. What is the ontological status of such items? A natural answer would be that they have no ontological status at all, precisely because they do not exist. But there are reasons to suppose otherwise, as was argued in the Islamic world well before Avicenna by Muʿtazilite theologians [T2–3]. One of the signature doctrines ascribed to “the Muʿtazilites,” even though in fact not all Muʿtazilites accepted it, was that the non-existent (al-maʿdūm) is a “thing (shayʾ).”1 Non-existent things are, of course, not existent (mawjūd), but they are “real (thābit).” Sometimes this is put by saying that the non-existent item is already an “object (dhāt).” The point then will be that existents are only a smaller subset of a larger class of entities, which are things or objects. Or to put it another way, being real does not imply being existent, the way that being existent implies being real.
The Muʿtazilites argued for this claim in part on a Scriptural basis. Examples would be the Qurʾānic statements that when God “wills a thing (arāda shayʾan) He says to it, ‘Be,’ and it is” (36:82, cited at [T63]),2 and that “God has power over every thing” (2:284). These and other verses may seem to imply that the object of God’s will (irāda) is already a “thing” before He creates it, though not everyone would agree with that interpretation [T39]. God also knows what He can create before He creates it. So if we make the plausible assumption that whatever is known is some “thing,” this will give us another reason to accept the Muʿtazilite doctrine [T11]. Also, if God doesn’t make things “real” by creating them He wouldn’t have to change when He does so; on the other hand, if He has to bestow existence on what is already real, that too could involve a change, as pointed out at [T52].
Of course these ideas are generalizable to other agents and knowers, so the doctrine could be supported on non-Scriptural grounds. The latter-day Muʿtazilite Ibn al-Malāḥimī sets out for the reasoning for this, though he will not accept it himself: if someone has the power to do something then there must be a relation between this prospective agent and the thing the agent can do [T11]. This will hold whether the relation is between God and the universe or between a carpenter and the table she plans to make tomorrow. A similar argument applies in the case of non-existing objects of knowledge, such as things one knows one could make. To make this line of thought still more plausible, we may add that objects of knowledge and power must already be distinct from one another [T2, T19, T32]. For example the carpenter can distinguish the table she’s planning to make from the chair she’s planning to make, and even think that she will make this table and not that table. Al-Rāzī gives a series of further illustrations at [T27].
Though all this might seem intuitively convincing, it was rejected by a number of authors. In some cases they simply denied that anything can be distinct or individual before existing [T22, T25, T37, T45]. One clever criticism, found in Ibn Kammūna [T56] among others, is that if an individual table were already real before the table is made, then the existence of that table should also be real—after all, the existence is possible just as much as the table is. And this is the existence of that table. So the table’s existence would already belong to the table before the table exists, which is absurd in itself, and leaves God with nothing to do when He creates something [T43, T50–51, T56]. An alternative explanation for how we make distinctions between non-existents was put forward by the Ašʿarites and is well explained by al-Shahrastānī [T20]. We do so on the supposition (taqdīr) that the things in question already exist. So if we contrast a non-existent chair and a non-existent table as two different pieces of furniture, we are talking in a counterfactual mode, and saying what would be true if the table and chair existed. For the Ashʿarites, being non-existent is the same as being “negated” and this is incompatible with being “real” [T38, T64].
Another worry about inferring reality from distinguishability is that even impossible things can be distinguished (though this is denied at [T60]). For instance a second God is impossible, and is different from other impossible things like a round square. But we surely do not want to say that impossible things are “real” [T29–30, T57, T65]. Or maybe we do? Impossible things are arguably objects of knowledge [T15–16], if only because we know of each of them that it does not exist. A related issue is the very idea of absolute non-existence itself, where what is at stake is not, say, this not-yet-existing table, or even an unspecified non-existent table, but nothing at all. One can think about this as simply the denial of existence, without further restriction. It seems unlikely that absolute non-existence is an object of power. What would it be if God or anyone else made it exist? But it’s harder to say whether or not absolute non-existence or “pure negation” is an object of knowledge [T35–36, T58, T61, T67–68].
One reason our authors were interested in the “absolutely non-existent” is that, as al-Samarqandī pointed out [T61], it seems to provide an exception to what we may call the “affirmation principle,” which is stated succinctly by al-Rāzī as follows: “everything of which a positive predicate is said inevitably is positive (kull mā kāna maḥkūm ʿalayhi bi-ḥukm thubūtī fa-lā budda wa-an yakūna thubūtiyyan).”3 This rule goes back to Avicenna, who remarked that “if an attribute is existent, that to which it is attributed is necessarily existent”4 (see also [T5]). The rule can of course be used to support the Muʿtazilite doctrine. If I can affirm predicates of non-existent things, then they must be “positive” or “real” (thābit) [T27].
Let’s stick then with true predications about non-existent things, like when we say that “Peter Adamson’s sister is human” or “tomorrow’s breakfast will be healthy.” What exactly makes it possible for us to assert such truths? This brings us to a final major argument for the reality of the non-existent, namely that things have at least some of their properties whether or not they exist. These will be their essential properties, for instance the rationality of Peter Adamson’s sister, which must belong to her even if she doesn’t exist, since all humans are rational and she is human. (Other properties, like the occupation of space, may emerge only once the thing exists [T2].) This notion of “essential independence” was present from early on in the kalām tradition [T2], and Avicenna’s essence-existence distinction offered a powerful framework for expressing it [T6–7]. Following Avicenna, for instance, ʿUmar al-Khayyām clearly accepted essential independence [T9] [T10].
If creation is simply bestowal of existence with an essence that already has its distinctive properties [T17, T21], this would explain why it is possible to distinguish between non-existing things: they already have the attributes that make them what they are. As al-Sāwī says, no cause is needed to make human to be animal, since it belongs to the essence of human to be animal. A cause is, rather, needed to make human exist, and any cause that does this will inevitably and simultaneously make an animal exist [T18]. Yet a number of texts included here question the notion of essential independence. One idea, closer in spirit to Avicenna’s apparent position, was to say that essences are indeed independent in respect of their essential features, but still need to be created by God [T49, 66]. Already Avicenna’s student Bahmanyār thought that essence or “true reality (ḥaqīqa)” occurs only along with existence [T8]. Similarly, the Ašʿarites argued that God gives things their properties, even the essential ones, by creating them, so that these properties belong to things only when the things exist [T4, T23, T40, T63]. This would avoid the consequence that essences already have some mysterious form of reality, hovering in logical space as it were, while they wait for God to give them existence [T24].
As for al-Rāzī, he bases his analysis on a distinction between “the essential (al-dhātī)” and “the essence (al-dhāt),” with the former being a part of the latter [T31]. Al-Rāzī differentiates between making a constitutive essential feature, making the whole essence, and making the necessary concomitants of that essence. So to borrow al-Sāwī’s example, animal is something “essential” to human and is a part of the essence of human. Thus one would need to make animal on the way to making human. Of course these two acts are never performed in isolation from one another, but they are distinct. This analysis was however rejected by al-Āmidī [T44] on the grounds that an essence is not a real composition of its essential features. Therefore al-Rāzī’s distinction has no bearing on the real act of the maker, which is just one act that causes the essence to be realized, with all its constitutive parts and necessary concomitants. Al-Ṭūsī also insisted that quiddities are indeed made by God when He makes the things with the quiddities exist [T53], the distinction between essence and existence being only conceptual. Asking a question like “does God make human to be human” is misleading, since for human to be an object of making, it must already be an existing quiddity [T54]. He presents this position as being, among other things, a critical response to Sufis who claimed to have had mystical experiences divulging that essences are not made, since they are eternally in God’s knowledge. Al-Ṭūsī finds the Sufi position to be suspiciously close to the Muʿtazilite one [T55].
It did not escape readers of Avicenna that his metaphysics, with its fundamental contrast between essence and existence, seemed a good fit for the Muʿtazilite doctrine that things are real before they are made to exist [T26]. Nor, it seems, did this escape Avicenna himself. Though he distanced himself from the Muʿtazila, he was clearly engaging with their doctrine in his own treatment of non-existence and “thing-ness (shayʾiyya).”5 But Avicenna made a significant contribution to the “realist” view on non-existents, by proposing that they in a sense exist after all (at least the possible ones). The reason they are “real” is that they exist in the mind [T5]. What we normally refer to as “existence” is in fact existence in concrete reality (fī al-aʿyān).6 This was seen by Avicenna’s successors as a real breakthrough, to the point that it became the majority view on the status of non-existent objects [T25, T46, T48, T59, T62, T68]. For them, even impossible objects may be supposed to exist in the mind [T46]. A view like this was embraced by Ibn al-Malāḥimī, following his “master” Abū al-Ḥusayn al-Baṣrī. He speaks of the non-existent as something “conceptualized (mutaṣawwar)” [T12]. When we use the standard Muʿtazilite argument for the reality of non-existents, that they are objects of knowledge, what we are talking about is concepts. Knowledge is a relation between the knower and the object, but the object could be conceptual or mental, rather than extramentally real [T13–15].7
Not for the first time, al-Rāzī emerges as the defender of an important minority position. He rejects the mental existence solution [T27] [T30], in part on grounds we have already mentioned: some non-existent things are impossible, and according to him these cannot exist even in the mind (at [T29] he is explaining the Avicennan position that he will attack in other works). But al-Rāzī does not want to retrench to the traditional Muʿtazilite view [T39], so he tries out a couple of alternative solutions. One, which he may or may not mean with full seriousness, is that the “non-existent” things actually are existent, but not in the usual way. They may be Platonic Forms [T33], a proposal that goes all the way back to antiquity [T1].8 But as al-Abharī pointed out, the example of impossible objects would seem at least as problematic for this proposal as it is for mental existence [T47]. Also in [T33] al-Rāzī speculates that non-existents could be in the corporeal realm but “hidden” from us, which seems rather ad hoc and would also not account for impossible items. Al-Kātibī nonetheless seems to like the idea, and interprets it as placing non-existent object in the Active Intellect [T48].
Fortunately al-Rāzī also has a more convincing alternative account, which is a kind of “reductionist” solution [T34].9 Take again the case of Peter Adamson’s sister. We are dealing here with a composite of two things that do exist, namely Peter Adamson and sisters. Since the “parts” of this complex notion exist, we are actually not talking about a non-existent thing after all. This proposal has the advantage that it could handle impossible or contradictory things: a round square would be simply a composition of round and square, both of which do exist. Al-Rāzī anticipates a possible objection, too, which is that some non-existents are simple, not composite: he gives the familiar example of a second God (or actually an “opposite to God”). But that too can be reduced to a something that does exist, namely the real God, since this non-existent is thinkable only on analogy to Him (see also [T16, T42]).
In all, the topic of the non-existent provides a nice window into philosophy in the 12–13th centuries: it displays the intimate connections between kalām and falsafa, with Avicenna’s view being assimilated to an earlier Muʿtazilite position. It also shows how an Avicennan view could be taken up by a wide range of intellectuals, many of whom did not consider themselves to be falāsifa. Characteristically, al-Rāzī emerges as both an acute expositor of other views and defender of an original account with some philosophical plausibility. And finally, the topic connects in obvious ways to more recent metaphysical reflections, especially those surrounding the Austrian philosopher Alexius Meinong (d. 1920) and his postulation of “intentional” objects that are real, but not existent.
Texts from: Aristotle ap. Alexander of Aphrodisias, Abū Rāshid al-Nīsābūrī, al-Anṣarī, Avicenna, Bahmanyār, al-Khayyām, al-Sāwī, Ibn al-Malāḥimī, al-Shahrastānī, Ibn Ghaylān, al-Suhrawardī, Fakhr al-Dīn al-Rāzī, Bābā Afḍal, al-Āmidī, al-Abharī, al-Kātibī, al-Ṭūsī, Ibn Kammūna, Bar Hebraeus, al-Nasafī, al-Samarqandī, al-Ḥillī.
Non-Existence
[T1] Aristotle, Peri Ideon, Reported by Alexander of Aphrodisias, On Met., 81.25–82.7 [trans. G. Fine]
[the non-existent objects of thought are Platonic Forms]
If, whenever we think of man, footed, or animal, we think of something that is, and of none of the particulars (for the same thought remains, even when they have perished), then clearly there is, besides the perceptible particulars, something we think of whether they are or are not. For we do not then think of something that is not. This is a form and idea. He says that this argument also establishes ideas of perishing and perished things, and in general of perishable particulars, such as Socrates and Plato. For we think of them and we retain an appearance of them and preserve it even when they no longer are. For there is some appearance even when they no longer are. Indeed, we also think of things that in no way are, such as hippocentaur and chimera. So neither does this sort of argument prove that there are ideas.
