And this is just to say that a teacher, in teaching something to a student, has to somehow bring about one or more of these states in the student. The way or ways the teacher does this i.e., brings them about in the student, these are called ‘the modes of teaching’ (anha’ al-ta’lim). Now the modes of teaching will differ in two ways. First, in accordance with the differences in the things used (in the teaching) to bring about one or more of those three states in the student; and, second, with the differences in the respects in which many of those things are used. What are these things? And what are these respects?

The Second Teacher (you should now have some sense of why he has that title in the tradition) answers the first question by stating (ibid., 87, lines 12–16):

These things are many. Among them is the use of expressions that signify a thing, the definition of a thing, the parts of its definition, its particulars and its universals, the descriptions of a thing, its properties, its accidents, a thing similar to it as well as its opposite, division, analogy, induction, syllogism, and placing the thing before the eye. All of these except syllogism are useful in facilitating understanding and conception.

He then explains, with a view to the second question, how all these things (excluding syllogism) may be employed in doing that i.e., in bringing about, when teaching something, a conception of the thing in the student. There are three strategies that may be used. In what follows, I want to present to you a translation of what he says about a sub-set of the third type.

Generally stated, the third type consists in ‘substituting one of the things mentioned above (except syllogism) for the thing itself’ – for that substitute might make conception of the thing easier on the student. And one particular way someone might do this is by ‘substituting an accident of the thing for the thing itself’. But having noted this, Fārābī then goes on to make a critical point by objecting to some of the Presocratics, as well as Plato, for employing this mode of teaching in scientific contexts, where it is problematic and best avoided because, he implies, it can foster mythological thinking. In political and rhetorical contexts, however, he notes that it is both unavoidable and useful. Here, “political” should be taken in a broad sense i.e., encompassing what we would now call ‘partisan politics’ but also civic education, legislation, religious and moral formation, exhortation, public deliberation and, more generally, the discourse through which a community is directed toward common ends.

Anyway, the full passage (full of interesting material) from The utterances employed in logic (ibid., 90-92) runs as follows:

As for substituting the accident of a thing for the thing, Aristotle completely avoids this mode of teaching in philosophy. Likewise with substituting what resembles a thing for the thing – Aristotle avoids it except in a few cases. And it may be possible for these substitutions to be composed in various kinds of composition. For example, the accident of a thing gets substituted for the thing, then that accident is substituted for what resembles it. And this latter is the worst among the modes of teaching. Worse than all of that is what is composed in compositions much greater. An example of that is substituting the universal of a thing for the thing, and then the universal for its property, and the property in turn for an accident of it. Thereafter, taking what resembles that accident in exchange for the accident, and then placing the name of that resembling thing in exchange for the resembling thing itself. The hearer and learner is then rendered very remote from the intended thing. This mode of substitution was used by many of the followers of Pythagoras and those who preceded Plato. Among the companions of natural science, Empedocles used it. Of this mode is the account that Plato mentioned in the book known as Timaeus: namely, that the Demiurge took the straight line, split it and then bent it from straight to circular, split that lengthwise into two circles, and then divided one of the circles into seven circles. Due to that, the heaven undergoes circular motion. This is the worst of the possible modes of teaching. Aristotle explicitly regarded this mode of teaching as base, stating: “As for those people, their efforts are directed only toward making themselves understand; their efforts are not directed toward making us understand. Rather, they are remiss in that. It is known that what they say is familiar among themselves, but what they posit through these statements lies beyond our understanding. Likewise, it is not necessary to examine the statements of those whose philosophy proceeds through myths.” Through these ways, then, are composed the statements called symbols and enigmas. Perhaps these [i.e., symbols and mysteries] are not repudiated except in the modes of philosophical instruction. As for in rhetoric and statements employed in political affairs, perhaps nothing other than these is necessary.

The Islamic philosopher Ibn Rushd thinks that philosophy is no different in this regard. It is not something one dabbles in e.g., indulging a curiosity here, picking up an argument there, and an objection some other place, etc. It is a science, and, like every other science, it makes demands upon the person who wishes to practice it. On this basis (as well as others), he seems to write off kalām at one place in Tahāfut al-tahāfut (TT) as unnecessary and undesirable, stating the following:

If you are of the people of a natural disposition (fitra) prepared to receive the sciences, and of the people of persistence and [also] leisure, then you are fit for looking into the books and sciences of the philosophers, in order to be informed about what is in their books of truth or its contrary. But if you are deficient in even one of these three, then it is apt for you leave that and tend to the outward revealed law, and not to look into these doctrines newly introduced into Islam. For if you are of their people, you will belong neither to the people of certainty nor to the people of the revealed law.

That is, one should enter into philosophy proper if one has the requisite qualifications. The first of them is the natural aptitude to receive the sciences; the second, the intellectual steadfastness to pursue them; and the third, the leisure required to devote oneself to them. As noted in a previous post (here), philosophy is a craft. It requires years of training, the patient acquisition of its principles, and the ability to follow a demonstration wherever it leads. We readily accept this in the case of physics, biology, or any other science. By Ibn Rushd’s lights, there is no reason to treat e.g., metaphysics, differently. However, if one does not have the requisite conditions, one should cruise along with the outward teaching of religion. Our mutakallim is problematic because he seems to occupy a confused halfway house between these two: neither here nor there, neither a philosopher nor an ordinary believer.

Let me illustrate the contrast between the three for Ibn Rushd in terms of how they’d approach a given theological issue (mas’ala). Take the question of tawḥīd (monotheism) - how many gods are there?

The philosopher investigates it (assuming it is not self-evident) as one would investigate any other object within a science. Just as the geometer investigates the properties of a triangle from the principles of geometry, or the chemist investigates hydrogen from the principles and methods of chemistry, the philosopher begins with the principles proper to general metaphysics (e.g., being, unity, causality, necessity, etc.) asks what follows from them in this particular (divine) case. Consequently, God’s uniqueness, if it is established, emerges as the conclusion of a demonstration. What the philosopher does not do or should not do is begin with the proposition that God is one and then search for arguments that will preserve it. The proposition itself is what must be understood and (if isn’t self-evident) demonstrated.

