Iʿtibārī Concepts in Suhrawardī

Shihāb al-Dīn al-Suhrawardī (d. 587/1191) famously criticised the central concepts of Avicennian metaphysics as merely mind-dependent (or iʿtibārī) notions. This paper aims to show that despite his critique, Suhrawardī held that these concepts are meaningful, indeed necessary for human cognition. By the same token, it is argued that their re-emergence in Suhrawardī’s ishrāqīmetaphysics is not amatter of incoherence. Although the paper’s findings can be generalised to hold of all iʿtibārī concepts,mutatis mutandis, our focus is on the concept of substance, mainly because of the importance of the concept of ‘dusky substance’ in ishrāqī metaphysics.

Oriens 48 (2020) 40-66 existence is a product of the human mind and its peculiar way of understanding quiddities.1 In this reconstruction, much hinges on how we understand the crucial notion of iʿtibār ʿaqlī or iʿtibār dhihnī. The difficulties of translating the term are well-known, and considerable uncertainty prevails as to what exactly Suhrawardī means by it. For example, iʿtibārī concepts have been described as "intellectual fictions,"2 "mental constructs,"3 "beings of reason" that "are products of our thought about things,"4 and "mere concepts,"5 which suggests that they have little if any basis in reality, and that their role in Suhrawardī's new ishrāqī philosophy is to be merely explained away as so many errors of previous generations. On the other hand, iʿtibārāt have been described as innate to the human mind, and even likened to Kant's transcendental categories,6 which suggests that they are concepts that we are bound to use in all cognition.
My aim in this paper is to investigate how Suhrawardī's critique of the central concepts of Peripatetic metaphysics as merely iʿtibārī relates to the ishrāqī alternative outlined in the second part of the Ḥikmat al-ishrāq. For this purpose, I will momentarily set aside the historically central discussion concerning the concepts of essence and existence, and focus instead on the concept of substance. This is for two reasons. First, the concept of substance makes a puzzling reappearance in an important passage early on in that central ishrāqī text. Secondly, denying the extramental reality of most of the other iʿtibārāt (existence, the modalities, genera, number, relation, and privation) seems intuitively less problematic than that of substance. For instance, the question of the reality of existence is a convoluted metaphysical problem, in which both alternatives are kaukua Oriens 48 (2020) 40-66 grounded in ulterior concerns, and no choice in favour of one or the other can be made solely based on their capacity to explain everyday phenomena. On the other hand, there are strong intuitive grounds for the mind-dependence of concepts like number: did my grandfather's right leg, the moss on my gravestone, and your mobile phone really constitute a set of three before we did the counting? But substance seems more difficult to reduce to the mind. If there are no substances, what remains of the robust things on the stability of which we base our functioning in and communication about the world? Alternatively, how can we conceive of change if there are no substances against the persistence of which we can measure that which changes?
In the following, I will begin with a brief derivative sketch of the sixth/ twelfth-century debate that provides the context for Suhrawardī's critique of the iʿtibārāt. I then turn to an equally concise review of Suhrawardī's standard arguments against the reality of the iʿtibārāt, with the special aim of showing that he recognises the problems that result from denying the reality of substance. I will then address his attempts at tackling those problems, which I argue show that Suhrawardī did not conceive of the iʿtibārāt as arbitrary mental constructs but rather as necessary entailments of the first-order concepts in which they are grounded.7 These considerations provide the basis for my analysis of the role the concept of substance has in the light metaphysics of the Ḥikmat al-ishrāq. In the end, I venture the claim that substance, and perhaps other iʿtibārāt as well, is indeed more like a transcendental concept-that is, a condition of possibility for the awareness of a certain kind of objectsthan a conceptual fiction. As a consequence, the focus of this paper is less on the critique of the iʿtibārāt, about the provenance of which in the twelfthcentury reception of Avicenna we now have excellent recent studies,8 and more on making better sense of Suhrawardī's alternative to Peripatetic metaphysics. 7 Thus, when I speak of Suhrawardī's critique of the iʿtibārāt, I mean his critique of the assumption that they have a reality independent of the mind. Suhrawardī does not mean that we should refrain from using them. Similarly, when I speak of denying the reality of the iʿtibārāt, I mean this mind-independent reality. Obviously, iʿtibārāt are real insofar as they are concepts in real minds. 8 For instance, Wisnovsky, "Essence and Existence in the Eleventh-and Twelfth-Century Islamic East," and Benevich, "The Essence-Existence Distinction." Oriens 48 (2020) 40-66

The Sixth/Twelfth-Century Background