[T2] Al-Nīsābūrī, al-Masāʾil fī al-khilāf bayna al-Baṣriyyīn wa-al-Baghdādiyyīn, 12.6–19; 14.2–8; 21.15–22.2; 22.6–7; 23.4–7
[Muʿtazilite doxographical account of their own position]
Know that the doctrine of the two masters Abū ʿAlī [al-Jubbāʾī] and Abū Hāshim [al-Jubbāʾī] is that the substance (al-jawhar) is a substance while not existing (fī ḥāl ʿadamihi). The master Abū ʿAbdallāh [al-Baṣrī] asserted the same, though sometimes the apparent meaning of his statements suggests that the attribute of space-occupation (al-taḥayyuz) occurs for the non-existent. But any sort of qualification (al-ḥukm) that cannot occur for [the substance] as such, will only occur once [the substance] is existent. Therefore, he made “existence” a condition (sharṭ) for this qualification [i.e. space-occupation], as well for the fact that it bears accidents, and for the fact that it can be perceived by the two senses. [On the other hand], our master Abū Isḥāq [b. ʿAyyāsh] believed that the non-existent is not specified by any attribute which would distinguish it from other things. Rather it is distinguished only by a “prospective” attribute (bi-ṣifa muntaẓira). So he did not affirm for the substance an attribute additional to its being space-occupying, existent, or being in any given direction (jiha), and said that the disagreement concerns space-occupation only. Our master Abū al-Qāsim [al-Balkhī al-Kaʿbī] believed that the non-existent is described neither as substance nor as accident. He ruled out the application of any such name for the non-existent apart from our calling it “something” (shayʾ), “object of power (maqdūr),” “object of knowledge (maʿlūm),” and “that which becomes space-occupying (mutaḥayyiz ʿanhu).” Sometimes he also described it as “affirmed as real (muthbit),” since his doctrine concerning reality (ithbāt) differed from ours, according to the definition which we will mention later. […]
[God does not make the substance to be a substance]
[14.2] If someone says: why would [God] have to be able to bestow existence upon [a substance] and make it blackness, if He were able to bestow existence upon it and not make it a substance? It may be replied: if the substance could be a substance due to the Agent (al-fāʿil), and blackness could be blackness due to the Agent, then no generic attribute would be established for any object (dhāt) nor could we say that it is possible for one object and impossible for another. If this were the case, every object would have to be able to become a substance or blackness. The occurrence of such an object with this or that description would depend on the choice (ikhtiyār) of the Agent. […]
[distinction argument]
[21.15] Another proof [that the substance is a substance while not existing]: we have already shown that every object of knowledge can be known in a detailed way (ʿalā ḥadd al-tafṣīl). For if it cannot be known in a detailed way, then neither can it be known in general (ʿalā ḥadd al-jumla) as was made clear in the books. This being established, since we have learned that one can only know in a detailed way if [the object of knowledge] has an attribute by which it is distinguished from everything else, objects of knowledge in all cases, regardless whether they are existent or non-existent, must have attributes by which they are distinguished from everything else.
Another proof: if God the exalted wants to create a substance, He must surely intend (yaqṣidu) to bestow existence on something He knows will necessarily occupy space once [22] it exists. This would not be the case, if [substance] were not distinguishable for Him from other [kinds of entities], yet it can be distinguishable only if it is specified with an attribute. […]
[counterargument to the Muʿtazilite position and a response]
[22.6] Another [counterargument]: attributing existence (wujūd) to the substance is nothing additional (zāʾid) to its being a substance. So if one says that [substance] is a substance in all cases [i.e. even during non-existence], this is like saying that it is existent in all cases [including non-existence]. […]
[23.4] As for the second point mentioned, this is a serious mistake. For attributing existence to a substance is something additional to its being a substance. What proves this is what we already inferred about the fact that [the substance’s] being a substance cannot be due to the Agent, whereas it has been established that its existence is due to the Agent. An attribute that does occur due to the Agent cannot be the same as an attribute which cannot occur due to the Agent.10
[T3] Al-Shahrastānī, Nihāyat al-aqdām, 151.1–10
[a report of the Ashʿarite and Muʿtazilite positions on the non-existent]
The Ashʿarites do not distinguish between existence (al-wujūd), reality (al-thubūt), thing-ness (shayʾiyya), essence (dhāt), and concrete being (ʿayn). It was al-Shaḥḥām from the Muʿtazilites who came forward with the statement that the non-existent is something, an essence, a concrete being, for which he affirmed [in non-existence] certain specific features that one would [normally] associate with them in existence, like the subsistence of accident in substance, its being an accident and a color, and its being black or white. Most of the Muʿtazilites followed him on this point, except that they did not affirm [in non-existence] the subsistence of accident in substance, nor “space-occupying” for substance, nor its receptivity of accident. Another group disagreed, including those who refused to ascribe any [notion to the non-existent] apart from thing-ness. Others rejected even this ascription, like Abū al-Hudhayl and Abū al-Ḥusayn al-Baṣrī. Others said that “thing” is the everlasting (al-qadīm). As for that which is temporally originated (al-ḥādith), it is called “thing” only metaphorically, and in a broad sense. Jahm ibn Ṣafwān by contrast came to the idea that [only] the temporally originated is something, whereas God—may He be praised—is the one who makes things to be things (mashiʾ al-ashyāʾ).11
[T4] Al-Anṣarī, Ghunya, vol. 1, 284.2–10
[essential dependence and modalities]
If they say: attributes are divided into necessary (wājib) and possible (jāʾiz). The necessary ones do not depend on the Agent for its necessity. As for the possible ones, they depend on the volition (khiyāra) of the Powerful (al-qādir). If He wants, He provides them. If He does not want, he does not. The only [attributes] that behave like this are origination (al-ḥudūth) and existence (al-wujūd). We say: you are passing arbitrary judgment. For what you called “necessary,” like being a substance, being an accident, being blackness, and so on, is on our view merely possible. All of these depend on the volition of the Powerful. The existence of a substance and an accident means nothing but they themselves (dhātuhumā). Nor does their being brought into existence mean anything other than making them real objects (ithbāt dhātihimā), after they were not any “object,” “thing,” or “itself.” As for those attributes you affirm as following upon origination, like space-occupation, the subsistence of the accident in the substance, and so on, you have said that they necessarily follow origination. Yet, the attributes which you called the attributes of the things themselves (ṣifāt al-anfus), like being a substance, being an accident, and being an object (dhātiyya), according to our view, follow origination in just the same way as space-occupation and the subsistence of an accident in a subject of inherence.
[T5] Avicenna, Shifāʾ, Ilāhiyyāt I.5, 25.8–26.17 [trans. Marmura, mod.]
[the non-existent is mentally existent]
It’s true to say that “the thing is that about which information is given.” But when it is further said, “the thing may be absolutely nonexistent,” this stands in need of investigation. If by “the nonexistent” is meant the nonexistent in concrete individuals, then it could be the case, for it is possible for a thing that does not exist extramentally to exist in the mind. But if anything else is meant, this would be false, and one could give no information about it at all. It is known only as [something] conceptualized in the soul. As for the notion that [the non-existent] might be conceptualized in the soul as a form that refers to some external thing, definitely not.
This goes for the informative statement because information is always about something realized in the mind. One never gives affirmative information about the absolutely nonexistent (maʿdūm muṭlaq). And even if information about it is given negatively, then in some respect it is given existence in the mind. For our saying “it” entails a reference (al-ishāra), and it is absurd that there should be reference to the nonexistent that has no form in the mind, in any respect. For how could one apply “thing” to the non-existent, given that when we say “the nonexistent is such-and-such,” the meaning of this is that the description “such-and-such” occurs to the nonexistent (ḥāṣil li-al-maʿdūm)? There is no difference between the occurring (al-ḥāsil) and the existent, so it would be like our saying, “this description exists for the nonexistent.”
[the affirmation principle]
In fact, we say: [26] what describes the nonexistent and is predicated of it must either exist for the nonexistent and occur for it, or not exist and not occur. If it does exist and occur for the nonexistent, then it must, in itself, be either existent or nonexistent. If it is existent, then the nonexistent would have an existing attribute (ṣifa). But, if the attribute exists, then that to which it is attributed necessarily exists (idhā kānat al-ṣifa mawjūda fa-al-mawṣūf bihā mawjūd). The nonexistent would, then, be an existent, but this is impossible. If [however] the attribute is nonexistent, then how can that which is in itself nonexistent exist for something? For that which in itself does not exist cannot exist for the thing. […]
[26.11] We say that we have knowledge of the nonexistent simply because, when the concept (al-maʿnā) arises in the soul alone, without any reference (lam yushar) to anything external, then what is known is just that very thing in the soul. The assent, which is made between two parts of what has been conceptualized, is to the effect that in the nature of the thing known, it could have an intelligible relation to what is external, even though there is at the moment no such relation. What is known is nothing but this.
According to those who uphold this [other] view, there are items among the set of all objects of information and knowledge that have no thing-ness in non-existence. Whoever wants to be acquainted with their doctrine should go have a look at the raving remarks they have made, which hardly deserve any attention.
[T6] Avicenna, Dānishnāma, Manṭiq, 15.1–16.5
[essential independence]
The third [condition for essentiality] is as follows. You know that there is nothing that would bestow a given notion (maʿnā) upon a given particular, rather this belongs to it by virtue of itself. For instance you know for sure that nothing makes human to be animal, or four to be number. Otherwise, in the absence of that thing, human would not be animal and four would not be number, which is absurd. The meaning of our saying that one thing makes another thing to be “such-and-such” is that the latter thing is not “such-and-such” in itself, but rather something outside it makes it “such-and-such.” If something in itself cannot help being “such-and-such,” then nothing makes it “such-and-such.” Of course whatever makes a human also makes an animal, but it does not make human [16] to be animal, since human is animal in itself, and four is number in itself, and black is color in itself. This is not like white for human, since there is something which makes human white, which is either within the nature of [human] or outside it. Nor does being (hastī) belong to human in this way, for there has to be something which bestows being upon human.
[T7] Avicenna, Dānishnāma, Ilāhiyyāt, 38.10–39.3
[essential independence]
This notion of being (hastī) is neither essential (dhātī) nor is it the quiddity (māhiyyat) for those ten categories. We have shown this above. So one cannot say something made human a substance, or made blackness a color, but one can say that something made [them] existent (mawjūd). Thus each of the ten [categories] has a quiddity that is not due to anything [else], such as the fact that four is four, or that it is a number with some given description. [39] In Arabic its being is called “that-ness” (anniyya). Quiddity is one thing, that-ness another. The that-ness of [the ten categories] is distinct from quiddity, since [that-ness] is not an essential notion; so it is an accidental notion.
[T8] Bahmanyār, Taḥṣīl, 283.17–284.2
[bestowal of existence]
When the agent (al-fāʿil) [284] gives (afāda) existence [to something], he just gives [it] its true reality (ḥaqīqa), and its true reality just is its “being existent (mawjūdiyya).” From all this it is clear that the existence of something is its being among concrete individuals, not something through which it is (mā yakūnu bihi) among concrete individuals.
[T9] Al-Khayyām, Risāla fī al-wujūd, 105.10–106.2; 110.8–13
[argument for essential independence, with response]
They said: for humanity, existence is a notion acquired (maʿnā muktasaba) from something else. Yet animality and rationality belong to it by itself. Nothing made (yajʿulu) [human to be animal], nor did anything cause [it]. For the Creator, who is great, did not make human to be a body, for instance. Rather He made [human] to be existent. Furthermore, when the human exists, he cannot but be a body. So they said: things being so, [106] existence must necessarily be a notion additional to the human among concrete individuals. How could it be otherwise given that [existence] is nothing other than something acquired (maʿnā mustafād) from something else. […]
[110.8] As for the solution to the doubt raised by the People of Truth—namely that existence is something acquired and not anything else, so how can it fail to be additional12 among concrete individuals, on this account—it is that it is the object itself (dhāt) which is acquired, nothing anything else, since the object is non-existent and then exists, so that the object itself is acquired.