Alternatively, one may accept God’s uniqueness as a revealed truth. One then believes that God is one, worships Him as such, and allows that truth – embedded within the rest of religion and its laws - to order one’s life, socially and politically. There is nothing untoward about this position. It is a respectable posture of imān rather than scientific investigation and, as we saw, is an attitude most apt for folks who lack the aptitude, seriousness, and leisure demanded by science.

But what, then, is the mutakallim doing? He does not approach the issue of tawḥīd as an open question, since he begins with a firm conviction about it and its place within his religion. Thus, he does not (test this for yourself) regard its denial as a genuine possibility. Yet neither is he, like the lay believer, content simply to receive it as a truth of revelation. Instead, the mutakallim enters into rational disputation in order to prove what he already holds to be certain. His inquiry, in other words, begins with its conclusion already fixed. As a result, for him the task is not so much to discover whether God is one as it is to come up with arguments that will establish, mainly for others, that He is so. The trouble with this approach, from the ḥikma point of view, is that the argument is being tested by whether it delivers the required conclusion. Hence, if it turns out to be a failure, the doctrine would not be given up but rather another argument would just be sought for it instead. Reason is therefore not operating as an independent route to knowledge but as an instrument recruited in the defense of something not only known in advance but, in lots of cases, committed to as non-negotiable in advance. In Epitome of the Metaphysics, Ibn Rushd characterizes those with such an MO as follows:

These people - as we have said more than once - do not hold these views because rational inquiry has led them to them but rather in order to vindicate things they initially based on their truth and agreed upon among themselves. Consequently, they seek to falsify whatever opposes those them and to establish whatever supports them.

This, I take it, is the force of the closing remark in the TT passage above. The mutakallim then belongs to neither class; he is not one of the “people of certainty,” since his mode of proceeding is not properly scientific, nor is he one of the “people of the Law,” since he is no longer content with the outward and simple teaching of revelation. Indeed, in the case of the layperson, kalām may do worse than merely fall short of knowledge (ʿilm/epistēmē). By treating metaphysical-theological questions not as objects of science, but (again: test this for yourself) as occasions for deploying a grab bag of argumentative moves to prove or defend - against e.g, atheism or rival religions – antecedently held beliefs, it may weaken the very conviction it was meant to generate. So it ends up unsettling iman without attaining knowledge. Thus, the Qāḍī’s verdict: if one is capable of philosophy, one should pursue it seriously; if not, know you role and stop talking (takallum) about such matters i.e., especially in educational and political contexts.

Given the above considerations, I don’t think it would be too off the mark to say that, for the Commentator, an ideal Islamic society is one in which there are Peripatetic and Athari types but no Ash’aris, Maturidis, and their like.

In the Aḍḥawiyya, Avicenna identifies three factors, all closely connected. He states:

As for the group that construed the Return as belonging to the body alone, what drove them to this is what the Law came with as regards the return of the dead. Subsequently, they surmised that the thing that is considered the essence of man is the body. And then, in their excessive hatred for the philosophers and their passion for opposing them, they went so far as to deny that the soul or spirit has any existence whatsoever, and that bodies become living [only] in virtue of a life created in them, whose existence is not the existence of the soul for the body but rather is an accident from among the accidents created in them.

So the first factor is their adherence to the surface level (ẓāhir) meanings of scriptural language about the Return. Since scripture depicts the afterlife in markedly bodily terms (e.g., resurrection of the dead, physical pleasures and pains, etc.) the theologians took the Return to consist solely in the restoration of the body. The second is their materialist conception of the human being: they rejected the existence of an immaterial soul capable of subsisting apart from the body and consequently regarded the essence of the human being as an essentially bodily entity. The third is their hostility toward the philosophers and their desire to oppose them (as they held a view reckoned by the theologians as contrary to what the ẓāhir of Scripture required). According to Avicenna, this opposition became so extreme that the theologians not only denied that soul or spirit possesses any (immaterial) existence, but also maintained that a body is alive merely in virtue of an accident called “life” (ḥayāt) that is created in it. On this account, life’s belonging to the soul is, not the soul existing for the body, but rather just one accident (among the others) inhering in the body.

From the way Avicenna frames it, he appears to present these factors as stages in a connected development (not merely as three independent considerations). As if to say: a literal adherence to the surface meaning of scriptural language about the Return first encouraged the theologians to conceive the human being in exclusively bodily terms. This materialist anthropology, once adopted, was then hardened and carried to its furthest conclusions by their polemical opposition to the philosophers who, in their eyes, were teaching something contrary to Scripture with their dualism about man. Thus, their hostility toward philosophy did not originate the bodily interpretation of the Return; rather, it reinforced that interpretation and drove them to deny the immaterial soul altogether. The sequence may therefore be represented as follows: a surface-level reading of scripture led to a materialist conception of the human being, and that conception, intensified by anti-philosophical polemics, culminated in the denial of any immaterial principle of human life. As result, they ended up, naturally enough, committed to a purely materialist account of the Return.

There’s broader lesson to be learned here for the perspicacious. It is that the surface-level language of Scripture, if handled without interpretive care, can misguide i.e., be metaphysically misleading. Of course, the point is not that Scripture is false, but that the form in which it addresses its audience (in particular in its use of sensible, corporeal imagery) does not necessarily provide a transparent (or exhaustive) account of the realities to which it refers. When such language is treated as though it were itself a precise metaphysical theory, one may illegitimately move from e.g., scriptural descriptions of bodily resurrection, to claims about the nature of the human being, the soul, etc. Indeed, the passage from the Aḍḥawiyya suggests that this is precisely what occurred: a particular reading of eschatological language gradually generated an entire materialist anthropology. Although the Shaykh develops this hermeneutical point more fully only after the passage quoted above, it casts that passage in a revealing light. Substantive metaphysical positions should not be established merely on the basis of Scripture’s surface meaning, but must also be subjected to demonstration and to careful reflection on the purpose, register, and intended audience of scriptural discourse. By the same token, affirming or denying such contested metaphysical positions should not, without further argument, be made a basis for takfīr or similar forms of exclusion. If the surface language of revelation does not by itself settle the underlying metaphysics, then disagreement over that metaphysics belongs principally to the domains of philosophical demonstration and interpretation, rather than serving as a straightforward criterion by which the boundaries of faith are policed.