The context for Suhrawardī's discussion of the iʿtibārāt is the sixth/twelfthcentury debate launched by the Avicennian distinction between essence and existence. As is well known, the distinction is crucial for Avicenna's proof for God's existence. Given the distinction,9 Avicenna states that for every contingent thing, the fact that its essence exists is due to a cause that is distinct from the essence, because the essence alone does not entail its own existence. If the cause of its existence were another contingent thing, we would have to pursue the question further. And since an infinite regress of causes amounts to ungrounded existence, Avicenna concludes that there must be an essence that exists by intrinsic necessity, or an essence that entails its own existence, and this is God.10 The distinction is also crucial for Avicenna's famous tripartite division of the status of essences, namely considered in themselves, in concrete, and in the mind.11 If the same essence can figure in all these considerations, the essence in itself must be neutral with respect to whether it exists in concrete or in the mind. This in turn suggests that there is a robust distinction between the essence and its existence. But whether the distinction is real or valid but merely conceptual was the precise bone of contention in the twelfth-century debate.12 A real distinction between essence and existence, in the sense of 9 For which Avicenna provides an argument. It seems that the debate arose when some post-Avicennian Ashʿarite theologians, most importantly ʿAbd al-Malik ibn ʿAbdillāh al-Juwaynī (d. 478/1085), adopted the initially Muʿtazilite idea of aḥwāl (sg. hāl) in order to deal with the problems concerning God's attributes.14 The aḥwāl are adverbial modes of being that belong to real entities without being entities in their own right. Since they are not entities, the theologians believed that they are not subject to the principle of the excluded middle, and that they can legitimately be said to neither exist nor not exist.15 Such an intermediate class was useful in solving problems concerning God's attributes. For instance, the theologians held that the attribute of eternality (qidam) is predicable not only of God but also of all His other attributes, including itself. However, if attributes are entities, this gives rise to an infinitely regressive series: eternality is eternal by virtue of a second-order eternality, which must also be eternal, and so forth. In order to cut the regress, Juwaynī held that only God and His first-order attributes are entities, whereas having an attribute is a mere ḥāl that lacks a metaphysical status of its own.16 By the turn of the fifth/eleventh century, the notion of ḥāl was conjoined to the Avicennian distinction between essence and existence as well as the debate concerning universals, and thereby the initially theological concept became a general metaphysical notion.17 A generation or so after Juwaynī such a general 13 Wisnovsky, "Essence and Existence in the Eleventh-and Twelfth-Century Islamic East," 29-30; cf. Benevich, "The Essence-Existence Distinction," 208, 237-42. 14 For the steps in this process, see Robert Wisnovsky, "One Aspect of the Avicennian Turn in Sunnī Theology," Arabic Sciences and Philosophy 14 (2004) theory of aḥwāl was criticised, first it seems by ʿUmar Khayyām (d. 517/1123), and then by Muḥammad ibn ʿAbd al-Karīm al-Shahrastānī (d. 548/1153), and it is to this strand that Suhrawardī's critique belongs. A crucial feature of the criticised version of the theory of aḥwāl is the idea that the truth of our statements about the world is always based on an adequate correspondence between the terms of our statements and metaphysically distinct elements of reality, whether existing substances and accidents or aḥwāl.18 Both Khayyām and Shahrastānī replaced the aḥwāl with iʿtibārāt existing in the mind, already relying on many of the arguments that Suhrawardī would later apply.

The Argument against Substantiality
The concept of substance that Suhrawardī sets out to criticise is rather uncontroversial. In his paraphrase, substance is something that exists externally so that it does not subsist in another.19 He then appends this definition with a threefold distinction between different kinds of substances-the corporeal instantiations of infimae species (such as the favourite example of a horse), the metaphysically constitutive parts of these instantiations (that is, prime matter and form), and incorporeal substances (such as human and celestial intellects).20 This familiarity notwithstanding, the reader of the Ḥikmat al-ishrāq soon finds that the definition of substance is only presented in order to refute the claim that substance can provide the foundation for the metaphysical analysis of being. The central argument Suhrawardī employs for this purpose is that the unipress & Bonn University Press, 2018), 323-50; and Wisnovsky, "Essence and Existence in the Eleventh-and Twelfth-Century Islamic East," 37-40. 18 Benevich, "The Classical Ashʿari Theory of aḥwāl," 154-8; "The Metaphysics of Muḥammad b. ʿAbd al-Karīm al-Šahrastānī," 331-40.

19
ḤI I.3.3.52, 42-3; cf. III.2.1, 232-3. He adds that substances do not subsist in another by being "completely diffused" (ʿalā sabīli l-shuyūʿ bi-l-kullīya) in it. This qualification is made in order to distinguish the formal and material parts of concrete things from their accidents, which are in their subjects by being "completely diffused in them." Form and matter can also be said to be in the whole of which they are parts, but not in this manner of complete diffusion. 20 Suhrawardī, al-Mashāriʿ wa-l-muṭāraḥāt supposition that substances are real, or that substantiality is a real constituent of concrete beings, leads to infinite regress.