[T10] Al-Khayyām, Jawāb ʿan thalāth masāʾil, 167.8–9
[essential independence]
From this it is clear that if one says, “oddness necessarily exists for three-ness,” one means that it belongs to three-ness without any cause (musabbib) or maker (jāʿil), and the same holds of all essential and necessarily concomitant [attributes].13
[T11] Ibn al-Malāḥimī, Muʿtamad, 363.14–15 and 23–24; 364.21–24; 371.1–2
[Muʿtazilite arguments from connection, specification, and intention]
The one with the power (al-qādir) to create substance has it before [the substance] exists, and inevitably has a connection with that over which he has power (taʿalluq bi-al-maqdūr). Therefore the substance is already an object (dhāt) before it exists, so that the one who has the power [to create it] may be connected to it. […] [363.23] There must inevitably be some relation (iḍāfa mā) between the one who has power and that over which he has power (al-qādir wa-al-maqdūr), in order that it may be open to him to bestow existence. And a relation can only be affirmed of two objects (dhātayn). This relation is the connection (taʿalluq) we are referring to. […]
[364.21] Each of us knows what he does before he does it, and knows the acts of others before they bring them into existence. He can distinguish between what he has the power to do and what he cannot, and between the types (al-ajnās) of things over which he has power. Knowledge and the making of distinctions (al-ʿilm wa-al-tamyīz) inevitably have a connection to the non-existent (taʿalluq bi-al-maʿdūm), and there can be a connection only between two things (shayʾayn). So it is established that the non-existent is something (al-maʿdūm shayʾ), and is an object, while in the state of non-existence, in order that knowledge and distinguishing can be connected to it. […]
[371.1] The one with power wills (yurīdu) to bestow existence upon substance, or upon blackness, so inevitably there must be both objects, so that there may be an intention (qaṣd) to bring either of them into existence.
[T12] Ibn al-Malāḥimī, Muʿtamad, 353.23–354.6; 354.15–16
[agrees with Abū al-Ḥusayn al-Baṣrī that non-existent things are conceptualized in the mind]
The correct view on this issue is what our master Abū al-Ḥusayn [al-Baṣrī]—may God have mercy upon him—put forward. Namely that the non-existent is known, but is neither an “object” (dhāt) nor “thing” (shayʾ), as [those notions are understood] among the theologians (al-mutakallimīn). The fact that it is known means that it is conceptualized (mutaṣawwar), [354] but not that it is in itself a concrete item. If it is not in itself a concrete item, then neither does it become concrete through knowledge, in order that knowledge could be connected to it as to a concrete item. Rather it has the same status as what the thinking person conceptualizes in the cases of “secondary eternal feature,” “power,” “knowledge,” or “life” alongside [God], may He be exalted; or as what one conceptualizes in the cases of circles, shapes, or forms, before they arise [extramentally]. All this is conceptualized and real in the imagination (thābit fī wahm) of the thinking person according to our masters, the followers of Abū Hāshim [al-Jubbāʾī]. They are not “objects” or “things,” even though the thinking person does know them. Don’t you realize that we deny a second eternal [God], may He be exalted, and deny power and life of [the second God]? And one can’t make denials about that which one is neither thinking nor conceptualizing. […]
[354.15] It’s claimed that what is meant [in the Qurʾānic references to “thing” (shayʾ)] is that which is real in the soul, in terms of what is imagined and conceptualized (mā yuthbitu fī al-nafs mimmā yatawahhamu wa-yataṣawwaru). This is what the grammarians call “thing”.
[T13] Ibn al-Malāḥimī, Muʿtamad, 365.1–10
[knowledge as a relation to a concept, not to an object]
We generally concede your point that knowledge inevitably has a connection to what is known (taʿalluq bi-al-maʿlūm). But what do you mean by “connection of knowledge to the non-existent?” If they say: we mean by this that there has to be an “object” in non-existence and a “thing” that is itself concrete (shayʾ mutaʿayyin fī nafsihi), so that a relation may be established between it and the knowledge of it, since relation can be established only between two “things.” One may reply to them: prove first that the non-existent is something concrete in itself, so that you may establish that the knowledge of it has the aforementioned connection! For our part, we do not admit that the knowledge of the non-existent has such a connection [to what is known]. Rather, we say that knowledge is connected to what is known with respect to what that known is in itself. If [what is known] is concrete in itself, like existing things that are distinct from one another, either through themselves or through the attributes that are established in them, then knowledge is indeed connected to them as concrete items. But if [what known] is not concrete in itself, like that which is conceptualized and imagined, the knowledge does not have the sort of connection to it that it has to existent things. Rather, its connecting to it means its having a relation to what is conceptualized, not to that which is in itself concrete. And likewise for distinguishing [between objects of knowledge].
[T14] Ibn al-Malāḥimī, Muʿtamad, 369.21–370.3
[knowledge is by means of conceptualization]
Knowledge is of two kinds. First, knowing the true reality of something and its quiddity, like knowing the true reality of volume (al-ḥajm). This knowledge concerns neither (a) the existent nor (b) the non-existent, as you [sc. opponents who think the non-existent is extramental] would say. For (a) if we supposed that all substances are non-existent, the knowledge [of volume] would not be banished from the world. Whereas if it were knowledge of some [particular] existing volume, then with the elimination of what is known, the knowledge would also have to be eliminated. Nor is it (b) knowledge of the non-existent. For as we have seen, there is no volume in non-existence. So the right view is that knowledge concerns neither the existent nor the non-existent, rather it is the conceptualization (taṣawwur) of the true reality and quiddity of volume. The proof that knowledge of the true reality of something is not connected to a concrete item of knowledge (maʿlūm muʿayyin), but is rather conceptualization (taṣawwur), is that [370] every thinking person necessarily knows the true reality of the eternal and the true reality of the originated. If this had to be knowledge of some concrete known item, then since nothing is eternal but God, may He be exalted, this would have to be knowledge of Him, may He be exalted. But then it would follow that [all] responsible people (al-mukallafūn) must necessarily know God, may He be exalted, but this is false.
[T15] Ibn al-Malāḥimī, Muʿtamad, 356.9–17
[knowledge must have a relation to some object of knowledge]
Some followers of Qāḍī al-Quḍāt [ʿAbd al-Jabbār] defined the non-existent as what is known without being existent. They said: our saying “what is known” excludes an analogue (mithl) of God, the exalted. For that would not be “known,” since knowing that there is no analogue of God is a knowledge without anything that is known (ʿilm lā maʿlūm lahu). […]
[356.15] But their statement that knowing that there is no analogue to God is a knowledge without anything known, is not true either. For one cannot conceive of knowledge without any relation to what is known (iḍafa ilā al-maʿlūm). This will be either a concrete item or a product of conceptualization (immā mutaʿayyin aw mutaṣawwir). How can one speak about knowledge when there is nothing that is known?
[T16] Ibn al-Malāḥimī, Muʿtamad, 367.6–24
[rebuttal of alternative solutions for impossible objects of knowledge]
It is reported that some of them claimed: the knowledge that there is no second [God] alongside God the exalted is a knowledge which is connected to what is known (mutaʿalliq bi-maʿlūm). Further, some of them said that this knowledge concerns God the exalted Himself (bi-dhāt), to the effect that there is no second for Him. And some of them said it is rather the knowledge that substances and accidents do not resemble Him. But these two solutions are wrong. […]
[367.10] For the knowledge that is connected to something is either the knowledge of it in itself (bi-dhātihi), or is knowledge of it as having some attribute or state (ḥāla). But knowing that there is no second alongside God, the exalted, is not knowledge of Him in Himself, the exalted. For this is knowledge that there is no powerful, eternal object (dhāt) apart from God Himself, may He be exalted. So how can it be knowledge of Him in Himself, the exalted? Nor is it knowledge of Him as having some attribute or state. For according to them, His being one (kawn wāḥidan) is not among His states. Besides, in order to know this [i.e. that there is no second God] one has to know Him in Himself, the exalted. For this is knowledge that there is nothing analogous or similar to Him. And inevitably there must arise knowledge of Him, the exalted, in order that knowledge may arise that He has no analogue. Just as knowledge of the substance is inevitably involved in knowing that blackness does not resemble substances, even though the knowledge that blackness does not resemble substances is not just the same as knowledge of substances. So our own approach is proven: even if we assumed that God Himself were denied—and He is exalted above this!—still we could know that there is no second, eternal being in existence.
The second solution is also wrong. […] [367.20] If knowing that there is no second alongside God, the exalted, amounted to knowledge that substances and accidents do not resemble Him, then once one knows they do not resemble Him, one could no longer entertain the possibility that there exists some eternal being who is neither a body nor an accident and resembles [God]. But obviously one can entertain this possibility.
[T17] Ibn al-Malāḥimī, Muʿtamad, 376.7–21
[Can God make a thing to be something else?]
Then there is their argument that if the existence of a space-occupying item were identical to the [space-occupying] object itself (dhāt) and were due to the Agent, then he could make it be a space-occupying blackness, or bitterness. […]
[376.9] We say to them: even if we accepted that, for instance, space-occupation is an attribute and likewise for other attributes of “objects,” it would still be open to us to say that the Powerful [initially] makes the object, and only then does it have to be space-occupying, or motion, or blackness, once it has come to be an object (ʿinda kawnihi dhātan). […] [376.16] Further, we say to them: if we admitted that the Powerful makes the object according to the attribute of space-occupation, or the attribute of blackness, it would still be open to us to reply to you that the Powerful cannot make one and the same object be space-occupying, blackness, and motion in combination. If they ask: why not? For these attributes do not rule each other out, rather they can co-occur. Then we say to them: this is impossible given what we already mentioned, about the impossibility that would follow if an object were a body, blackness, and motion in combination. For the absurd (muḥāl) cannot follow from the possible (al-ṣaḥīḥ).
[T18] Al-Sāwī, Baṣāʾir, Manṭiq 37.22–38.2
[essential independence]
The third [feature of the essential] is that it is not acquired (mustafādan) from another thing. So, the human is not animal due to some cause that would make him animal. Rather he is animal in virtue of himself. For, if it were due to a cause, then it would be possible to suppose that the human is not an animal, by supposing the absence of the cause. Now, this does not mean that animal exists through itself, without any cause that would make it existent. Not at all! Rather, we have in mind that there is nothing that makes the human an animal. Admittedly, whatever makes a human also makes an animal, in making a human, since the human is an animal, and his origination is the origination of an animal. But if one were to say that [this cause] made a human, [38] and only then bestowed animality upon him, this is not so. For [in that case] humanity would be made subsistent without animality, and only then would animality be imposed upon it from the outside, which is absurd.
[T19] Al-Shahrastānī, Nihāyat al-aqdām, 152.15–153.5
[Muʿtazilite argument from distinctions in the non-existent]
The first [approach] is for you to affirm specificity and universality also in the non-existent, so that you will say that it includes what is necessary, like the impossible, and includes what is possible, like the contingent, and includes that which is impossible in itself, like the agreement of contradictories, and includes that which is impossible through another, like what differs from what is known to be the case. So these classifications are imposed on the non-existent. Thus you take non-existence in general, and specify it in these ways. Now, if the non-existent were not [153] something real (shāyʾ thābit), then universality and specificity could not be realized in it, nor could the distinction between one class and another.
[Muʿtazilite argument from connection]
The second approach is for you to concede that what is denied (manfī), and the non-existent, are known. You can impart information about it, and think about it. What is knowledge connected to [here], and what does connection even mean, if there is not at all something metaphysically real?
[T20] Al-Shahrastānī, Nihāyat al-aqdām, 153.6–14
[Ashʿarite response to the argument from distinctions: such distinctions are made within the non-existent only by supposing it to exist]
Those who deny [that the non-existent is something] say: we affirm in non-existence neither specificity nor commonality. Rather, its specificity and commonality comes down to mere verbal expression (al-lafẓ al-mujarrad) and intellectual supposition (al-taqdīr fī al-ʿaql). In fact knowledge is not connected to the non-existent as such, but only on the supposition of existence (ʿalā taqdīr al-wujūd). Unqualified non-existence is known and grasped intellectually by supposing unqualified existence, as opposed to specific non-existence, that is, the non-existence of something concrete. One may indicate a realized existent, speaking of “the non-existence of this thing,” or one may entertain it as a supposition in the intellect, and speak of “the non-existence of this supposed item.” For instance, “resurrection” is entertained by supposition in the intellect, and then one may deny it in the present, or affirm it in the future. Thus non-existence can be rendered specific, common, or known, only with respect to existence or the supposition of existence. Therefore, knowledge is connected to the existent.