Contrast it with one we get from the Islamic philosophical tradition (ḥikma), as articulated by the Don (Ra’is) i.e., Avicenna, in one of his treatises. There he describes it as something considerably more demanding: a person equipped to attain truth, realize the good, perfect the soul, and become - within the limits of human capacity - an intelligible likeness of reality itself. If Sellars’s philosopher is just someone broadly familiar with the terrain, Avicenna’s philosopher is someone perfected by what that terrain contains. Let’s consider why more closely.

In Treatise on the divisions of philosophy and their details, the Shaykh defines ḥikma as follows:

On the what-it-is of philosophy: philosophy is a theoretical art from which the human being profits the obtaining [i] of the whole of existence as it is in itself and [ii] of what is obligatory from among that which one ought to acquire by performing it, in order to, by that, [iii] ennoble and perfect his soul, becoming an intelligible world reflecting the existent world, and [iv] prepare for the supreme otherworldly felicity, all according to the extent of human capacity.

A few words are in order towards unpacking some of the elements of the account:

The Shaykh calls ḥikma a “ṣināʿa naẓariyya” i.e., a theoretical craft. It is a theoretical craft or skill as opposed to a practical one (like carpentry); and it is a theoretical craft or skill as opposed to being an organized collection of true propositions (like a science e.g., geometry). Ṣināʿa indicates a stable, rule-governed capacity acquired through effort (e.g., study and practice). Philosophy, therefore, is a kind of artistic ability i.e., that allows one to easily do something properly i.e., in an ordered way.

To do what though? To gain two things: [i] apodictic beliefs alone i.e., about the entirety of reality as it is and [ii] true beliefs about what ought to be done/not done and, additionally, the doing/not-doing of them.

As regards [i], this identifies the object of the theoretical branch of philosophy (as opposed to the practical one (for which see below)): reality considered independently of our desires, conventions, and activities. Philosophy is a skill that allows us to probe reality profitably (i.e., for scientific certainty) when we’re confronted with it – analogously to the way carpentry is a skill that allows its possessor to profitably handle wood when he encounters it.

As regards [ii], this is the practical half of philosophy. Some things exist independently of us; others – such as actions, virtues, political arrangements, and states of character, etc., - come to be through human choice. So philosophy discerns not only what is true (and false i.e., theoretical philosophy), but also what is good (and evil) and so ought (and ought not) to be realized. As such, it is a skill that allows one to understand the right things and which of them to bring about and which not, where these latter are not merely additional objects to be understood. Consequently, the art that is philosophy also governs action and produces the appropriate perfections (the virtues) in the agent.

What is the point of gaining, in virtue of the art of philosophy, [i] and [ii]? On the Avicennian view, it is two fold: [iii] and [iv].

The basic idea is that the possession of theoretical and practical wisdom has a transformative end – this worldly [iii] but also otherworldly [iv].

In knowing the true and in doing what is right, the human soul actualizes capacities that are perfective of its kind, intellectually and ethically. Intellectual perfection comes through apprehending reality; ethical perfection comes through ordering desire and action according to the good. To the extent that it does this, and does it well, it is [iii] rendered a nobler instance of its kind - so much so that, ideally, the structure of mind-independent reality as a whole comes to exist within the soul in an intelligible mode. The soul thereby becomes a kind of intellectual microcosm corresponding to the external world: the perfected soul mirrors within itself the way things are, not by materially containing them, but by apprehending their intelligible forms, principles, causes, and order. This ennoblement of the soul through the sciences and the virtues also has a soteriological component [iv]. It constitutes the flourishing of the human soul as the kind of entity it is, not only in this worldly life, but also by preparing it for its ultimate felicity in its post-mortem life - its true and enduring life. So this flourishing and felicity [iii-iv], as the consummation of its (theoretical and practical) perfection, is that for the sake of which we gain [i-ii] through the art of philosophy. What the soul becomes through knowledge and virtuous action determines, not its being, but its well being i.e., the quality of the life it enjoys, both now and once separated from the body.

The contrast between the two accounts is stark indeed.

The gist of it is this: all cognition seems to be tied to the cooperation of sense and intellect. That is, the senses give us particulars, while intellect forms universals from such particulars. However, since immaterial entities are not available to the senses in principle, the issue is how could the intellect form any real concept of them, let alone affirm their existence. Here it is as Ibn Miskawayh states it:

An objector says: if no universal is known by the intellect, and no particular is known by sense, except that each of the two has the other accompanying it - so that confidence in the testimony of one is secured by the testimony of the other - and if someone who lacks a sense also lacks sound thought concerning its universals, just as he lacks it concerning its particulars, then from where do we have confidence in affirming a thing that is not one of these two? For it would be one of them, yet it wouldn’t have the testimony of the other through the path by which it is reached.

The force of the objection depends on the assumption that the intellect has no independent access to realities beyond what sense first supplies. Miskawayh says that the proponent of this sort of view has made ‘the intellect subordinate to the sense’. That assumption, however, is questionable and indeed, upon scrutiny, false.

But that aside for now. Let us unpack the thinking in it bit more as it is worth reflecting over. It seems to be as follows:

Prima facie, every genuine concept is grounded in sense-experience; that is to say: of the familiar things we have concepts, we first encountered them through sense. For example, this white thing, that sound, this animal, that hot object, etc. Then, our minds abstracted universal concepts from these sensory particulars. For example, from many particular humans that we’ve seen, we form the universal concept “human.” From many particular horses, we form “horse.” From particular tokens of whiteness, we form “white.” You get the idea.