Know that substantiality is not anything added to corporeality in concrete […]. Instead, making ( jaʿl) something a body is identical to making it a substance, since according to us substantiality is nothing but the perfection of the thing's quiddity in the respect that it is independent of a substrate in its subsistence. The Peripatetics have characterised it (ʿarrafūhu) as not existing in a subject, but the denial of subject is negative whereas existing (al-mawjūdiyya) is accidental. If their defender says that substantiality is another existing thing, it will be difficult for him to explain and establish this to the opponent. If it were another thing existing in the body, it would have existence not in a subject, and so it would be attributed with substantiality, and the argument would return to the substantiality of substantiality, regressing infinitely.21 There are two phases in this version of the argument. At first, Suhrawardī denies that substantiality is a real constituent of concrete things by questioning the validity of its definition. Substance is defined by appealing to existence ('what exists not in another'), and since existence both is extraneous (or "accidental") to the essence of substance and has already been argued to be a mere iʿtibār in earlier paragraphs of the same section,22 substance will be equally unreal as its definiens. Moreover, the definition employs a negative attribute as a differentia ('what exists not in another'), and negative features are by necessity dependent on the mind that performs the negation. This leads to the dense second phase of the argument. According to Shams al-Dīn al- Shahrazūrī (d. after 687/1288) substance by compromising the epistemic status of the definition.23 Since substance is a highest genus, it cannot be defined by means of a more general concept, nor can it be attributed with a proper differentia. Hence, all attempts at defining substance have only pragmatic validity, and the best one can do is characterise it in a way that elucidates the concept to one's interlocutor and thereby makes the positive reality of substance easier to perceive. Substantiality will still be a metaphysically real constituent of individual substances, albeit one that is indefinable. Shahrazūrī dismisses this defence as an instance of what he calls the Peripatetic custom of "postulating [allegedly] known realities for the many unknown consequences of [the Peripatetics' own] statements".24 The question of definability constitutes the background for the real argument. According to Suhrawardī's reconstruction of the Peripatetic theory, substantiality is the feature due to which a concrete substance (such as a body) is a substance, or something that is not in a subject. Being a feature of a body cannot be constitutive to substantiality, for if substantiality were essentially in a body, or in anything else for that matter, substantiality would be an accident by definition, and the body a substance by accident. Since this is absurd, substantiality must be a substance in its own right, which leads to the infinite regress: if substantiality itself is a substance, it must be a substance because it has substantiality in a second order, and so forth ad infinitum. The natural conclusion is that substantiality is a purely iʿtibārī notion, existing only in and dependent on the mind.
From an Avicennian point of view, the argument is problematic because it makes the highly questionable assumption that constitutive features of essences must be entities in their own right. Suhrawardī all but ignores the possibility that substantiality is constitutive to the body in the sense that it is universally entailed by all bodies in the same way as a species entails its genus-an interpretation that comes peculiarly close to Suhrawardī's own view.25 It thus seems that instead of Avicenna himself, his immediate target was the aforementioned amalgamation of Avicennian metaphysics and the theory of aḥwāl. Interestingly, however, Suhrawardī remains silent about the central bone of contention of the earlier opponents of that theory, namely that the aḥwāl violate the law of the excluded middle,26 which suggests that he represents a later, 23 Shahrazūrī settled stage in the development of the theory of iʿtibārāt. But the earlier critics of the aḥwāl also recognised that the theory was introduced to solve pressing metaphysical problems, such as the problem of universals. These motives survive in Suhrawardī's discussion of what is perhaps the most potent counterargument to the claim that substance is an iʿtibārī concept.

The Truth of iʿtibārī Statements
In al-Mashāriʿ wa-l-muṭāraḥāt Suhrawardī introduces three arguments in defence of the real existence of the iʿtibārāt. According to the first argument, our distinction between real and merely imagined existence is inexplicable without a corresponding real distinction between something's having and not having existence.27 The second argument is, interestingly, based on the law of the excluded middle. The iʿtibārāt come in exhaustive pairs or sets-for instance, if there is no possibility, then the thing that does not have possibility must be either necessary or impossible. But were that the case, there would be no contingent things, which has devastating consequences for the Avicennian distinction between God and the created world by means of the modalities.28 While these two arguments may be relatively easy to deal with by means of an iʿtibārī interpretation of the relevant concepts,29 the third argument is more interesting, for it poses the question of how to explain the fact that for any given object, some iʿtibārī concepts are evidently valid but others just as evidently invalid.
They have said: If these things were intellectual predicates and not among the affairs that have reality (lā umūran fī dhawāti l-ḥaqāʾiq), the mind could connect them to any chance quiddity. Thus, anything that the mind associates with existing in concrete would have come to exist in concrete.30 is in a context distinct from his discussion of the iʿtibārāt; see, for instance, T III, muqaddima, 2, 175-6. 27 MM III.3.6.102, 344.11-5; the argument is already presented by Juwaynī (Benevich, "The Classical Ashʿari Theory of aḥwāl," 156-7 Although the concept of substance is not specifically mentioned here, it is clear that the argument concerns the iʿtibārāt in general, including substantiality. Hence, if substantiality were merely iʿtibārī, we could classify percepts as substances arbitrarily. For instance, it would not be the case that I perceive the apple tree as a substance because it is an instantiation of a maliform species that endures the variation of its accidental attributes. Instead, I could just as well conceive it as a substance because it is of a certain height, because it is about to bloom, or because it happens to stand three meters from my window. Although each of these arbitrary applications of the concept of substance overlaps with the same conglomerate of features, the things that they pick out are shown to be distinct as soon as the tree proceeds to bear fruit or I prune its top branches. In general terms, if the distinction between substances and their accidental attributes were arbitrary, the ways in which individual cognitive subjects carve out the world by means of their concepts would be entirely subjective, and we would have no means to decide in favour of one over the other.