[T21] Al-Shahrastānī, Nihāyat al-aqdām, 155.4–156.13
[things have intrinsic features without yet existing]
The secret of our doctrine is that substances and accidents have essential attributes which belong to them in themselves, rather than being dependent on an act of the Agent and the power of the Powerful. For we can conceive of substance as substance or as concrete (ʿayn) or as an object (dhāt), and also conceive of accident as accident, as object, and as concrete, without our considering it as something existent created by the power of the Powerful. That which is created and temporally originated requires the Agent in respect of its existence if and only if it is in itself contingent of existence and non-existence; it is when preponderance is placed on the side of existence that it requires the Preponderator. So the Agent has an effect (athar) through His capacity and power only on existence.
We say then: what belongs to [the originated] in itself was prior to existence. This is its substantiality or accidentality, which is something (shayʾ). What belongs to it due to the power of the Powerful is its existence and its occurrence (ḥuṣūl). What follows upon its existence is its space-occupation and its receptivity to accidents. This is a necessary intellectual judgment which no intellect can reject. On this basis one may answer the question about the effect of the bestowal of existence (al-ījād). For what is effected by [God’s] power is nothing but existence, and the Powerful bestows nothing except only existence. The contingent in itself requires the Powerful only in respect of existence. Don’t we say that the contingency of the contingent, insofar as it is contingency, is something it has through itself, and that in this respect it has no need for the Agent, as it is not up to the Agent to make it contingent? Rather it is for the preponderating (tarjīḥ) of one side of [156] the contingency [over the other] that it requires the Agent. So it is known with certainty that the essential features are not to be related to the Agent, but only whatever befalls them in terms of existence and occurrence. We say that if the Agent wants to bestow existence upon substance, He must inevitably distinguish (yatamayyaza) substance from accidence in its true reality, in order that His intention (al-qaṣd) of the bestowal of existence towards [the substance] may be realized. Otherwise, if substance and accident are not distinguished in non-existence by some feature (amr) and some true reality, and this feature and true reality are not something real (shayʾan thābitan), then He will not be able to form an exclusive intention for substance instead of accident, or motion instead of rest, or whiteness instead of blackness, and so on. Specification with existence (takhṣīṣ bi-al-wujūd) can be conceptualized only when what is specified (al-mukhaṣṣaṣ) is concrete and distinguished for the one who specifies (al-mukhaṣṣiṣ), so that you don’t wind up with substance instead of accident, motion instead of rest, or whiteness instead of blackness. Know then that the true realities of the genera and species are independent of the act of the Agent. For if things are not distinct in themselves, then bestowal of existence and creation (ikhtirāʿ) are inconceivable; instead the occurrence of things coming to be in various ways would be a matter of coincidence and luck (ittifāqan wa-bakhtan).
[T22] Al-Shahrastānī, Nihāyat al-aqdām, 159.15–160.5
[individuation argument]
We say: suppose somebody indicates a concrete substance (jawhar bi-ʿaynihi) and asks you whether this substance was, already before it existed, something real and a corporeal substance insofar as it is this one (hādhā), or instead absolute substance and something universal, not specified as this one. If you answer that it was the same concrete substance (bi-ʿaynihi jawharan), then its being indicated (ishāra ilayhi) as this one must already have been realized, [160] and that one, which was indicated (mushār ilayhi) [prior to existence], must have been this one, since nothing else shares in being “this one” which is not this one. If on the other hand it had been absolute substance before existence, and not this one, then that one [prior to existence] would not be this one [now being indicated], and this one [now being indicated] would not be something [real]. The absolute, as such, would not be this one and this one would not be that one, nor would that one be this one. So that which is real in non-existence would not have existence realized for it, and that for which existence is realized would not be real.
[T23] Al-Shahrastānī, Nihāyat al-aqdām, 160.6–161.15
[God gives things their essential properties in creating them]
They mentioned that the essential attributes [of created things] are not related to the Agent, rather it is existence that is related to Him. To this it is said: there are attributes affirmed of a non-concrete thing through which it becomes concrete; and then there are attributes affirmed of a concrete thing through which it belongs to a kind or species. The former are called “essential (dhātī) attributes” only in the sense than that they are expressions of it as a concrete object (ʿibārāt ʿan dhātihi al-muʿayyina). So its existence, substantiality, concreteness, and objecthood (dhātuhu) all express one and the same thing (ʿibārāt ʿan muʿabbar wāḥid). Just as it stands in need of the Bestower of existence for existence, so it needs [Him] for its objecthood, concreteness, and substantiality. Possibility of existence is the same (huwa bi-ʿaynihi) as the possibility of reality (thubūt). It is dependent on the Bestower of existence both for its concreteness, in order to be a concrete being, and for substantiality, in order to be substance. Otherwise it would follow that nothing depends on the Bestower of existence in any respect or with regard to any attribute, apart from existence. In fact, insofar as existence is a state (ḥāl) that does not have [a further] existence as an attribute, even existence would not require the Bestower of existence. So this would imply the existence of something eternal [other than God]. But instead, specified existence does require the Bestower of existence, due to the attribute of contingency. Furthermore, existence is not realized for the contingent existent in general (ʿāmm). Rather, [161] contingent existence is qualified with this or that attribute, and existence is realized for it.
Hence, it is established that a concrete particular requires the Bestower of existence. Yet existence is realized for it only when [the Bestower] wills it, and He wills it only when He knows it before bestowing existence upon it. So its existence is rendered specific (yatakhaṣṣaṣu) as an accident or a substance in the knowledge of the Bestower. Thus what is known is rendered specific as what is willed, and what is willed is specified as existence, which is the same as the substance. Not that it is in itself (fī dhātihi) a “thing” in such a way that it would be specified with existence only after it is “something.” Then extramental “thing-ness” would be specified as substance, and general extramental substantiality would be specified as this substance.
Again, existence is the most general of the attributes of existents. The bestowal of existence upon something more general does not entail the existence of something more specific. So if the white were related to the Bestower of existence only in respect of its existence, then it would only become existent [when God creates it], not white. Instead, if one puts it the other way around, saying that He bestows existence on it as black or white, and only whiteness [or blackness] is related to the Bestower of existence, but from the existence of the more specific the existence of the more general necessarily follows, then this would be more like it from a reasonable point of view. We say therefore that the white is opposed to the black in virtue of its whiteness; if one denies blackness, one denies existence. For it is unreasonable to deny blackness and leave existence standing.
[T24] Ibn Ghaylān, Ḥudūth al-ʿālam, 74.12–18
[the absurdity of real essences prior to existence]
Existence is among the universal features that exist in the mind. For it is among the features that all existents share, and the existents are distinguished by these features from the non-existents, which are conceptualized in the mind. The existence [of existence] is only in the mind, like “being color (al-lawniyya)” and the like. This is proven by the fact that, if the existence of objects were related to the Bestower of existence, and given by Him to them—or as they put it, “true realities are not acquired (mustafāda) from the Bestower of existence, rather it is existence that is acquired from Him”—then the objects would be in themselves things that [already] occur (ashyāʾ ḥāṣila), upon which existence would be emanated. In fact [existence] would be added to all its attributes, which would already be something (shayʾ) prior to existence, and only then would existence be emanated upon them. This is one of the false judgments of the estimative faculty.
[T25] Al-Suhrawardī, Talwīḥāt, 176.13–23
[arguments against taking the non-existent to be a thing]
It is also claimed that the contingent non-existent is a thing, while the negated is the impossible, and the contingent is real before existence.
[existence argument]
It may said to him: the non-existent quiddity is not existent, and its existence is negated and denied: yet it is contingent. This invalidates your claim. Furthermore, if “non-existent existence” is affirmed for a quiddity, as in his doctrine about the contingent, given that any attribute affirmed of something is attributed to it, it results that non-existence is attributed to existence, but this is absurd.
[individuation argument]
Furthermore, if that which is indicated by “this” is real prior to existence, then it is already “this,” so it would be existent prior to [its] existence. If on the other hand it is not [yet] “this,” then “this” would not be contingent [i.e. possible] prior to existence.
[mental existence solution]
Rather, what is existent in the mind may be non-existent in concrete individuals, and vice-versa.
[T26] Al-Rāzī, Muḥaṣṣal, 55.5–7; 59.2–14
[Avicenna as agreeing with the mainstream Muʿtazilite view]
The non-existent is either impossible of reality (mumtaniʿ al-thubūt), so that it is uncontested that it is pure negation (nafī maḥḍ), or it is possible of reality. According to us, and to Abū al-Hudhayl and Abū al-Ḥusayn al-Baṣrī from among the Muʿtazilites, it is pure negation; as opposed to what the rest of the Muʿtazilites think. The crux of the disagreement is that they claim that black’s existence is additional to its being black, and furthermore that this quiddity can be independent of the attribute of existence. […]
[59.2] Abū Yaʿqūb al-Shaḥḥām, Abū ʿAlī al-Jubbāʾī and his son Abū Hāshim, Abū al-Ḥasan al-Khayyāṭ, Abū ʿAbdallāh al-Baṣrī, Abū Isḥāq ibn ʿAyyāsh, al-Qādī ʿAbd al-Jabbār ibn Aḥmad and his students claimed that possible non-existents are already objects (dhawāt), concrete beings, and true realities, before entering into existence. And they claim that the effect of the Agent (taʾthīr al-fāʿil) does not consist in making them to be objects, but in making those objects to be existent. They agreed that these objects are distinct as individuals, and that the reality of every kind of these non-existents is infinite in number. As for the philosophers (al-falāsifa), they agreed that the contingent items are their quiddities, not their existences, and that it is possible to strip those quiddities of extramental existence. For we can intellectually grasp “triangle” even if it has no extramental existence. As for the question whether one can strip [the quiddity] of both extramental and mental existence, Ibn Sīnā wrote in the first Maqāla of the Ilāhiyyāt of the Shifāʾ that this is indeed possible, though some [philosophers] do not allow this. […]
[59.11] They also agreed that quiddities are not made (ghayr majʿūla). They said: whatever is necessitated through another is eliminated along with that other thing. If blackness belonged to black through something else, then when that other thing was eliminated, black would stop being black. But saying that black is no longer black is absurd.14
[T27] Al-Rāzī, Arbaʿīn, vol. 1, 91.5–92.23
[“distinction argument” that the non-existent is a thing]
Non-existent objects are in themselves distinguished, and inevitably, all [things] that are distinguished from one another are in themselves concrete true realities. This is what we mean when we say that the non-existent is “something.” There are several ways to show that non-existents are indeed distinguished one from another while not existing:
First: we know that tomorrow the sun will rise from the East, not from the West. Both of these sunrises are presently non-existent, yet we already know the distinction between them. This shows that distinction applies to the non-existent. Second: we are able to move to the right or to the left, but not to fly to the sky. So one of the non-existents is distinguished from another, given that we can do one and not the other. This distinction applies while they are not existing, so the application of distinction to the non-existent is established. Third: we find ourselves wishing for fortune, children, well-being and happiness to come to us, and that no sort of illness or misfortune befalls us, even as both types are non-existent. If one non-existent were not distinct from another through its concrete quiddity and specified true reality while failing to exist, it would be impossible to wish for one and worry about the other. Fourth: the non-existent is divided into two types, the impossible and the possible. Doubtless each [92] of these types is distinct from the other in itself and in its true reality. That is why the willing, capable [agent] cannot bestow existence on the impossible, but can do so for the possible. If there were no distinction between the possible and the impossible in themselves, this would not be true.
[rejection of the mental existence solution]
Let it not be said: these items, while not existing in extramental reality, do exist in the mind, which is why distinction may rightly be applied to them. For we say: you must say either (a) that these non-existents exist in the mind, or (b) that the knowledge of them exists in the mind. (a) But the first is wrong. For we know the sun and the moon. If these objects of knowledge existed in the mind, then when someone imagined many suns, or many moons, or a sea of mercury, or a mountain of jewels, these things would exist in his mind. This is necessarily known to be false. (b) As for the second option, namely that what is present (al-ḥāḍir) in the mind is the knowledge of these things, we concede this. But our investigation concerns the object of knowledge, not knowledge [itself]. For even without these objects of knowledge existing in the mind, we know that they are distinct in themselves, and that their true realities are distinct. It is irrelevant whether they exist in the mind or not. This is what was to be shown.