If so, then sense and intellect seem to always accompany one another in this epistemic process. Without intellect, it seems that sensation would have given us only confused particulars; likewise, without sensation, intellect would have had no material from which to form a universal concept. Thus, the testimony of one is confirmed by the other i.e., particular sense impressions lead to universal concepts, and universal concepts go back to sense impressions. And this cooperation between sense and intellect explains why when we lack a sense we are prevented from having the concepts pertaining to that sense. As the Aristotelians says, man faqada hissan faqada ‘ilman. Examples: for someone born blind cannot have a proper concept of color in the way a sighted person does, because they have never had the relevant visual impressions. Someone born deaf cannot have a proper concept of melody or pitch in the way a hearing person does. Yes, they may know words and or analogies, but they lack the original sensory basis from which those concepts are formed. In that respect, they lack the relevant concepts proper. On the Aristotelian view, then, Mary couldn’t not know anything about color as, in the absence of sensing some color or other, she couldn’t have the concept of color to begin with.

If all that is right, then if something is in principle beyond all sense-experience, it seems we cannot have a proper concept of it. Now ‘gods’, ‘souls’, and the like, are supposed to be non-physical items – in that they are not visible, audible, tasteable, smellable, and touchable. We do not see one god here and another there, as we see one horse here and another there. Nor do we touch, hear, smell, or taste ‘souls’. Hence, if all concepts arise from sense-impressions, and there is no sensory source in principle from which the concepts of them could be taken, then we have no proper concepts of ‘god’ and ‘soul’. They are mere written or verbal marks. And if we lack proper concepts of them, then our affirmation of such items is empty and merely verbal and, consequently, without any sense and unjustified.

In sum, if every intellectual universal is accompanied by sensible particulars, and every sensible particular can give rise to some universal conception, we cannot have trust and confidence in an affirmation of something that lacks this reciprocal support. For unlike in ordinary cases, where intellect and sense validate one another, with putative immaterial entities this co-ordination is absent, and the mind lays claim to something for which sense gives no corresponding testimony. Consequently, any reasoning that employs such terms will equally suffer i.e., it will be baseless. Belief that there are such things is therefore undermined.

According to naturalist philosopher Kim (Physicalism, or something near enough. Princeton: Princeton University Press 2005), we would do so by pairing cause A with effect C, rather than D, and cause B with effect D, rather than C, by appealing to spatial relations: direction, distance, orientation, trajectory, and so on. Thus, it is because gun A was positioned and aimed in the direction of C, rather than D, that C was struck by a bullet from A rather than from B. Likewise, it is because gun B stood in the relevant spatial relation to D that D was struck by B rather than by A. This illustrates, Kim thinks, a general feature of causal explanation. For we do not explain the causal facts merely by listing the causes and the effects; we also explain how a particular cause is paired with a particular effect. In paradigmatic cases, this pairing is secured by spatial relations: the relative location of the cause, its direction of activity, its distance from the effect, and the path or medium through which its influence is transmitted. Spatial relations therefore play a central role in individuating causal connections among physical things.

Kim then uses the above as a premise in an argument against substance dualism – the thesis that the human soul/mind is a non-physical substance. The gist of that argument goes like this (Second Figure syllogism):

No human soul/mind is spatial
All causal-pairings require spatial relations
Therefore, no human soul/mind can stand in causal-pairings

The argument is clearly valid. The substance dualist will grant the minor. But, obviously enough, the controversial premise is the major. For Kim, the upshot of the conclusion is that (mental) causation is impossible for non-physical minds.

Kim motivates this last thought by asking us to imagine two psychologically synchronized individuals, Smith and Jones. Whenever Smith wills to raise his hand, Jones wills to raise his hand as well, and their hands always rise simultaneously. The question is then: why is Smith’s willing causally paired with Smith’s hand rising rather than Jones’s, and vice versa? If minds were spatially located, their spatial relations to bodies could explain such causal pairing. But if minds are non-spatial entities, it becomes unclear what grounds the causal connection between this mind and a this body as opposed to that body.

Interestingly, eight centuries earlier the Islamic theologian-philosopher Fakhr al-Din Rāzī (d. 1210) advanced an argument against Avicennian substance dualism that can be plausibly characterized as a Kim-style causal pairing problem. For it concerns something near to Kim’s worry, namely, what fixes the attachment of this immaterial soul to this body rather than another. However, though both pairing problems try to exploit the same feature in the dualist model, one difference between them is that Kim’s version frames the problem synchronically (i.e., at one time, why is this mental cause paired with this bodily effect?), whereas Rāzī’s frames it diachronically (i.e., across time, why is this same soul still paired with this same body?).

The Rāzīan argument unpacked

Rāzī raises the objection in Exalted Topics, in the volume dedicated to issues of spirits and souls. He states that if, as the Avicennians hold, the soul is an immaterial substance merely “attached” to the body by governance or employment, then, it seems possible that the governor of this body could be replaced by another governor. And if that is possible, then bodily continuity would no longer secure personal identity. Hence, seeing a given person e.g., Zayd, again would not be enough to know for sure whether he is same human being one initially saw, since the original soul may have departed and another may now be governing the same body. What Rāzī seems to be urging is that, once the soul is conceived as an immaterial substance externally related to the body by an extrinsic relation (= governance and employment), there seems to be no principled ground for the continued pairing of this soul with this body as opposed to that one. In other words, if the relation of rule/management is extrinsic and detachable, then it is at least prima facie possible that the original governor of Zayd’s body should depart and another immaterial governor take its place. In that case, when we see Zayd’s body again, we couldn’t be certain whether it’s the same human being, since the initial human being would have been replaced if the governing soul had been replaced. In this respect, Rāzī anticipates the core structure of Kim’s causal-pairing problem: the dualist owes an account of what individuates the relation between a particular non-spatial soul and a particular body instead of another.

And like Kim, Rāzī motivates the argument with an analogy. Suppose, he says, we see a certain house and allow that its ownership may be transferred from one owner to another. When we later see the same house, we would not be certain that its owner is the original owner; that would remain doubtful. The visible continuity of the house does not secure the continuity of its owner, because ownership is an external and transferable relation. Rāzī’s point is that, on the dualist theory, the soul-body relation is similarly external such that nothing seems to prevent the attachment of this immaterial substance to this body from being severed and replaced by the attachment of another immaterial substance. Consequently, seeing Zayd’s body again would not secure certainty that the same human being remains, any more than seeing the same house again secures certainty that it has the same owner.