Such an extreme form of nominalism is not a conclusion Suhrawardī is prepared to draw: As regards the third argument, that is, their saying that "if these were mental, the mind could connect them to any chance quiddity, and they would be true of it," it is false. It is not a condition for a mental affair to have an equal relation to all quiddities. Is something's being particular not a mental affair? [Yet] we are not to connect it to any quiddity we want, but [only] to certain quiddities of which it is true because of a specificity of theirs (khuṣūṣihā), and likewise for being a genus, being a species (al-nawʿiyya), impossibility, and what is like that. The mind connects iʿtibārāt only to what it observes them to be suitable for because of a specificity in the quiddities (bimā yulāḥiẓu ṣulūḥahā lahu li-khuṣūṣin li-l-māhīyāt).31 Suhrawardī's stance is clear, albeit not particularly persuasive: each extramental object has a specificity that determines which iʿtibārāt are suitably attributed to it. He adds that the mind somehow observes the suitability between an iʿtibārī concept and an object due to this specificity, but it is difficult to see how reference to something so vague could be a real answer to a substantial problem. Its vagueness notwithstanding,32 the same point is made in the Ḥikmat al-ishrāq, with one further qualification: 31 MM III.3.6.103, 347. 32 It is perhaps worth noting that aware of this shortcoming, Mullā Ṣadrā makes an extended kaukua Oriens 48 (2020) 40-66 It is not the case that if something is a mental predicate, such as the generality predicated of something for instance, then we can connect it in the intellect to any chance quiddity so that it is true. Rather, [we can only connect it] to what is suitable for it due to its specificity (limā yaṣlaḥu lahu bi-khuṣūṣihi).33 The passage does not give us a clue about the nature of specificity either, but it does add one seemingly obvious but crucial point, namely that the relation of suitability concerns truth. Whatever the specificity in the extramental object may be, it is special because it makes the attribution of the iʿtibārī concept to that object true. This "truth-making" capacity cannot be explained by means of correspondence between the mind and the world, as Suhrawardī and his commentators emphasise.34 The most extended treatment of this problem is in an answer to the following question in al-Mashāriʿ wa-l-muṭāraḥāt: When we have analytically separated ( faṣṣalnāhumā) in the mind the general and the differential meaning of a simple reality, such as black, do both of them correspond ( yuṭābiqa) to the external black as such, or does one of them correspond to [one] thing and the other to another thing? If both correspond to the black as such, there is no difference in the intellect between the two or between either of the two and the form of black, for under this assumption the form of black corresponds to the external [black], as does the differentia on its own and the genus on its own.35 Suhrawardī's opponent argues that unless there is a real distinction between the general and the differential constituents of a perceived object, such as an instantiation of black, the distinction in the mind will collapse. genus 'colour,' and the differentia 'absorbs sight') all correspond to the same external object, and if the relation of correspondence is transitive and symmetrical (an implicit but necessary assumption to make the argument work), then the mental terms turn out to be indistinct. And when we introduce white to the picture, we get the further absurdity that black and white are indistinct; for if the concept 'colour' corresponds to both white and black, then given that correspondence is both transitive and symmetrical, the concept 'white' corresponds to black.36 Interestingly, Suhrawardī's answer denies that the correspondence theory of truth can be straightforwardly applied in all cases: What you said in the question concerning correspondence (al-muṭābaqa) is also valid, but not all that is predicated of something is predicated because of correspondence to a concrete form, for particularity is predicated of Zayd, as is the concept of reality insofar as it is reality, yet they are not two forms belonging to his essence or to any of his attributes. Instead, they are two attributes of his that only come to be in the mind. Now, correspondence does come into consideration in the case of attributes which have existence in the mind but also in concrete, such as black and white, but in the case of the reality of colourness the attribution is iʿtibārī ( fī lḥaqīqa al-lawniyya waṣf iʿtibārī), and likewise in the case of genera and differentiae. Hence, black is one reality, the existence of which in the soul is like its existence in concrete, and it has nothing essential in any respect, nor does it have any parts.37 Some mental content, for instance my conception of black, is true because it corresponds to an extramental object. Hence, Suhrawardī is a realist about at least some concepts. But regrettably, at least as far as I am aware, he never states decisively which concepts, and why, he takes to be real. However, in light of the scattered remarks he does make, it seems that he accepts all really instantiated infimae species.38 Their truth is a matter of correspondence between the mind and the world. 36 These absurdities are spelled out in the immediately following section (MM III.3.7.115, Be that as it may, straightforward or immediate correspondence is not the ground for the truth of many other types of mental content. For instance, logical concepts, such as this being a particular instantiation of black or having mind-independent reality, have no distinct counterparts corresponding to them in the extramental world, and the same is true of the abstracted generic and differential features of the object, like its being colour or absorbing vision. Yet it seems intuitively plausible that statements constituted by such content can be every bit as true of the extramental object as statements that are true due to correspondence. The burden of showing how that can be remains on the iʿtibārī theorist. Perhaps in a related attempt, slightly later on in the same chapter of the Mashāriʿ wa-l-muṭāraḥāt Suhrawardī states that simple concepts, such as 'black' and 'white,' that are true by way of correspondence are "principles" (mabādiʾ) for iʿtibārī concepts, like the genus concept 'colour.' Furthermore, he says that at least such iʿtibārī concepts that are fiṭrī, that is, those that we "innately" hold to be true of their respective "principle" concepts, "end at them."39 This ending at principles is naturally interpreted to mean the reference of iʿtibārī concepts, and thereby the truth condition of related propositions: such concepts primarily refer to the relevant first-order concepts, and by means of them to worldly objects. By the same token, propositions including iʿtibārī concepts are true by virtue of the truth of relevant first-order "truthmaker" propositions, which in turn are true because they correspond to actual states of affairs in the extramental world.40 In terms of contemporary metaphysics, we could say that the iʿtibārī concepts are grounded in those first-order concepts.