On the basis of these four demonstrations, it is established that one non-existent is distinguishable from another while it is non-existent. This being established, we say: the distinction (imtiyāz) of one thing from another depends upon each of them being in itself a concrete true reality and concrete quiddity. For their mutual distinction is one of the features (aḥkām) of those true realities, and one of their attributes. Yet it is absurd that an attribute or a feature can be affirmed (thubūt) without the bearer of the attribute being realized (taqrīr). So it is established that non-existents are distinct, and realized, and that distinction can only be accomplished once true realities and quiddities are realized. This means it surely must be the case that non-existents are objects (dhawāt), quiddities, and true realities. This is what was to be shown.
[T28] Al-Rāzī, Arbaʿīn, vol. 1, 95.11–25
[against the distinction argument in T27]
The answer is: we say that we find in our souls awareness (shuʿūr) and perception (idrāk) of many forms, and there is agreement between us and you that they are neither quiddities nor true realities. Rather, they are pure negation and outright non-existence. The first form [of pure negation] is knowledge about impossibilities. For we judge it impossible for there to be a companion (sharīk) of God, and provide proofs for this. If we couldn’t conceive the companion of God, we could not have predicated impossibility of it. For assent (taṣdīq) is impossible without conception. We say that the companion of God is impossible, that the conjunction of existence and non-existence is impossible, and that the presence of a single body in two places at the same time is impossible. And we distinguish between all these cases of assent. Mental awareness and intellectual distinction are present in these forms [of negation], even as all scholars (ʿulamāʾ) are in agreement that these impossibilities are neither objects, nor true realities, nor quiddities.
[T29] Al-Rāzī, Mabāḥith, vol. 1, 136.18–24
[mental existence solution, argument from impossibles]
As for the first argument they gave, it is based on the their ignorance about the fact that quiddity has existence in the mind, as we have already shown. This gains support from the fact that we can conceptualize the impossible, and imaginary forms like the form of Zayd and the form of ʿAmr or of a concrete horse, while [the Muʿtazilites] agree with us that these are not extramentally real. Likewise, if we intellectually grasp existence and non-existence, neither of them are real objects while they are non-existent. And there is no getting out of these convincing arguments that force absurd consequences on the opponent (al-ilzāmāt) with a merely verbal defense. So we know that these conceptualized quiddities are existent in the mind.
[T30] Al-Rāzī, Mulakhkhaṣ, fol. 80r14–16
[why objects of knowledge need not be real]
Their rationale is that the non-existent is an object of knowledge, and everything known is real. The major premise [sc. that everything known is real] is rejected with [the examples] of impossibles, imagined objects, and also existence itself, since it is an object of knowledge, but is not described as [itself] existent and real.15
[T31] Al-Rāzī, Sharḥ al-Ishārāt, vol. 1, 48.12–49.6
[essential independence]
I do not sufficiently understand this talk [about essential independence]. For the essential is a part of the essence, and an essence’s part is distinct from the essence [as a whole]. It is acknowledged that a relation to one thing is distinct from a relation to something else, especially for those who say that only one effect can proceed from one simple thing. From the fact that it is impossible to make the essence without also making the essential, it just follows that the act of making the essence is the same as the act of making essential. So it remains possible that the act of making the essence, while being distinct from the act of making the essential, depends on it and is posterior [49] to it. It is for this reason that the act of making an essence is impossible without the act of making the essential. Instead, the truth is that the act of making the essential is prior to the act of making an essence, and the act of making an essence is prior to the act of making a necessary concomitant, because quiddity is the cause of the concomitant. This is the truth.
The ancients might have meant by [essential independence] that the cause of the essential is a cause for the essence by means of those essential features: when the essential features are realized, there can be no delay before the realization of the essence, since the essence is an effect of the conjunction of the essential features; and when the essence is realized there can be no delay before the realization of the necessary concomitant of a quiddity, since the concomitant of a quiddity is an effect of the quiddity. This is my view on this issue.
[T32] Al-Rāzī, Sharḥ al-Ishārāt, vol. 2, 218.1–4
[distinction argument]
On might say: such conceptions are not distinguished one from another prior to their existence. Rather, before they exist, we know that they will be distinguished from one another, but only once they exist. But this is unconvincing. For if there is nothing to which the intellect can refer, or to which it can ascribe distinction, then it is absurd to judge that one of them will be distinct from the others once they exist.
[T33] Al-Rāzī, Mulakhkhaṣ, fol. 79r9–21
[Platonic Forms as an alternative to mental existence]
Those who believe in the reality of [mental existence] argue that we conceptualize things that have no extramental existence, and predicate of them distinction from everything else. So the conceptualized is existent, because positive predicates (al-aḥkām al-thubūtiyya) are predicated of it. It is not among concrete individuals, so it is in the mind.
Answer: we do not concede that we conceive of objects without extramental reality. Admittedly, they may not be present to us (ḥāḍira ʿindanā), but why can’t one say that everything that can be conceived of and imagined has an existent form that subsists by itself (ṣūra mawjūda qāʾima bi-nafsihā), or in some hidden body (fī shayʾ min al-ajrām al-ghāʾiba)? When the soul turns towards them, it perceives them. These are the exemplars (al-muthul) of which the great Plato speaks. We will mention Aristotle’s argument against them, with a response.16
Know that there are only three options concerning knowledge of non-existents: (a) if the non-existent knowable objects are not real at all, but one [non-existent] is distinguishable from another, then pure non-existence should be distinguishable in a way that if it were existent, then no [distinction] would be added to it. But this is wrong. So if [the non-existent] is real, then it is either (b) in the mind, and this is wrong, in light of the arguments already mentioned; or (c) in extramental reality. [In this case] it is either present (ḥāḍir), which is obviously wrong, or hidden (ghāʾib), which is what we have mentioned.
[T34] Al-Rāzī, Mabāḥith, vol. 1, 500.7–19
[non-existents are known by knowing their parts]
Every object of knowledge has to be distinguishable from other things. And everything that is distinguishable from another is existent. Therefore, every object of knowledge is existent. Or to turn this around: that which is not existent is not an object of knowledge. Yet we know many things that are non-existent, for instance we know the non-existence of a companion of God, and the non-existence of an agreement between opposites. So how are these two points compatible?
We say: The non-existent has to be either (a) simple (basīṭ) or (b) composite (murakkab). (a) If it is simple, like the non-existence of an opposite to God, this can be intellectually grasped only thanks to its resemblance to something that does exist. For instance one might say: for God the exalted there is nothing whose relation to Him would be like the relation of blackness to whiteness. If we didn’t know the opposition which occurs between existent things, it would be impossible to know the non-existence of an opposite to God, the exalted.
(b) If [the non-existent] is composite, like knowing the non-existence of an agreement between blackness and whiteness, then knowledge of this is accomplished only due to the knowledge of its existing parts (bi-ajzāʾihi al-wujūdiyya). For instance, we may intellectually grasp “blackness,” “whiteness,” and “agreement” on a given occasion. Then one may say that “agreement,” which is grasped as something existing, does not arise out of “blackness” and “whiteness.”
The upshot then is that the non-existence of the simple is known only through an analogy (bi-al-muqāyasa) drawn to something existing, while the non-existence of the composite is known only by knowledge of its simple [parts].17
[T35] Al-Rāzī, Mabāḥith, vol. 1, 130.16–21; 131.2–10
[truths about negation as counterexamples to the affirmation principle]
[The rule is wrong] because negation (salb) is said to be opposed to affirmation, but negation has no reality in itself, despite opposition to affirmation being attributed to it.
If you say: negation has an intellectual form, and has reality in the intellect, then we [the opponents] say: to the contrary, its being real in the intellect does not make it opposed to reality. Rather, it is a type of [reality]. Indeed it must be real, insofar as it is opposed to reality. […]
[131.2] Response to the second [argument]: the mind makes a form present to itself, and predicates of it that this form is not grounded in the extramental, and that there is nothing extramental that corresponds to it. This then is what one means by “conceiving of negation.” Then, [the mind] makes another form present to itself, and predicates of it that there is something extramental that does correspond to it. Then [the mind] predicates of one of the two forms that it is opposed to the other: not insofar as both of them are present in the intellect, but insofar as one of them is grounded in the extramental, the other not. The subject of the predication of this opposition is an intellectual form that exists in the described way [i.e. only in the mind]. This is what the philosophers (al-ḥukamāʾ) mean when they say that the opposition between negation and affirmation is realized only in utterance and thought, but not extramentally.
[T36] Al-Rāzī, Mulakhkhaṣ, fol. 80v21–81r11
[is absolute non-existence an object of knowledge?]
It is well known that absolute non-existence cannot be known, nor can any informative statement be made about it (yukhbaru ʿanhu). Rather it is the non-existence that is related to existents that is known, and about which an informative statement is given. But this calls for a twofold investigation. (a) When we say “no informative statement about absolute non-existence may be given,” this is already an informative statement about it, so this is a contradiction. (b) Non-existence is a part of the notion (mafhūm) of relative non-existence. And you have to understand something before you can understand it as being related to something else. So relative non-existence may by understood only after non-existence as such is understood; thus absolute non-existence is inevitably an object of knowledge.
Yet there are also problems with saying that absolute non-existence is an object of knowledge. For absolute non-existence is neither concrete, nor does it have reality or distinction. How then can the intellect refer to what cannot at all be distinguished or be concrete? As for talk of “mental form,” you already heard which [problems] arise here. But even if one concedes [that there is a mental form], problems remain. For mental form can be grasped by the intellect correctly only if it corresponds to the extramental. But this can be accomplished only if something is in fact realized extramentally. This is the crux of the problem, and we ask God the exalted to help us to deal with it.
[T37] Al-Rāzī, Muḥaṣṣal, 56.5–10
[individuation argument]
The non-existent black is either one or many. If it is one, then if unity is a necessary concomitant of [its] quiddity, it cannot perish, so that [blackness] could not be multiplied numerically in existence. If however [unity] is not a necessary concomitant, then one can suppose it to be removed, since if something is contingent, no absurdity follows from its removal. And if unity is removed, multiplicity is the result. But [multiplicity] can be realized only if there are two things that are distinct in concrete being (bi-al-huwiyya). Next, if the reason for the distinction [between the two non-existent blacks] is a necessary concomitant of the quiddity [of blackness], then the two would differ in quiddity, which is a contradiction. If on the other hand [the reason for the distinction] is not a necessary concomitant [of the quiddity of blackness], then something would be subject to temporary attributes while not existing. If one allows this, one should also allow it to be a subject of inherence for motion and rest, succeeding one another, all in pure non-existence. This is nothing but sophistry.18
[T38] Al-Rāzī, Mulakhkhaṣ, fol. 80r11–14
[specificity and generality argument]
If the non-existent is equal in scope to the negated (al-manfī), or more specific than it, and whatever is negated is not real, then nothing non-existent is real. If however it is more general than [the negated], it cannot be pure negation, for otherwise there would no longer be any difference between the general and the specific. Thus [the non-existent] would be real and would be said of what is negated, so the negated would be real, which is a contradiction.
[T39] Al-Rāzī, Arbaʿīn, vol. 1, 90.20–91.3
[God’s power over non-existents]
The statement of [God] the exalted, “God has power over every thing” (Qurʾān 2:284), yields an argument from the verse (al-āya), namely that the word “thing” ranges over quiddities. So it follows that God the exalted has power over these quiddities. But for His power over them to be genuine, it must be able to have an effect (tuʾaththira) on these quiddities in respect of their occurrence and failure to occur (taqrīran wa-ibṭālan). Given [91] this, God the exalted is prior to the occurrence of these quiddities, because that which has an effect is prior to the effect. Given this, it is established that quiddities are, one and all, pure negation and absolute non-existence in eternity.
[T40] Al-Rāzī, Arbaʿīn, vol. 1, 99.17–22
[God’s power over quiddities]
The Powerful makes quiddity to be quiddity just as much as He makes quiddity to be existent. The argument you use to reject the dependence of quiddity upon the Agent would also imply the rejection of the dependence of existence upon the Agent. For if existence depended on the Agent, then existence would cease to be existence as soon as one assumed the non-existence of the Agent, but this is absurd.