But, Rāzī concludes, no rational person actually doubts that it is exactly the same person that we witness the second time around. Thus, there is a tension between (1) what substance dualism requires if true i.e., doubt, and (2) what common sense requires i.e., certainty. And given that we are as a matter of fact certain about the matter, it is substance dualism that must be given up.

Here is the full passage from Exalted Topics II.2, 46 (ed. Saqa, Beirut: Dar al-Kutub, 1987/1407, vol. 7):

If ‘soul’ were an expression for an ‘immaterial substance that is attached to this body by way of employment and governance’, then it would not be prima facie implausible for the attachment of this soul to this body to be severed and for it to become attached to a second body, or vice versa. On this supposition, if we were to see the body of Zayd and then see it again later, we would inevitably remain doubtful as to whether this human being is [the same as] the first human being. For, on the supposition that the first governor of this body has been replaced by another governor, that human being has been replaced by another [human being]. This is just as when we see a house and allow that the ownership of that house may be transferred to another owner: upon seeing that house a second time, we would not then be certain that its owner is [the same as] the first owner. Rather, we would be doubtful about it. Likewise, in the present case, doubt must remain as to whether this human being whom we now see is the same human being whom we initially saw. But since this doubt does not occur to any rational person, and since it has been established that, on the supposition that the human being is something other than this body, this doubt must occur, we know that the doctrine that the human being is something other than this body is a false doctrine.

I base myself on something Sijzī says in the Maqālīd about divine power and the general doctrine of radical tanzīh (divine transcendence) the Ismailis defend. Here’s the relevant passage from that work (ed. Poonawala 2011, 54):

T1. As for the power of the Originator (God) the exalted, it transcends the need for a capacity. Indeed, the power of the Originator the exalted is so hidden no one has any way to grasp it. For it is beyond (khārija) all the beings and, further, its modality is not relatable to intelligence. […]. It has thus been established that the Originator the exalted originates all things by making being [ex-nihilo] (ta’yīs), not by way of craft and nature. And that which is made by ta’yīs, the modality of power over it is veiled from all knowledges. Whoever characterizes God by ‘power’ is [only] not ascribing to Him the features of incapacity.

In light of T1, Sijzī would not feel the force of the objection as stated because it (he might say) assumes a metaphysical framework that is theologically inadequate. In particular, the objection presupposes at least three things all of which are to be rejected:

1. that “power” is a property applicable to God and creatures in the same sense,

2. that creation is grounded in such capacity-property, and

3. that “having” or “exemplifying” a property is a condition that must be in place prior to creating anything

Re 1, Sijzī says that to describe God as “powerful” is only to mean to negate incapacity-making features from God. Accordingly, a term such as “power” does not pick out some positive attribute in God as it does in creatures. Rather, it functions as a mere negation. Hence, the claim “God must have the property of power” misfires in treating a purely semantic claim as if it is a metaphysical one.

Re 2, the objection assumes that what God does is to be understood in terms of a, or something like a, causal capacity. But the language of ibdāʿ (origination) is meant to break precisely that analogy. Origination is not “production” grounded in a capacity; as Sijzī says, it is not like craft or nature (ṭabīʿa) but something sui generis. The best way to bring to closer to understanding to say that it is a bringing into being out of total nothingness. Even with this, though, the “how” or modality of this relation remains a mystery (“not relatable to intelligence”). So the demand that God must already “have” a capacity in order to create/originate is, from Sizji’s perspective, an illegitimate projection of creaturely explanatory schemas onto what is beyond being altogether.

Re 3, Sijzī might say circularity only arises if “having power” is treated as some ‘thing’ (property) that, as such, must itself be brought into existence as one entity among others. Sijzī would deny that because for him everything that falls under “thing” (which would include both concreta and abstracta) is originated. And recall that talk of God’s power is just supposed to be a way of denying any incapacity-making features of God. As for the modality of what God does in itself, that, again, is beyond the entire framework in which “properties,” “capacities,” and “having” make sense. Thus, there is no question of God’s ‘having’ or ‘exemplifying’ the property of power because “property” as such belongs to the originated order.

Given these considerations, the bootstrapping objection for Sijzi would likely be a category mistake insofar as it attempts to force what is transcendent into the same conceptual space as the created realm.

As for the second horn of the objection, Sijzi wouldn’t say he’s denying, without qualification, that power is a property. Rather, he’s denying that power is a property in the divine case (as opposed to the creaturely case) for special reasons. Hence, universals, properties, etc., all belong to a realm that the Originator transcends. And within that realm, there a ways to make the relevant distinctions among created things. So the denial that “power” is a property in God then simply reinforces the sort of sharp divide Ismaili thinkers pushe for in general i.e., between (i) the originated domain, where intelligence (human and non-human) discerns attributes, properties, and relations, and (ii) the Originator, which is beyond all such categories.

In sum, Sijzī would say the circularity only arises if you assume that what God does (origination) must be explained in terms of the very conceptual machinery (properties, capacities) that itself (we have good reason to think) belongs to what is originated. Absent that assumption, the bootstrapping problem never gets off the ground - though at the cost of accepting that God’s relation to the world is, as he says, fundamentally a mystery i.e., opaque to intelligence.

For a brief characterization of Ismaili radical tanzīh and a justification for it, see an earlier post here.

Here is the result of my labors. Let p = corporealism i.e., the thesis that ‘God is a body’, and q = idolatry i.e., devotion to, assistance from, etc., intermediaries. And suppose both p and q are falsehoods.

The argument at Guide I.36, distilled, then looks like this:

1. If x is of a greater ontological rank than y, then a false belief about x is worse than a false belief about y

2. God’s essence is of a greater ontological rank than anything else

3. Thus, a false belief about God’s essence is worse than a false belief about anything else (from 1 and 2)

4. P pertains to God’s essence alone, while q pertains to something else besides

5. Thus, belief in p is worse than belief in q (from 3 and 4)

6. Of two false propositions, belief in the worse one is more blameworthy

7. Thus, belief in p is more blameworthy than belief in q (from 5 and 6)

In sum, Maimonides urges that corporealism is worse than idolatry because it misconceives the divine nature itself whereas idolatry just misdirects worship i.e., it does not necessarily entail a mistaken conception of God’s nature.