Such a general answer still leaves open a number of crucial questions. First of all, the examples of black and white are special in the sense that they are simple sense data, for which it is sensible to say that our knowledge of them as well as their differences is fiṭrī.41 But what about the more complicated, substantial infima species concepts, like 'horse' or 'human?' It seems that a great deal of learning is required for their acquisition, and that we can go wrong in Oriens 48 (2020) 40-66 the process. Secondly, is the application of all iʿtibārī concepts to their firstorder truthmaker concepts as unproblematic as the application of 'colour' to 'black' and 'white,' or is there room for doubt and error in at least some cases? I will postpone the first question to the next section of the paper, because it is immediately related to the question of the role of the concept of substance in Suhrawardī's ishrāqī metaphysics. Let us, however, conclude the present section with a brief reflection on the second question. Suhrawardī repeatedly states that at least the iʿtibārī concepts of existence and thingness are fiṭrī: as soon as I am aware of a black object, I am not only naturally aware that it is coloured, but also that it exists and that it is something.42 Yet unlike being coloured, existence and thingness are not constitutive to the essence of the black object in logical analysis. Hence, the grounding or truth-making relation between first-order concepts and iʿtibārī concepts must accommodate different kinds of relation. What is more, perhaps there are other iʿtibārī concepts that are not similarly obvious, such as the modalities or the concepts of substance and accident. Insofar as we can conceive of genuine debates about their applicability, it seems that some extent of investigation is required not only for acquiring the concepts in the first place, but also for justifying their application to any given first-order concept. Yet even if that were the case, the grounding and truthmaking relation I am offering as Suhrawardī's solution to the problem of the truth of iʿtibārī concepts could still hold on a general level-only in some cases the relation is obvious and immediate, whereas in others it may require justification and intermediate argumentative steps.
These more detailed questions will require further study that brings together different parts of Suhrawardī's metaphysical texts, instead of focusing only on his explicit treatment of the question of the iʿtibārāt. However, there seems to be no prima facie reason why such an analysis could not be carried out. By way of a more serious counterargument, someone might say that reconstructing the relation between iʿtibārī and first-order concepts in terms of grounding and truthmaking is anachronistic, because there is a more natural way of understanding it: if iʿtibārī concepts are true of first-order concepts that are true by way of correspondence, it seems that they naturally fall into the class that the Arabic logicians commonly called secondary intelligibles (maʿqūlāt thāniyya). Now, Avicenna restricts this term to specifically logical second-order 42 T I. concepts, such as 'universal,' 'particular,' 'genus,' 'species,' or 'differentia,'43 and with the exception of one passage, Suhrawardī seems to have followed suit.44 It is true, however, that after the sixth/twelfth-century, the iʿtibārāt are commonly spoken of as secondary intelligibles, and this eventually leads to a distinction between philosophical ( falsafī) and logical (manṭiqī) secondary intelligibles.45 Why should we simply not follow the tradition? I have no objection to calling the iʿtibārī concepts philosophical secondary intelligibles. The question is, how does this label help us understanding the underlying problem? One would still have to explain how the secondary intelligibles are related to the primary intelligibles, or first-order concepts, and how the philosophical secondary intelligibles differ from logical ones in this respect. In Fedor Benevich's words, "[o]ne might wonder how existence," for an example, "can be the subject-matter of metaphysics if it is a secondary intelligible."46 And yet there doesn't seem to be a single thinker in the later Islamic tradition willing to exclude the iʿtibārī concepts from metaphysical analysis. On the contrary, the later thinkers commonly agree that although the analysis that yields the distinct iʿtibārī concepts does depend on the mind, the validity of applying them to first-order concepts does not, and because of this their study does not boil down to mere conceptual analysis but constitutes genuine metaphysical investigation.
Thus, Suhrawardī's critique of iʿtibārāt is based on a metaphysical departure from a robustly realist view concerning all our scientifically valid concepts to a moderate realism that distinguishes between concepts that do correspond to distinct constituents of mind-independent reality and concepts that are grounded and made true by the former. If my interpretation is on the right track, the distinction comes with a theory of how the two kinds of concepts are metaphysically related to each other. This theory gives Suhrawardī the basis for  (MA thesis, McGill University, 1994). 46 Benevich, "The Essence-Existence Distinction," 220 (fn. 59).
Oriens 48 (2020) 40-66 a metaphysical analysis of conceptual relations that goes beyond the mere registering of intensional distinctions between extensionally identical concepts.47 For instance, it is true that the concept of substance is extensionally identical to the conjunction of all substantial infimae species concepts, and it is also true that the concept of substance is intensionally distinct from each infima species concept. But as Suhrawardī's analysis shows, this is not all we can say about their relation, for we also know that the infima species concept, unlike the concept of substance, has a distinct correspondent in the world, and that because of this difference the two concepts stand in a hierarchical relation to each other-'substance' is entailed by, or grounded in, the infima species concept. In the final analysis, it would be a mistake to take Suhrawardī's nuanced critique of the iʿtibārāt for their flat rejection. By the same token, we should not label Suhrawardī a nominalist or a conceptualist pure and simple,48 for although he is a conceptualist about some concepts (the iʿtibārāt), he is a realist about others (the infimae species) and has a theory of how the mind-independent concepts are grounded in the real ones. In the following, final section I try to argue that this is true also of the ishrāqī metaphysics he presents in the second part of the Ḥikmat al-ishrāq.