[T41] Bābā Afḍal, Taqrīrāt, 648.8–650.8
[endorses non-existent things]
As for the first option, which is that [“thing” (chīz) and “existent” (mawjūd)] are equivalent in respect of generality and specificity, this is false. For the condition for equivalence in generality and specificity is that truth is preserved through conversion. For instance, you might say “every body possesses dimensions,” and when you convert it to “whatever possesses dimensions is a body” it remains true. But this not so in the case of the expressions “thing” and “existent.” For when you say “whatever is existent is a thing,” this is true. But when you say “whatever is a thing is existent,” this is untrue. For if we take something that does not exist outside the soul, and whose becoming existent is contingent, insofar as it is contingent, and the contingency is its attribute, then it is a thing. Although it is a thing from the point of view that it is contingent, it is not existent. So, from the point of view that it is contingent, it is a thing and the expression “thing” truly applies to it, while from this same point of view, that it is contingent, the expression “existent” does not truly apply to it. So it was right to deny that “every thing is existent,” even though it is true to say that “every existent [649] is a thing,” given that as we have said, whenever something contingent does not exist, it is a thing without being existent. […]
[649.8] Another argument that “thing” is more general than “existent.” If you say that one of the following three conditions must apply to “thing”: either (a) its existence is necessary, or (b) its existence is contingent, or (c) its existence is impossible, then this division is correct, without repetition, nor is any of the options false. Yet if you say that “existent” is one of the three, either its existence is necessary, or contingent, or impossible, the third option is false, since the existent cannot fall under to the option of the impossible. […]
[650.1] Having shown that “thing” is more general than “existent,” we need to realize that existence is an attribute of a “thing,” which is related to it in such a way that one can say “the being of the thing (hastī-yi chīz)” but not “the thing of the being.” This relation indicates beforeness and priority. So thing comes first and being comes after. But one of these two must be the case: either [“thing”] is prior in respect of being, or in respect of something else. Yet it cannot be prior in being, because being is precisely what comes after the “thing”; so there is nothing else apart from the thing-ness (shayʾiyyat), which is the thing itself (dhāt-i chīz). So it is in itself (bi-dhāt) that “thing” can have priority over the attribute of existence.
[T42] Al-Āmidī, Abkār al-afkār, vol. 1, 107.3–17
[conception of the impossible]
The intelligent scholars agreed that there cannot be any knowledge without objects of knowledge (lā maʿlūm la-hu). However Abū Hāshim disagreed with them about the case of knowledge that is connected to the impossible, refusing to define “thing” as the object of knowledge. He said: knowing that co-occurrence of the opposites is impossible, and the denial of a second God (may He be exalted), or that part cannot be equal to the whole, and so on, constitutes knowledge without any object of knowledge. At the same time, he agreed that knowledge is connected with the impossible, and that were are its knowers. So he agreed in the meaning but disagreed in expression. For that [the impossible] is known means simply that knowledge is connected to it. Otherwise, how can the existence of knowledge without the object of knowledge be conceptualized, since the knowledge and the object of knowledge belong to the class of correlative terms, where neither of them can be grasped intellectually without the other? […]
[reductionist solution]
[107.14] Knowing that two opposites cannot co-occur is [just] the knowledge of the opposites, and they both are objects of knowledge. Knowing that there cannot exist any second God (may He be exalted) is the knowledge of the existence of God, and it is the object of knowledge. Similarly, the knowledge of every instance of the impossible has to be knowledge that is connected to an object of knowledge. It is absurd to say that there is knowledge without any object of knowledge.
[T43] Al-Āmidī, Ghāyat al-marām, 244.14–245.24
[two arguments against the reality of the non-existent]
The true opinion and the correct approach is to say: if objects (al-dhawāt) were real in non-existence, then once they do exist, either something new would occur (yatajaddada) to them which they lacked while they were non-existent, or not.
(a) If one says that the first is the case, then [what is newly added] would also have to be either a substance or an accident, or a state (ḥāl) additional to these two. But it can be neither substance nor accident, because they have been already supposed as real from the very beginning. After all there is no distinction in this respect between one substance and another, nor between one accident and another. If however [what is newly added] were an additional state, this would be based on the theory of states, which has been already refuted.
(b) If one says that the second is the case, then there would be no difference between existence and non-existence, which is absurd. And thus any talk about origination and existence would be absurd. This absurdity is forced only on those who posit entities as real in non-existence and realized in eternity. So they are not real, and realization is through origination, and reality only applies to the substantial and accidental objects themselves, nothing else.
Furthermore, let’s take blackness and whiteness and say: if they were [245] real objects in non-existence, then either they would require a subject of inherence, in which they would subsist, or not.
(a) But it is impossible that they do not require [it]. For otherwise, once they exist, either they [still] require it, or not. (a1) But it is absurd to say that they don’t require it. Otherwise, there would be no difference between substances and accidents. (a2) But if they do require it, then they must do this either in light of their essences (bi-iʿtibār dhawātihā), or in light of something having to do with their existence (bi-iʿtibār amr wujūdihā). (a2a) But they cannot require it with respect to their existence, since existence as such is, according to the opponent [i.e. the Muʿtazilites], a single judgment that is common to both substance and accident. So if an accident requires the subject of inherence with respect to its existence, then substance would too, but this is impossible. So it remains only (a2b) that [blackness and whiteness] require a subject of inherence [when being existent] due to their essences. But then there would be no difference between their existing and their not existing. For what would require [a subject of inherence] while existing would be exactly the same as what is real while it is non-existent.
(b) If on the other hand they do require a substrate of inherence in which they subsist, and if we posit blackness and whiteness as succeeding one another in a single substrate of inherence, on the side of existence, then either they would already subsist in this subject of inherence before existing, or one of them would subsist in this subject, and the other in another subject. (b1) It cannot be that one of them subsisted in this subject of inherence and the other in another. Otherwise, when [the second one] came to exist in this subject, it would have to undergo transmission (al-intiqāl), but the transmission of accidents is absurd. (b2) So it remains that both [blackness and whiteness] really co-occur in [the same substrate of inherence] in the state of non-existence. But if this were the case, one could not say that their co-occurrence is impossible when they exist. For this impossibility would be either due to their essence, or due to their existence. (b2a) But their mutual exclusion and contrariety cannot be due to their existence, since existence applies to them both with the same undifferentiated meaning. (b2b) So it remains that their mutual exclusion is due to their essence, so that if there were no contrariety while they do not exist, there would be no contrariety once they do exist, either. But their mutual exclusion and contrariety is real when they do exist, so it is real when they do not exist, too. So it follows from the fact that they are real in non-existence that they cannot be in the same subject of inherence, given the necessity of contrariety. (b1 conclusion repeated) Nor can they be in two different substrates, given that an impossible transmission [of accidents] would be necessary once one supposes that they succeed one another. But it follows from the impossibility of their subsistence in a subject of inherence that they cannot be real in themselves, given that they can have neither subsistence nor reality apart from non-existence. This was the desired conclusion.
[T44] Al-Āmidī, Daqāʾiq al-ḥaqāʾiq, Manṭiq, 56.14–24
[essential independence: reaction to T31]
If someone says: the essential is a part of the essence, but a part of an essence is distinct from [that] essence, and a relation to one of two distinct things is distinct from a relation to the other. Hence, even if one rules out that an essence be made without making the essential,19 it does not follow that making the essence is just the same as making the essential.
We say: if, by saying “the essential is a part of an essence, but a part of an essence is distinct from [that] essence,” he intends that the concept (al-mafhūm) of the essence can be detached from the concept of the essential, he is mistaken. Otherwise the essence could be realized without the essential [being realized], which is absurd. But if he intends that the concept of the essence is additional to the concept of the essential, even though the latter enters into the concept of the former, then this is true. This being so, we know that an essence comes down to the conjunction of the essential features, and it is not distinct from them. This being so, the maker of an essence through an act of making20 is either making its essential features through that act of making, or is not the one who does this through that act of making. But if it is not their maker through that act of making, then neither is it the one who makes the essence, since the essential enters into the concept of the essence. But this conflicts with what was assumed. So it remains only that it is their maker through that act of making.
[T45] Al-Abharī, Kashf al-ḥaqāʾiq, 252.2–9
[individuation argument]
Contingent quiddity is not realized among concrete individuals without existence. For whatever is realized among concrete individuals possesses some concrete being (huwiyya) in the concrete individuals. And whatever has concrete being in concrete individuals is individuated (mushakhkhaṣ) or is disposed to individuation. Whatever is like this is existent. But no quiddity that lacks existence is existent, so no quiddity that lacks existence is realized among concrete individuals.
[T46] Al-Abharī, Muntahā al-afkār, 283.19–23
[the impossible and mental existence]
The impossible has a form in the intellect. Otherwise one could not predicate impossibility of it. The same goes for the non-existent. In refutation of mental existence, it was said that if the hot and cold existed in the mind, then opposites would co-occur (ijtimāʿ al-ḍiddayn) in a single locus of inherence (maḥall).21
But in response we say: we do not concede that there is an opposition between the universal hot and the universal cold. Rather one affirms opposition only between extramental hot and cold.
[T47] Al-Abharī, Kashf al-ḥaqāʾiq, 251.6–9
[against al-Rāzī’s appeal to Platonic Forms]
Regarding his statement: why do you say that we can conceive of things that have no extramental existence?22 We say: because we can conceive the aforementioned notions (al-mafhūmāt), some of which cannot exist extramentally. But these impossible things subsist neither in themselves nor in hidden bodies.
[T48] Al-Kātibī, Ḥikmat al-ʿayn, 5.14–16.9
[mental existence as an account of non-existents]
You should know that we conceptualize things that lack existence in extramental reality, and we predicate positive predicates of them. But the subject of an existential attribute (ṣifa wujūdiyya) must itself be existent, because to affirm an attribute of something presupposes affirming [6] that thing. Since it is not in concrete individuals, it has to be in the mind. Thus is established the theory of mental existence; and also because the universal true realities have existence only in the mind, given that everything that exists among concrete individuals is individual. […]
[response, echoing al-Rāzī’s appeal to hidden objects]
[6.7] What one ought to say: we do not concede that we conceptualize things that lack existence in the extramental reality. Rather, whatever we conceptualize has an existent form, either self-subsisting or in some existent thing that is hidden from us. This is what the philosophers (al-ḥukamāʾ) thought, for they agreed that all things are contained in the Active Intellect.23
[T49] Al-Ṭūsī, Talkhīṣ al-Muḥaṣṣal, 84.8–13
[the sense in which God makes things to be what they are; reaction to T26]
Those who say that quiddities are not made, do not say that they are not originated (mubdaʿa). Rather they say: when a quiddity is posited (furiḍat) [as existing], then its being that quiddity is not due to the Maker’s act of making, but this necessarily attaches to it once we have posited that quiddity [as existing]. But the Muʿtazilites’ claim that “the effect of the Agent does not consist in making objects to be objects” is not like this. For they make the objects non-existent, yet real in eternity, without the Agent having any effect (taʾthīr) [on them]. Having made objects equal in respect of being objects, they were obliged to affirm the attributes of the genera; otherwise everything would be of the same kind.
[T50] Al-Ṭūsī, Talkhīṣ al-Muḥaṣṣal, 77.13–18
[defense of the Muʿtazilite view against al-Rāzī]
[Al-Rāzī] said:24 [the Muʿtazilites are wrong to claim that non-existents are real] also because then objects would be eternal, and not subject to [God’s] power (maqdūra). And on their view, existence is a state (ḥāl), so for them it would not be subject to [His] power either. But if neither the object nor existence depend on the Agent, then existent objects will be independent of the Agent.
I say: they could respond that making an object to be qualified by existence is something over and above either of the two [sc. the object, and existence], like the composition that is additional to the parts. This [joining of existence to the object] is due to the Agent, and it does not follow from the independence of the parts that the composite is independent of Him.
[T51] Al-Ṭūsī, Fī al-nafī wa-al-ithbāt, 22.8–16
[does God add existence to already real objects?]
Those who affirm the reality [of the non-existent] say: it is existence’s being attributed to the object (mawṣūf-i budan-i dhāt bi-wujūd) that is through the Agent. Those who reject it ask: is existence’s being attributed to an object something that is known, or not? (a) If it is something known, then this is either (a1) the object itself or (a2) something else. (a1) If it is the object, then since the object is real while not existing, existence must be attributed to it. (a2) If it is something other than the object, then existence’s being attributed to the object is some further object of knowledge over and above the object. Whatever is an object of knowledge is something, and is real. However, there was [at first] no “being attributed,” and then it came to be (padīd āmad). So it follows that it was not something, and then it came to be. But this is not their doctrine! (b) Still, if its being25 attributed is not known, then neither is the act of the Agent. So the act26 would not be known, and from this it would follow that the Agent does not know what He does. This is not [their] doctrine either.