In On the rational soul, 271 (ed. al-Ahwani, tr. by yours truly), he states:

The felicity of the human soul comes about when its substance is rendered perfect, and this is accomplished when it is purified through knowledge of God and action for God. Its purification through action for God consists of (a) its being purged of vile and wicked qualities of character, (b) its being far removed from blameworthy attributes and evil and offensive habits by following reason and religious law, and (c) its being adorned with good habits, praiseworthy qualities of character, and excellent and pleasing traits by following reason and religious law. […]. The purification through action is accomplished by methods mentioned in books on ethics and by assiduous performance of religious duties, both legal and traditional, such as observances relating to the body, one’s property, and to a combination of the two. For being restrained at the places where religious law and its statutes place such restraints, and undertaking to submit to its commands, have a beneficial effect on rendering the soul that ‘incites to evil’ [Qur’an 12:53] obedient to the rational soul which is ‘at peace,’ [Quran 89:27] – I mean, making the bodily faculties of the soul, the appetitive and the irascible, subservient to the rational soul which is ‘at peace.’

Ramazan Karim!

In other words, Socrates is saying that thinking is a formative process that involves inner tension and sustained effort. It is not about arriving quickly at the right conclusions. The slow pain of mental labor is a welcome sign; for to struggle with a problem is already to be engaged in the kind of intellectual activity that makes insight possible. By contrast, a mind that feels no such strain is – at least in the ordinary case* - in danger of falling into complacency and unproductiveness.

Consider now the risk AI poses if we routinely outsource our thinking to it. When the hard work of formulating, analyzing, and struggling with ideas is regularly delegated to an external system, those “pains of labor” that Socrates treats as signs of intellectual fertility are bypassed. The result is a kind of cognitive barrenness: we obtain answers, yes, but without the inner process through which understanding is gestated. One has results, sure, but without performing those (formative) mental activities, said results are not really one’s own. Absent that struggle which cultivates the capacity to think, we’re deprived of what the Islamic philosophers call ‘the habitual intellect’ (‘aql bi’l-malaka). Thus, because one has ceased to undergo the inner travail that makes insightful thought and understanding possible, one ends up neither ‘conceiving’ (taṣawwur) well nor being able to produce unknowns from knowns in a healthy (read: valid) way.

In defense of AI, one might push back by saying something like: ‘rather than a contraceptive, AI is - like Socrates himself - more akin to a midwife i.e., in assisting minds to give birth to good ideas. At least, that seems possible. What speaks against it?’

What speaks against it is the fact that midwifery can only be practiced by those who, having birthed children of their own, are now past the child-bearing years. The barren or sterile - which AI is as it’s incapable of genuinely producing new ideas - cannot be “entrusted [with] the duties of midwifery” (149b); for machine nature, even more so than human nature, “is too weak to acquire skill where it has no experience” (149c).

The upshot: as a kind of contraception that risks making one sterile, it is on the fuqaha’ to reckon its use (in this manner) forbidden in the shar’.

*I say in the ordinary case because, according to the ḥukamā’, the prophets are an exception to the rule.

A theological consequence of the refutation is that the claim that one can literally point toward God - whether upward, outward, or in any spatial sense - is a mistake arising from being in thrall to one’s estimation (wahm), which tends to extend, without warrant, physical predicates to what is not a body. The Shaykh’s argument therefore undermines positions found, unsurprisingly, among ḥashwī theologians like Ibn Taymiyya but also, very surprisingly, among certain Aristotelian philosophers such as Averroes (from whom the former in fact inherited this assumption).

Avicenna Against Naturalism And Corporealism (ms Ayasofya 4853)

498KB ∙ PDF file

Download

Download

If so, then we might wonder: okay when you are not actively entertaining the P. Theorem, where is it (ontologically speaking)? Some of the options on the table here are: that it does not exist at all when not thought. Alternatively, that it exists lodged somewhere in the brain. A third view is that it exists outside the brain, in a domain of its own as a Platonic Form.

The ḥukamā’ (Avicennians in particular) rule all these out in favor of a fourth view: that it exists in a disembodied mind or intellect that serves as its storehouse. According to them, this separate mind preserves all real intelligibles – be they conceptions or propositions - that we are not actively thinking.

The theologians - I believe Ḥillī (13^th century) was the first - object to their argument by noting that the reasoning they advance for that thesis also applies to vain intelligibles e.g., false beliefs. If that is so, then that separate intellect will have to serve as the container of those as well. But this is inconsistent with their doctrine. For they hold that the separate intellect only thinks truths, and if it contains falsehoods as well then the truth-falsehood dichotomy is undermined.

In Revealing the intent i.e., of the Tajrīd, Ḥillī reports that when he raised this objection to Ṭūsī, the latter was unable to answer it. Perhaps this is so; perhaps not. I have been unable to locate any passage in Ṭūsī’s corpus in which the objection is explicitly addressed. In any case, another thirteenth-century Avicennian philosopher did address it explicitly - namely, Ibn Kammūna. Notably, Ibn Kammūna engaged in a number of philosophically fruitful correspondences with Ṭūsī. It is therefore plausible that his response to the objection was inspired, at least in part, by ideas communicated by Ṭūsī.

Ibn Kammūna answers the objection is a letter to Fakhr al-Din Kāshī, who wrote to him summarizing the issue as follows:

There’s no doubt that just as the soul judges by factual, correct, and universal propositions, it likewise has false, universal beliefs that it is sometimes inattentive to and sometimes forgets and returns to with a new acquisition or without one. Based on this, one of two things follows: [1] either the actual impression, in that [separate] existent, of the forms of those believed vain intelligibles, or [2] the existence of another existent that preserves them, which is neither a body nor bodily, nor a soul like it. Both options are false. As for the first, that is because of the absence [in that separate existent] of any judgement that does not correspond to reality (nafs al-amr) and because of a nonexistence of a difference between the true and the false due to their distinction from one another by correspondence and its failure. […].