The Dusky Substance
An underlying objective in Suhrawardī's critique of substance and other iʿtibārī concepts may have been to clear conceptual room for an alternative to Peripatetic metaphysics. The foundation of the new metaphysics is laid in the second part of the Ḥikmat al-ishrāq, which begins with the axiomatic introduction of the concept of light (nūr) defined as ẓuhūr, that is, "manifestation" or "appearing." In this foundational sense, light or appearing cannot be explained, described, or defined by means of anything else49-instead, it provides the ground for the explanation of all other things. As a consequence, light is not the appearing of any further thing; in particular, it is not the appearing of any- thing like substance, but appearing as such, pure and simple. In other words, we can say that Suhrawardī begins by identifying being with being known: on the fundamental level, to be is to appear. Later on in the work, Suhrawardī situates this foundation in an emanationist framework: there is one source of all light, who in the superabundance of His luminosity gives rise to a series of further lights, in a manner familiar from Avicenna's cosmology.50 However, there are two important deviations. First of all, Suhrawardī emphasises that the differentiation of lights is exclusively bil-tashkīk, that is, all differences between them can be reduced to degrees of comparative perfection and deficiency,51 with no need for any essences distinct from light. Secondly, he maintains that the number of lights is innumerably greater than that of the celestial intellects in Avicenna's more sparsely populated system, for he claims that there must be a light corresponding to each of the fixed stars on the outermost sphere. This explosion of multiplicity in the order of emanation is explained by means of an intricate series of diffractions, reflections, and conjunctions of lights, which gives rise to a complex system of vertical (or hierarchical) and horizontal relations between lights.52 The details concerning this series remain sketchy, and by way of a conclusion, Suhrawardī confesses that for instance the intellectual order underlying the visible arrangement of the fixed stars cannot be encompassed by human knowledge.53 For a metaphysician with more general concerns, however, it suffices to know that its outcome is a comprehensive world of Platonic Forms.54 This theory of Forms is designed to avoid the features of Peripatetic metaphysics that Suhrawardī has criticised in his discussion of the iʿtibārāt. The Forms are not distinct Avicennian quiddities that determine existence, but of one and the same light in relative degrees of perfection. By the same token, they are not subject to categorical analysis, and finally, each Form is strictly one in itself in the sense that it does not have more basic constituents. Although Forms are products of the diffractions, reflections, and conjunctions of hierarchically superior lights, these higher lights are not their constituents the way genera and differentiae are constituents of species. To drive home this point, Suhrawardī uses the analogy of physical light, arguing that just as the lights from several lamps can merge into a single light, in which the respective contributions of the lamps are inseparable, similarly incorporeal lights can fuse into a Form that has no metaphysical parts.55 All well and good-but how are the Forms related to concrete things? In order to make sense of this, Suhrawardī has earlier introduced a fivefold ontological framework consisting of pure lights and accidental lights, dusky substances and their dark states, and what he calls "barriers" (barāzikh, sing. barzakh):

Things are divided into [what] is light and shining (ḍawʾ) in its own reality and into what is not light and shining in its own reality. […]
Light is divided into that which is a state of another, that is, accidental light, and into light, which is not a state of another, that is, incorporeal light or pure light. That which is not light in its own reality is divided into that which is independent of a substrate, that is, the dusky substance (al-jawhar alghāsiq), and into that which is a state of another, that is, the dark state (al-hayʾa al-ẓulmāniyya). The barrier is the body, and it is described as a substance that is pointed to ostensively ( yuqṣadu bi-l-ishāra).