[T52] Al-Ṭūsī, Fī al-nafī wa-al-ithbāt, 21.10–14
[change in God’s knowledge]
Those who affirm the reality [of the non-existent] say: if substance was not real in eternity, God the exalted did not know it as real; so when it came to be known as real, His knowledge has changed. Those who deny it say: a similar consequence follows for you. For God the exalted did not know substance as existent in eternity. Then, He knew it as existent. But in any case, this alteration is in what is known, and not in God Himself, may He be exalted!
[T53] Al-Ṭūsī, Ajwibat al-masāʾil al-rūmiyya, 1.6–8; 6.10–15
[essential independence]
Does a cause produce the existence of an effect, or its quiddity?
Response: the cause produces the effect itself. Then, the intellect distinguishes between the existence and the quiddity of [the effect]. There are no existence and quiddity here such that the cause would produce either of them, or both. […]
[6.10] Is the Necessary Existent the cause of the existence of existing things, or of both their existence and their quiddity? […]
[6.12] The response to this has been already provided in the first question: namely that He is the cause of existents27 other than Him, and the intellect distinguishes, in its conceptual consideration of every existing thing, between its existence and its quiddity. Some of them said that existence is through a maker, whereas quiddity is not made. But those who make this and similar claims are those who have not achieved verification in matters of intellectual reflection (taḥqīq fī al-maʿqūlāt).
[T54] Al-Ṭūsī, Murāsala bayna al-Ṭūsī wa-al-Qūnawī, 104.1–9
[essences are made]
I say: when they say “quiddities are not made” they mean that, for instance, blackness is not blackness by someone’s making it so. For if we first suppose blackness, and then introduce a maker’s act of making for it, the maker cannot make it different from what supposed in the first place. Likewise with existence: no maker has made existence to be existence, since it is impossible to make something occur that has already occurred (taḥṣīl al-ḥāṣil). But if we were to ask whether any maker can make blackness to be blackness, in the sense of asking whether [the maker] can produce something, namely blackness, so as to ask whether [the maker] can make blackness to be existent, then one would rightly respond: indeed, [the maker] can produce blackness, and make blackness to be existent. In fact, the truth is that all quiddities and existent things are made, and their maker is God the praised and exalted.
[T55] Al-Ṭūsī, Murāsala bayna al-Ṭūsī wa-al-Qūnawī, 57.3–6; 107.2–4
[the Sufi position, with reply]
What has been yielded by verified witnessing and veritable “taste” is that quiddities are not made, and have existence of a kind. It belongs to them insofar as they are concrete (taʿayyunihā) in the knowledge of the True, eternally and forever, in one and the same way. This however is due to the connection (taʿalluq) of knowledge to them and the intellection of the numbering of connections in accordance with the objects of knowledge, with both the connection and the numbering being eternal. […]
[107.2] I say: claiming that quiddities are not made, and that they have existence of a kind, is close to what the Muʿtazilites affirm. For they say that [quiddity] is real while it is non-existent, and they distinguish between reality and existence.
[T56] Ibn Kammūna, Sharḥ al-Talwīḥāt, vol. 3, 17.10–19.11
[Is the existence of what will come to exist itself previously existent?]
The first way [to argue that non-existents are not real]: is the existence of the non-existent quiddity, which you claim to be real in non-existence, itself real in non-existence, or not?
(a) If it is not real, then the [quiddity] is not existent in non-existence, since its existence is obviously denied and ruled out, given that we are talking now on the supposition that it is not real.28 Now, if this existence [of the quiddity] is denied, it’s obvious that it cannot be necessary. Nor can it be impossible, because otherwise the quiddity would never come to exist at all, so that what exists contingently would be impossible of existence, which is a contradiction. So the [existence of the quiddity] is contingent, and is negated, so that something negated is contingent, and vice-versa. This falsifies their claim that the negated is equivalent to the impossible and that everything contingent is real.
(b) If however the existence of that quiddity is real in non-existence, then, given that according to their teaching, every contingent non-existent is real, and furthermore the existence that is still non-existent belongs to the class of the contingent, it follows that the non-existent is existent, which is absurd. The reason this follows is that, so long as the quiddity is real in non-existence, and the existence that is specific to it is real in non-existence too, then if [the quiddity’s] being attributed with that existence was not real, then, given that this [18] attribution is contingent, the same absurdity results. For we know for sure that if an attribute is affirmed as real for something, then this thing cannot help but have it as an attribute, since its having it as an attribute means precisely that it is affirmed as real (thubūt) for it. If, however, it is affirmed as real, then quiddity, existence, and the attribution of existence to quiddity are all real. Then, the quiddity would be attributed with existence even while it is in non-existence, and it would be existent while it is non-existent, the impossibility of which is obvious.
[individuation argument]
The second way: the concrete individual, which is indicated as being “this,” must prior to its existence either be real—and then it would already be this particular individual, since it is real, without yet being existent—or it would not be like this before its existence.
(a) If it is [already this individual before existing], and (a1) if its existence, or the existence of its attributes, was not at that point required in order for it to be indicated as a particular, then it must have been particular while it was non-existent. Yet every particular is existent—this judgment is simply obvious. So a non-existent individual would be existent, which is a contradiction. (a2) If however it does require its existence, or the existence of its attributes, in order to be indicated as particular, then also, according to their doctrine, its indication as a particular has to be possible even while it is non-existent. So it would be existent, despite having been posited as non-existent. For the quiddity, its existence, and the existence of its attributes—all of them are contingent. So on their view, these would be real. Meanwhile these existences could have no further existences, since otherwise [the second-order existences] would be also contingent, and thus real, and one can repeat this argument, yielding an infinite regress, which is absurd. Therefore, the Agent would have no effect (taʾthīr), neither on quiddity, nor on its existence, nor on the quiddities of its attributes, nor on the existence of that attributes. This—apart from being a denial of the Creator, the exalted—implies that the features that enable one to indicate this individual as a particular would be all present while it is non-existent, so that it would already be a particular in that state. But conceptualization is possible for this only while it is existent; so that it would be existent before existence.
Even if they renounce their doctrine that existence is additional to quiddity outside the mind, and agree that it is identical to the quiddity, it won’t help them. For if the quiddity is real, and it is identical to its existence, then its existence is real. So, [the absurd] conclusion follows once again.
(b) All this applies if [they accept that] it is already this [individual] before existing. If however it is not like this before its existence, then if [19] it is not real at all, then it verifies the falsehood of [the statement] that everything contingent is real. If however [this individual] is real but it is not this one while being non-existent, then this one, insofar as it is this one, was not contingent before existence. Rather the quiddity of the individual was contingent—not taken as the quiddity of that individual with respect to what that concrete individual is, but taken as an unqualified quiddity (al-māhiyya al-muṭlaqa). However, we are not talking here about the unqualified [quiddity]. We are talking about the [quiddity] which is specified as this one. For it were contingent insofar as it is this one, then it would be real as this one, although the working assumption has been that it is not real insofar as it is this one. Thus, it would not be contingent insofar as it is this one. But if it is not contingent, then it is either necessary or impossible. Both options are wrong: if it were necessary, then it would not be non-existent, and if it were impossible then it would never come to exist.
As both option (a) and option (b) are wrong, if one assumes that everything contingent is real while being non-existent, then [this assumption] must be wrong. This is the conclusion sought.
[T57] Ibn Kammūna, Sharḥ al-Talwīḥāt, vol. 3, 16.15–17.7
[distinction argument and reply]
The first way [to argue that the non-existent is real]: the non-existent can be distinguished, and whatever can be distinguished is real. So the non-existent is real. The rationale for saying “the non-existent can be distinguished” is that the non-existent, for instance tomorrow’s sunrise, is an object of knowledge, and whatever is known can be distinguished. For we distinguish between the sunrise in the East and in the West, and also between the motion we can perform and the one we cannot. Also, because we may want something, and want to avoid something else, even while what we want, and want to avoid, are non-existent. The rationale for saying “whatever is distinguished is real” is that, by “real,” we mean simply that quiddities are in themselves concrete and realized, given that one could not distinguish this [quiddity] from that one if they were not yet realized. So quiddities are real while they are non-existent. […]
[17.6] Refutation of their claim that “whatever is known is real”: we can conceptualize what is impossible, like the partner of God the exalted, a sea of mercury, or Yāqūt Mountain, yet [the opponents] agree that these are not real.
[T58] Ibn Kammūna, al-Jadīd fī al-ḥikma, 85.5–86.2
[conceptualization of absolute non-existence]
No informative statement can be made about the absolutely non-existent, which is that which has no form in the mind or extramentally. But absolute non-existence [as such] does have a form in the intellect. In the intellect, one may predicate of it that it is opposed to both mental and extramental existence. This does not imply that opposites are true of one and the same thing, since there is no contradiction between absolute non-existence and existence in the mind. For what is true is not “something is either absolute non-existence or existent in the mind,” but rather “something is either absolute non-existence or not absolute non-existence,” and “something is either existent in the mind or not existent in the mind.” So the notion of absolute non-existence is represented in the mind, and it becomes an individual form, with individual mental existence occurring to that form. The [mere] mental reality that is related to no extramental reality eliminates extramental reality. The fact that [absolute non-existence] is conceptualized in the mind and distinguished from everything else, and is concrete in itself and real in the mind, does not rule out its being something that is related to nothing extramentally real. For that which is extramentally unreal is not predicated to be entirely unconceptualized. Rather it is predicated to be conceptualized insofar as it is not real in extramental reality and not to be conceptualized except insofar as it has this description. But the elimination of reality, covering both the extramental and the mental, is a conceptualization of that which is unreal and is not [86] conceptualized at all. Thus one can validly make predications of it, insofar as it is this conceptualization, but not insofar as it is unreal. There is no contradiction, due to the different subjects [of predication].
[T59] Bar Hebraeus, Ḥēwath ḥekhmthā, Met., 126.12–16
[mental existence solution]
The non-existent is not anything (lā īth law meddem hū). Yet some people apparently argued against this statement. They say: whatever is grasped intellectually, regardless whether it is possible or not, has an intelligible form (yuqnā). Otherwise we could not predicate anything of it, or deny anything of it. Yet everything that has an intelligible form is something (meddem). Therefore the non-existent is something too. We respond to them: as the intellectual form is something in the intellect, so it has thing-ness (meddemyūthā) and not non-existence in the intellect. Therefore being grasped intellectually and actual thing-ness are co-extensive with actual existence (īthūthā).
[T60] Al-Nasafī, Sharḥ Asās al-kiyāsa, 273.17–274.5
[there is no distinction for the impossible]
The second argument [for the reality of the non-existent] is that the contingent non-existent may be distinguished from the impossible. And impossibility cannot be positive (thubūtiyyan), so contingency is positive. […]
[274.3] Regarding the second [argument]: this is quite impossible. For distinguishing between two things requires conceptualizing both of them. Here this is quite impossible, since the impossible is something that cannot be conceptualized at all. This is why one says that it is among the absurdities, as it is neither established nor conceptualized in the mind.
[T61] Al-Samarqandī, Ṣaḥāʾif, 84.1–13
[the affirmation principle and absolute non-existence]
First argument [against proving mental existence using the affirmation principle]: we predicate of the absolutely non-existent (that is, the unrestricted (bi-lā sharṭ) non-existent) that it is opposed to the existent, even though it has no reality at all.
Response: regarding “opposed,” “distinct,” and “negated,” the gist of the meaning of all these is “what does not agree.” One can predicate this of the non-existent, because one can predicate the non-existent of the non-existent. But we are speaking only about what is existing. If you have in mind some other meaning [for “opposed,” etc.] then we will not concede the correctness of the predication.
Second argument: If what you have mentioned were true, namely that predications can be made only of that which somehow exists, then it would be correct to say that one cannot make any predication of the absolute non-existent. But the consequent is false. For, (a) if the subject of predication in this case is really the absolutely non-existent, a contradiction would follow.29 (b) If on the other hand it is not [the absolutely non-existent], then predication is possible, so [what you have mentioned] would be wrong.
Response: as for (a), we say that one cannot predicate anything existing of the non-existent, here however this has not been done.30 As for (b): the meaning of the absolutely non-existent prohibits any predication of it only insofar as it is non-existent (mā dāma maʿdūman).31
[T62] Al-Samarqandī, Ṣaḥāʾif, 87.5–8
[three ways of talking about the non-existent]
We have two ways of arguing that the absolutely non-existent is purely negative while it is non-existent. The first is that, if it were true for non-existent blackness that it is blackness, then this proposition would be true either (a) extramentally (khārijiyya), but this is ruled out, because the subject is non-existent; or (b) in virtue of the true reality (ḥaqīqiyya), but in that case we would be taking it as existing (ʿalā taqdīr al-wujūd), and not while it is non-existent; (c) or mentally (dhihniyya). In that case it would be blackness in the mind, on which there is no disagreement.32
[T63] Al-Samarqandī, Ṣaḥāʾif, 90.5–13
[essential dependence]
Third argument [that the non-existent is real]: if blackness were blackness only when it existed, then its being blackness would be due to another. So if this “other” were eliminated, then the existent blackness would stop being blackness, which is absurd. This is the reason given by the philosophers (al-falāsifa) for the claim that quiddities are neither made nor created.