That is, Kāshī is urging that an analogous - or, if you like, a parody - argument can be constructed for the false beliefs many people hold. In that case, we would have to affirm that that separate mind knows all sorts of absurdities, which is contrary to what the ḥukamā’ hold – namely, that it knows only what is objectively true. Further, if the separate intellect is supposed to be the domain of just the factual, but both true beliefs and false ones exist in it, then the distinction between truth and falsity is done away with. This latter is Kāshī’s second point in connection with the first reason (toward the end of the above passage).

In the response to both points, Ibn Kammūna evokes the opacity of belief contexts. He states:

It is known to the lofty principles - which are designated as ‘angels’, ‘immaterial intellects’, and the like - that ‘[p*] Zayd, for example, believes at such and such a time that p’ where p is some universal, false belief. This [p*] is a correct belief [on their part] that corresponds to reality, and contains or entails p. Thus, if Zayd cognizes p, then neglects it, and then returns to it, the recovery of p is only possible for him due to its being preserved in those immaterial entities in the way mentioned. And it does not follow from this that an assent that does not correspond to reality (which is called ‘compound ignorance) has been ascribed to those immaterial principles. On the contrary, they conceptualize that false assent [p] in a way that is consequent upon the knowledge that [p*] Zayd brings about that assent. This [response] is among the plausible rebutting defeaters to the mentioned objection.

Let me explain.

Suppose Zayd believes p (where p = some false proposition e.g., ‘the world is temporally originated’). Suppose further that you know that ‘Zayd believes p’. Call the latter proposition p*. So, Zayd believes p and you believe p*. Now, your belief that p* contains or implies p. However, from your believing that p* it doesn’t follow that you yourself also believe that p. That’s a non-sequitur.

More generally (and technically) put, then, belief is closed off neither under semantic containment nor under logical entailment. Another (perhaps simpler) example to illustrate the point: suppose it is true that

• “You believe that ‘Zayd believes ghouls exist’”

And that this belief is true. This belief of yours entails the truth of

• ‘Zayd believes that ghouls exist’.

But from that, it doesn’t follow that

• ‘You believe that ghouls exist’

Likewise, your belief that ‘Zayd believes ghouls exist’ contains the proposition ‘Zayd believes ghouls exist’. However, it does not follow that you too believe that ghouls exist.

In other words, taṣdīq (assent), for the ḥukamā’, does not transmit across. So if a knower, K, assents to p*, then although p occurs as part of the content of p*, it is not itself asserted by K (when K asserts p*). The embedded proposition (p) is only mentioned, not itself endorsed (by K). For taṣdīq is about what a knower has actually grasped and assented to, not about everything logically or semantically connected to it.

These failures of closure, however, do not apply to taṣawwur (conception), which is the mental state of grasping a meaning or content but without assent (affirmation) to it or dissent (negation) from it. Thus, if K conceptualizes p*, and p is a component of p*, then it’s the case that K conceptualizes p as well. In other words, for them, taṣawwur, unlike taṣdīq, is closed under conceptual containment.

Given the above, it is then possible for some K to, all at once,

(1) believe p*,
(2) conceptualize p, but
(3) fail to believe p, and in fact
(4) believe the opposite of p

If so, then let K = that separate mind or intellect. That, then, is how, according to Ibn Kammūna, false propositions may exist in the lofty principles, with the distinction between truth and falsity maintained. Thus, the objection is resolved.

One might resist the account as follows: ‘I can see how the above closures in taṣdīq cases apply to Ks that are finite e.g., due to limited attention or limited inferential capacity, etc. As for a K that is infinite, it’s not clear. It rather seems that if such a K believes q and q entails r, then K believes r as well.’

Response: for an infinite mind, there is doxastic closure under entailment. Granted. However, in the case at hand i.e., p* and p, that still would not apply. For this case is not a failure of inferential capacity. That is, it is not about missing an implication or failing to draw a conclusion. Rather, the difference is attitudinal (and whether K is finite or infinite makes no difference). The attitude-scope distinction marks the boundary of what a propositional attitude - such as a belief or a doubt - actually attaches to. When K believes q, the belief has q within its scope; so, for an infinite mind, any logical entailments of q remain inside that same scope and are consequently also believed. By contrast, when K believes p* - where p* is “Zayd believes p”- the belief takes scope over the ascription of a belief, a mental state, not over p itself; p itself lies outside the scope of K’s belief, even though it appears inside the content of what K believes. And because propositional attitudes attach to a whole proposition rather than to every embedded component of it, there is no transmission of belief from p* to p. In sum, doxastic closure operates only on what is believed as believed, whereas embedded propositions remain merely represented, which is why belief isn’t inherited across attitude embeddings.

In Questions and responses between Biruni and Avicenna, second issue, Ma’sumi states:

It is most astonishing that [on their view] He exists in eternity - there being no time, light, darkness, and creation - and then some notion occurs to Him and so He creates things; that after always being begrudging, He starts to bestow and so enacts, by an infinite power, a finite effect. Then, loss and corruption occur to it and it is then brought back again. Is this not just the same as pointless [action]?

Here’s a straight-forward (and I think forceful) Avicennian argument against their view:

1. If something, x, is added to some substrate, y, such that it makes y fall under some substance-kind, then x itself is a substance

2. A vegetative-soul, when added to the appropriate substrate, makes it fall under e.g., the substance-kind ‘oak tree’

3. Thus, vegetative-soul is a substance

Now run the same argument for animal and human souls. Thus, they too will be substances. In fact, the same argument is applicable to whatever they consider to be a substantial form. Thus, all substantial forms will be substances.

The text:

Translation:

If you would like to have [radical tanzīh] revealed to you in a manifest way, observe then with the eyes of your intellect whether, prior to origination, there existed something it is possible for you to grasp with your intellect, or relate the being of the Originator to, or ascend towards by some effort. If it’s not possible for you to do that - the ascent coming to a halt below Him so that there is no hope of linking His being to some item - you will be certain that all the entities supposed and imagined by the intellect are provided in full on creation [only] after origination, and that the Originator Himself transcends them all.

This is a stage beyond the stage of reason and isn’t reachable except by [spiritual] witnessings and disclosures, not intellectual discussions. And each person is facilitated for what they’ve been created.

But what does Jurjānī mean by its being ‘beyond reason’? And why should we believe the claim? He doesn’t say.