[…] If light is cut off from a barrier, it does not need anything else to be dark, and so these barriers are dusky substances.56 If the Forms are pure lights, it seems natural to conceive of their individual instantiations as accidental lights. When an accidental light appears on a dusky substance, it renders the substance a barrier. The barrier is an intermediate entity between light and darkness, or that which appears and that which by definition cannot appear. To resort to a simile, the dusky substance is like a screen on which the accidental light appears: just as the accidental light requires the screen as a necessary condition of shining upon another,57 the screen can only be seen when it is illuminated. Hence, the barrier is not really a fifth sort of entity, but the meeting point of accidental light and dusky substance, yet at the same time it is the only way for either of them to appear. Light will not be accidental unless there is an other that functions as a barrier, whereas the substance cannot appear unless it is illuminated, and thus a barrier. The striking feature of this ontology is that the substance that Suhrawardī had relegated to the status of an iʿtibārī concept makes such a swift comeback. What is more, the description of the barrier as a meeting point between emanated light and receptive darkness is not entirely different from the Avicennian theory of the generation of sublunary contingent things. For Avicenna, instantiations of Aristotelian species forms are emanated from the active intellect only when sublunary matter is sufficiently prepared to receive them, and so the material substratum and the individual instantiation of the form can only be actualised together58-like the accidental light and the dusky substance when they meet in the barrier. But these similarities notwithstanding, there is a crucial difference between Avicenna's and Suhrawardī's accounts of concrete entities. Avicenna is firmly committed to the idea that both the individual substance and the matter and form that constitute it are real and metaphysically distinct things, independent of the mind thinking about them. For Suhrawardī, on the contrary, only the incorporeal light's appearing in another is independent of the mind. True to his earlier discussion of the iʿtibārāt, he maintains that "the substantiality of the dusky substance is intellectual and its duskiness privative," and that the dusky substance "does not exist insofar as it is like that," that is, insofar as it is a substance.59 Why then reintroduce the concept of substance in the first place? By way of answering this question, let us return to Suhrawardī's account of substantial infimae species. Of particular interest in this regard is an intriguing passage in the Muqāwamāt, a set of objections and responses collected as an appendix to the Talwīḥāt. In a context dealing with the iʿtibārāt, Suhrawardī engages in a discussion concerning simple and composite infimae species and the way in which they are perceived: The gist of recognising (maʿrifa) species in simple [things] is that what is taken as a species has a perfection of quiddity Primitive objects of sense perception, like black, the body as a three-dimensional continuous whole, or the human soul, are perceived as simple, which means that in perceiving them, we are not aware of any constitutive features. 61 We have already discussed the case of the black and seen that for Suhrawardī, concrete instantiations of colour are paradigm cases of simple percepts. The simplicity of the human soul (nafs) is also quite uncontroversial, for immediate perception of soul means being aware of oneself, which Suhrawardī, following Avicenna, takes to be a primitive feature of human experience.62 The case of the body seems more difficult, for one might object that a three-dimensional continuous whole is constituted by more basic geometrical entities, such as two-dimensional planes. However, Suhrawardī probably means that this is not how we primarily perceive bodies. Instead, we perceive them as simple, and only in mental analysis can we define the concept of body by means of the concept of plane. Be that as it may, for our concerns the more interesting class of species is that of the composite ones, all three examples of which are cases of substantial infimae species. They are characterised as enduring through the variation of their features, as expected of substances, an obvious example being a human being whose identity remains intact through a process of tanning. Later on in the same paragraph, Suhrawardī maintains that the infimae species concepts of both the simple and the composite kind, unlike their generic and differ-60 M III.41, 170-1. 61 An alternative interpretation is that by simplicity Suhrawardī means the absence of metaphysical constitution. These two kinds of simplicity, epistemic and metaphysical, need not rule out each other, of course. But since Suhrawardī excludes metaphysical constitution from the composite things he mentions here (see, for instance M III.41, 171-2, with humanity as an example), it seems natural to think that he has epistemic simplicity and composition in mind here. Suhrawardī does seem to recognise two kinds of metaphysical composition, though: composition in the sense of constitution (which he rules out) and composition from two distinct entities, with the composition of a human being from body and soul as an example. The latter kind of composition could be meant here, but since water is mentioned together with man and horse, this seems implausible. On water as a substantial species, see also MM III.2.5.60, 290-1. 62 Cf. M III.56, 186; for a general discussion, see Kaukua, Self-Awareness in Islamic Philosophy, 104-23.

kaukua
Oriens 48 (2020) 40-66 ential constituents, have real, mind-independent correlates: "The species [of animal] is an occurring quiddity [that is, not a mere iʿtibār], which is not specified by anything apart from relations, except (lā yukhaṣṣiṣuhā mā warāʾa liḍāfāt illā) such things that [can] be estimated to be replaced with the natural identities (huwiyyāt) remaining without them."63 He specifies that these things are real only when considered as natures, whereas if conceived in categorical analysis as species constituted by genera and differentiae, they are iʿtibārī like their generic and differential constituents. However, as natures they suffice to ground the correspondence relation between the first-order species concepts and the world, and thereby provide the truthmakers to the generic and differential concepts. But what does Suhrawardī mean when he says that the substantial nature "is not specified by anything apart from relations?" This is explicated at the end of the paragraph in a way that ties the discussion from the Muqāwamāt to our question concerning the role of substance in ishrāqī metaphysics. Suhrawardī says: We point attention to our saying "apart from relations" about the species, for the relations [between] simple accidents cannot be imagined (tawahhumuhā) to be replaced with the identities (huwiyyāt) remaining the same.64 The point is that substantial natures, like human, horse, or water, are perceivable only as enduring structural relations between primitive sensible variables. To put this in another way, the relations between the relevant sensible variables must endure for as long as the nature is to remain, but the values of the variables can change.65 At the same time, the composite relation can only be perceived through the primitively perceivable features that are interrelated in it. What is more, in order to perceive the endurance of the relation, which is a necessary condition for the distinction between the nature and its accidental features, Oriens 48 (2020) 40-66 one must combine the two aspects of variation and stability. The question is, is this simultaneous distinction and combination of the two aspects possible without the concepts of substance and accidents? In this regard, the substantial infima species of the Muqāwamāt comes intriguingly close to the dusky substance of the Ḥikmat al-ishrāq. Its endurance is not perceived at any moment but is merely assumed to hold diachronically despite the change of the immediately appearing features. Since it is not perceived, it is dusky and can never appear as such; and since this assumption of substantiality is based on a concept of the perceiving mind, it is iʿtibārī. Yet the assumption of substantiality seems necessary for the human perception of an important part of reality, namely the individual instantiations of substantial Forms.