Response: we do not concede that in this case [blackness] would remain existent, so that the absurdity would follow.
Some claimed that “thing-ness” applies only to extramental existence. But they apparently based this upon their denial of intellectual existence, which we have already shown to be wrong. Also, their claim undermines the saying of the exalted: “His command is that when He wills a thing (arāda shayʾan) He need only say to it, ‘be!’ and it is,” (Qurʾān 36:82) where the non-existent is called “thing.”
[T64] Al-Samarqandī, Ṣaḥāʾif, 89.1–4
[doubts raised about al-Rāzī’s specificity and commonness argument at T38]
[Al-Rāzī’s argument] calls for further inquiry. For if he means by “non-existent” the contingently non-existent, then the division is wrong. For these two, the non-existence and the negation, are in fact mutually exclusive (mutanāfiyīn). If however he means the absolutely non-existent [i.e. including both the contingent and the impossible], then we do not concede that if it were common and negative, then there would be no difference between the specific and the common anymore. Rather there would be a difference between what one can truly say of the negated and of the contingently non-existent.33
[T65] Al-Ḥillī, Kashf al-murād, 17.12–18
[response to the distinction argument]
Response: the answer is that distinguishability does not imply concrete reality (al-thubūt ʿaynan), otherwise absurdities would follow. The first of these is that sometimes, the known object is in itself impossible of existence, like the companion of God the exalted, an agreement between contradictories, etc. One of these may be distinguished from another. If distinguishability implied concrete reality, there would follow the reality of the impossible, even though they agreed with us on negating the reality (intifāʾ) of the impossible. Second, the known object is sometimes a composite of something imaginary and something existent, and it is agreed that no such thing is real in concrete being. Third, if having power to do things implied that they were real, then they would be negative, because there is no power to do what is [already] real. And likewise for the things that we wish.
[T66] Al-Ḥillī, Kashf al-murād, 63.23–64.14
[God’s proper effect is bestowal of existence, not necessity]
[Al-Ṭūsī] said: it is upon quiddity that [God] has an effect, with necessity as a concomitant (lāḥiq).
I say: this is a response to their third question, the gist of which is as follows. That which has an effect (al-muʾaththir) does so either [64] upon the quiddity, or upon existence, or upon the attribution of existence to the quiddity. But all three options are false, so [His] having an effect is false. The first, because everything which [depends] on something else is eliminated when that other is eliminated. But this is absurd, because necessarily, it is absurd for the quiddity to be other than the quiddity. For the subject of predication has to be realized while its predicate is real; and the quiddity is not realized while non-existence is predicated of it. […]
[64.9] The gist of the response: God does indeed have an effect upon the quiddity. When quiddity is posited (ʿinda farḍ), its realization necessarily follows as a concomitant, on account of its having been posited and as a consequence thereof. Still, the necessity cannot be an effect [of God]. For [the effect] is the bestowing of existence, whenever He posits [the quiddity] as existent. As for the situation before He posits the quiddity [as existent], He is by necessity able to bestow existence upon the effect, and this kind of necessity is prior to the existence [of the effect]. The difference between two kinds of necessity is obvious, and was mentioned in the logic. The mistake here arises due to equivocation on “necessity” which can mean two different things.
[T67] Al-Ḥillī, Asrār, 415.18–416.11
[absolute non-existence]
Thing-ness cannot be without one of the two kinds of existence. Someone who says that things can be absolutely non-existent is saying something true, so long as they mean by this that it is non-existent in concrete individuals, since the thing can be real in the mind and non-existent in extramental reality. But if they mean that it is non-existent both in the mind and in extramental reality, this is impossible, since the absolute non-existent [416] cannot be the subject of an informative affirmation or negation. For “this” involves an indication, but there is no way to indicate absolute non-existence, nor any mental form for it. How can anything be affirmed of the non-existent, given that affirmation would mean such-and-such a description occurring for the non-existent, and the occurring of one thing for another presupposes that [the latter] thing itself occurs? Our knowing the non-existent means that it has a form in the soul, but one through which [the soul] does not indicate anything in extramental reality. When we make an informative statement about the non-existent, we make the statement about such a form.
Some people affirmed that the non-existent is extramentally real, because they did not understand [the idea] of mental existence. For they made an informative statement [about the non-existent], and did not realize that the informative statements are only about something real in the mind, and then by means of this, about something that is real in concrete individuals—if it has any such reality. They also said that the non-existent is an object of knowledge and intention and can be distinguished, and everything that can be distinguished is real. Their argument is correct, but they did not grasp the difference between the absolutely real and the mentally real.
[T68] Al-Ḥillī, Kashf al-murād, 51.6–52.9
[conception of contradictories and non-existence]
[Al-Ṭūsī] said: the intellect can consider opposites and predicate opposition between them, which involves no absurdity.
I say: the intellect can predicate opposition between negation and affirmation, so it has to consider them at the same time. For opposition belongs to the class of relations and associations, and can be conceptualized only once one has conceptualized that to which it applies. So [the intellect] conceptualizes negation and affirmation at the same time; but there is no absurdity in their simultaneous co-occurrence. For opposition is not in relation to the mind but in relation to things in themselves (nafs al-amr). So one may conceptualize a form in the mind, and predicate of it that there is nothing extramental that corresponds to it. Then one may conceptualize another form and predicate of it that there is indeed something extramental that corresponds to it. Then one may predicate of one of them that it is opposite to the other; not insofar as both of them are present in the intellect, but insofar as one of them goes back to the extramental reality, while the other doesn’t. Indeed, the mind may conceptualize a form and its negation, because it distinguishes [between them] in the way explained above, and predicates opposition of the two forms [i.e. the form and the form of its negation] not insofar as they are present in the mind, but in the way that has been mentioned.
[Al-Ṭūsī] said: the non-existence of everything is conceivable, even its own non-existence and the non-existence of non-existence, by way of imagining it in the mind and eliminating it. It is in a sense real, and in a sense privation (qasīm). But predications do not truly apply to it insofar as it is unreal, so there is no contradiction.
I say: the mind can conceptualize all intelligibles, regardless whether they are existing or non-existing. It can apprehend the non-existence of everything, because it conceptualizes unqualified non-existence and can relate it to all quiddities. [The mind] can even apprehend [non-existence] as applied to itself, [52] so that the mind conceptualizes the non-existence of itself. In the same way, it can apprehend non-existence itself, meaning that the mind represents a form for non-existence, which is intelligible as being distinct from the form of existence, and it conceptualizes the elimination of [that form]. [Non-existence] is real insofar as it is conceptualized. For the elimination of reality, covering both extramental and mental, is conceptualized for that which is neither real nor conceptualized at all. [Non-existence] is real when it is taken as conceptualized and is the privation of the absolutely real, taken as its negation. There is nothing implausible here. For we say that the existent is either real in the mind, or not. So non-existence is the privation of existence, but when taken as a notion (mafhūm), it belongs to the real. The predication applies to the elimination of absolute reality [sc. non-existence] insofar as it is conceptualized, not insofar as it is unreal. There is no contradiction, because of the different subjects [of predication].
R.M. Frank, “Al-maʿdūm wa-l-mawjūd: The non-existent, the existent and the possible in the teaching of Abū Hāshim and his followers,” MIDEO 14 (1980), 185–209; F. Klein-Franke, “The Non-Existent is a Thing,” Le Muséon 107 (1994), 375–390. For the reception of the doctrine in our period see F. Benevich, “The Reality of the Non-Existent Object of Thought: the Possible, the Impossible, and Mental Existence in Islamic Philosophy (Eleventh–Thirteenth Centuries),” Oxford Studies in Medieval Philosophy 6 (2018), 31–61.
This passage also solicited attention from philosophers, perhaps precisely because they were in dialogue with the Muʿtazilites. See P. Adamson, “Al-Kindī and the Muʿtazila: Divine Attributes, Creation and Freedom,” Arabic Sciences and Philosophy 13 (2003), 45–77.
Al-Rāzī, Mabāḥith, vol. 1, 132.13.
Avicenna, Shifāʾ, Ilāhiyyāt, I.5, 26.4.
R. Wisnovsky, “Notes on Avicenna’s Concept of Thingness (shayʾiyya),” Arabic Sciences and Philosophy 10 (2000), 181–221; T.-A. Druart, “Shayʾ or res as Concomitant of Being in Avicenna.” Documenti e Studi sulla Tradizione Filosofica Medievale 12 (2001), 125–142.
See Deborah Black, “Avicenna on the Ontological and Epistemic Status of Fictional Beings,” Documenti e studi sulla tradizione filosofica medievale 8 (1997), 425–453; D. Black, “Mental Existence in Thomas Aquinas and Avicenna,” Mediaeval Studies 61 (1999): 45–79; T.-A. Druart, “Avicennan Troubles: The Mysteries of the Heptagonal House and of the Phoenix,” Tópicos 42 (2012): 51–73.
See further M.S. Zarepour, “Avicenna on Empty Intentionality: A Case Study in Analytical Avicennism,” British Journal for the History of Philosophy, published online: DOI
See further the chapter on Platonic Forms below.
For a possible Avicennan inspiration for this move see Benevich, “The Reality of the Non-Existent Object of Thought.”
In these passages, Abū Rashīd defends one Muʿtazilite position that substance is a substance in non-existence against another Muʿtazilite position that it is only an unspecified shayʾ. The Ashʿarite and the later Muʿtazilite position that the non-existent is not even a shayʾ is not discussed by Abū Rashīd at all.
Al-Shahrastānī’s account is based on al-Anṣārī, Ghunya, vol. 1, 279–281 as well as al-Juwaynī, Shāmil, 124–126.
Adding lā before yakūna.
For the continuation of this passage see [13T11].
In Maṭālib, vol. 1, 94.18–95.7, al-Rāzī ascribes a similar argument involving the issue of essential independence to the Muʿtazilites, as a support for their maʿdūm-šayʾ theory. Al-Samarqandī, Ṣaḥāʾif, 87.1–4 and 90.5–8 offers the same kind of parallel.
This argument is accepted in al-Abharī, Kashf al-ḥaqāʾiq, 253.8–11; al-Ḥillī, Kashf al-murād, 17.13–16; and Ibn Kammūna, al-Jadīd fī al-ḥikma, 84.10–15.
See further our chapter on Platonic Forms [T8].
Cf. Mulakhkhaṣ, fol. 126v16–22 and Maṭālib, vol. 1, 49–50. This solution is accepted in al-Kashshī, Ḥadāʾiq al-ḥaqāʾiq, fol. 128v8–15.
cf. al-Rāzī, Mabāḥith, vol. 1, 136.6–14.
Reading al-dhātī for al-dhāt.
Reading bi-jaʿl for yajʿalu.
See al-Rāzī’s arguments against inherence of forms in the mind in the chapter on Knowledge and Cognition from the “Logic and Epistemology” volume.
Deleting aw fī al-dhihn following the MS Majlis Shurā-yi Millī 2752, 110.16.
By contrast, in Munaṣṣaṣ fī sharḥ al-Mulakhkhaṣ, fol. 96v39, al-Kātibī accepts al-Abharī’s counterargument from [T47] against this position.
The text is a commentary on al-Rāzī, Muḥaṣṣal, 56.3–4.
Correcting nabūdan-i to būdan-i.
Correcting fāʿil into fiʿl.
Reading al-mawjūdāt for al-wujūdāt.
For this and what follows it is helpful to bear in mind that “negated” is taken by the opponents as mutually exclusive with “real.”
That is, because you have to predicate of absolute non-existence that one cannot predicate anything of it.
As the editor indicates, the point is that predicating the impossibility of predicating is not predicating something existing.
Cf. al-Urmawī, Maṭāliʿ, fol. 3r16–22.
See further the chapter on Propositions in the Logic and Epistemology volume.
Al-Rāzī’s specificity and commonness argument is equally dismissed in al-Abharī, Kashf al-ḥaqāʾiq, 252.16–253.5.