Let me elucidate this idea by means of an example. Consider a concrete individual horse. In Suhrawardī's terms, it is the appearing of the Platonic Form of horseness in and to another, that is, in the dusky substance of this individual, to you who are considering it. Now, considered in itself and as a Platonic idea, horseness is an atemporal, fully actualised presence of all the features that belong to horselike perfection. But encountered in another, as an individual horse, horseness is extended over time in the manner of a substantial nature that develops towards and flourishes in horselike perfection over the course of its existence.66 But the constantly varying immediately perceivable features, through which horseness appears, can only appear partially and in a consecutive manner in the individual horse. For instance, the horse's power and speed are merely implicit, or potential, when it grazes in the pasture. If we had access to the Form of horse as it appears in itself, all of these features would appear simultaneously, but when horseness appears to us in the barzakh of a dusky substance, the simultaneity is broken into a series of actual and potential appearing. We can perceive the whole in the instance only by means of the concept of substance, or the unchanging relation between the changing simple features in terms of the Muqāwamāt, that we assume to be manifested in the instance.
Thus, despite his denial of the mind-independent reality of substantiality, Suhrawardī needs the notion of substance to explain those features that are crucial to an accidental light's appearing in and to another but that cannot be derived from appearing alone. The momentary appearance of the horse at the pasture cannot betray its future behaviour at the race track. But although our perception of the individual horse as a thing that endures through time 66 Cf. MM III.3.6.113, 364. kaukua Oriens 48 (2020) 40-66 depends on subsuming its appearance under the concept of substance, from the metaphysical point of view the identity of the thing is not due to any constitutive substantiality-only the Form of horse has that kind of causal power.67 In the ishrāqī context, substantiality is no longer the foundation of metaphysics, but it is required to account for a certain mode of appearing, namely the appearing of concrete objects distinct from us. In this restricted role, the mind-dependence of substantiality is no longer a problem, for that mode of appearing is itself dependent on the mind, or in Suhrawardī's own words, "one who is not aware of himself cannot be aware of another."68 Appearing as an object requires a subject to which the object appears.

Conclusion: Ishrāqī Transcendental Concepts?
The status attributed to an iʿtibārī notion in Suhrawardī's metaphysics of light is by no means unique to substance. There are passages in which he admits that the modal concepts, the concept of existence, or the concepts of essence, reality, and quiddity are crucial for the adequate perception of extramental reality, even though these concepts do not have distinct counterparts in it.69 By means of these concepts, we can conceive of aspects of concrete things, such as their contingency and the entailed createdness, that are true even though they can never appear in concrete. Moreover, it is only by means of iʿtibārī concepts like 'relation,' 'genus,' 'differentia,' and 'species,' that we can articulate and know the similarities and differences between individual things on which our sciences are based. And as we have seen, the iʿtibārī concepts are all grounded in infima species concepts, which in turn are true by way of correspondence with mindindependent reality. On the other hand, the central ishrāqī notion of pure light, such as that of the Platonic Forms, is described as "light in itself and to itself" ( fī nafsihi li-nafsihi) whereas the accidental light is light in itself but to another (li-ghayrihi).70 Thus, 67 See the discussion of metaphysical causality that follows the introduction of the notions of dusky substance and accidental light in ḤI II. ḤI II.2.6.121, 83. The second part of this description is ambiguous and could also be translated as 'due to itself' or 'due to another,' denoting some kind of dependence relation.
Oriens 48 (2020) 40-66 accidental lights are there only insofar as there is a subject distinct from the source of that light that can perceive it. Ishrāqī metaphysics is thus a kind of phenomenalism. But to be a viable alternative to Avicennian substance metaphysics, it must be able to salvage the intuitions corroborating the latter. If you were to found metaphysics anew on a phenomenal concept like 'appearing,' one potent way of convincing your Peripatetic readership would be to accommodate their view in your new system, particularly if that view is perceived to fare especially well in terms of our everyday ontology of robust substances. I want to suggest that it is this kind of reductive explanation that Suhrawardī's ishrāqī philosophy hinges on: yes, Peripatetic metaphysics aptly describes the world as it appears to us, only we should not confuse that appearance with the foundations of reality. In light of these considerations, I conclude that it is indeed helpful to think of Suhrawardī's iʿtibārāt as transcendental concepts. But unlike Kantian transcendental categories, which mark the limits of thought and knowledge, the iʿtibārāt are applied to input from a reality with which we do have some immediate acquaintance and the true nature of which we can know independently of iʿtibārī assumptions. Although Suhrawardī is openly sceptical about our possibility of exhaustively knowing the realm of the Platonic Forms, we do at least know that they exist and that they are the principles of the world that we perceive. Perhaps there is even a way to know these Forms directly by way of experience (bi-l-mushāhada), entirely unconditioned by the iʿtibārāt. Nevertheless, if the reconstruction I have sketched out is anywhere on the right track, Suhrawardī seems to have felt a genuine need to accommodate Peripatetic metaphysics in the ishrāqī system as a broadly accurate account of the world as it appears to us.