At least three readings present themselves. First, it is ‘beyond reason’ in that:

1. It is beyond reason to conceptualize the meaning of monism

Second,

2. It is beyond reason to establish a valid argument for monism

And third,

3. It is beyond reason to establish a sound argument for monism

Suppose 1 is meant. It’s easy to see why that reading can’t be right. For when in context Jurjānī raises the issue of monism’s legitimacy, he characterizes it as the view that

…though existence is identical to the Necessary and is non-partitionable and indivisible, it has spread over the ‘structures’ of the beings and so manifested in them, and hence none of the things are free from it - indeed, it is their reality and the same as them, [they] only being distinguished and multiplied due to qualifications and determinations that are perspectival (as exemplified by the ocean and its appearance in the forms of multiple waves which, despite that, there’s nothing here in reality except just the ocean).

If the above is the view, it’s clear that it’s something intelligible i.e., not literally nonsense. This is just to say we can conceive it - and no mystical ‘witnessing’ was required for that. So reading 1 is ruled out.

What about 2? It’s also easy to see why that reading is false or, at least, implausible. Just consider the argumentation some historical philosophers have advanced for that thesis. Here’s one example: Spinoza in the Ethics, part I. If you grant the premises he lays down there, the rational road to monism is shockingly brief: for if God exists as a substance with infinite attributes, and no two substances can share an attribute, it follows that no other substance besides God exists or can exist. The reasoning is valid and, again, no “mystical” insight is required to see that!

So 2 can’t be it either. What about reading 3?

Two points. First, that reading would have to presuppose that monism is true – the very point at issue between monists and non-monists. So if 3 was meant, then Jurjānī was evidently begging the question against his counterparts. If you don’t think 3 presupposes monism’s truth, though, then in a way the non-monists might’ve even accepted that reading and explained themselves as follows: ‘yes, 3 is true – precisely because it is not logically possible to give a sound argument for what is false and monism is false!

Second, some monist thinkers e.g., Spinoza himself, would’ve definitely disagreed with reading 3 – for he obviously took the argument he was advancing in the Ethics to be sound. And further, even granting that 3 is the right reading, why should be believe it’s true? As noted, Jurjānī doesn’t say (in context).

Apart from readings 1-3, the meaning of ‘beyond reason’ remains beyond my ken.

Here’s an example i.e., the way Feser distinguishes between ‘natural philosophy’ and ‘natural science’. For he claims that though both are concerned with “only the empirical and material reality” (p. 4), the difference between them is that natural science is “concerned with the empirical and material world that happens as a matter of contingent fact to exist” (ibid.), whereas the natural philosophy “is not so confined” because “it is concerned with what any possible empirical and material world would have to be like” i.e., such that we can have scientific knowledge of it. With all this, he is suggesting that natural philosophy is broader, more fundamental, and deals with necessities the contingent findings of the empirical sciences don’t concern.

Critique

The above account appears muddled and unstable. I believe Feser’s proposed distinction collapses under Aristotelian principles. For according to these principles, “scientific knowledge” (episteme/‘ilm) is knowledge of what is necessary, not contingent. In other words, for something to count as scientific, it must be universally and necessarily true, not merely true of a particular configuration of the world i.e., the one that happens to exist.

If we take this notion of science seriously, then any true proposition from natural science - whether familiar, e.g., “salt dissolves in water,” or less familiar, e.g., “acids neutralize bases” - must be necessary; that is, true in any possible empirical and material world. Otherwise, it would be contingent, and thus not scientific in the Aristotelian sense. But if such propositions are necessary, then, by Feser’s own criterion, they would not be confined to “this empirical world”; they would apply to any possible empirical and material world in which those things exist. Consequently, these propositions would fall under natural philosophy in his sense. Thus, if natural philosophy studies the necessary features of any material world and natural science provides necessary truths about the material world, the two domains overlap completely in content. The upshot: the proposed distinction holds no more.

To preserve it, he might pushback by saying something like: ‘natural philosophy deals with metaphysical necessity, while natural science deals with nomological necessity (e.g., chemical laws and the like).’

If so, the response to that would be: the Aristotelian philosophy he claims is back with a vengeance undermines this very distinction too. For that philosophy grounds necessity in the natures of things i.e., it states things behave as they do by necessity, given what they are. So the solubility of salt in water, for instance, follows necessarily from the natures of salt and water. It is not a contingent feature of this particular world; in other words, in any possible world in which salt and water exist, salt will dissolve in water - assuming no impediments, of course. Hence, chemical propositions such as “salt dissolves in water” or “acids neutralize bases” turn out to be as necessary, within an Aristotelian framework, as the natural philosophical proposition “bodies are composed of matter and form.” All of them express truths that hold in any possible empirical and material world, since all depend on the natures of material substances as such. The appeal to metaphysical and nomological necessity is hence of no aid here.

In sum, Feser cannot maintain that natural philosophy uniquely concerns what “any possible empirical world would have to be like.” For, as we’ve just learned, on the Aristotelian view all genuine scientific knowledge – be it chemical, biological, or physical - is likewise about what any material world, constituted by such natures, must be like. Therefore, the proposed distinction between natural philosophy and natural science is unsustainable within the Aristotelian perspective Feser wishes to preserve. Either natural science, too, yields necessary knowledge of the material world and hence falls under natural philosophy, or it does not yield necessary knowledge and thereby ceases to count as science in the Aristotelian sense.

The answer he is given is instructive and revealing of how the function of divine discourse was generally conceptualized in the hikma tradition. Sijistani basically states that the Law addresses a broader audience than philosophy. Consequently, its speech must vary in mode and style e.g., being sometimes clear and reasoned, sometimes indirect and allusive, sometimes expressed through metaphor, and other times argumentatively. In this way, it speaks both to the intellect and to the imagination, guiding each type of person according to their capacity. When its purpose is understood along these lines, one sees that revelation and philosophy are not in conflict: what the Law gives in imagistic form, the philosopher reads more abstractly ; what satisfies the philosopher as indication, suffices the layperson as expression.

One day I asked Abu Sulaiman: Why wasn’t tawḥīd purified in the Law (sharīʿa) from the dross of conjectures and imagistic expressions as it was purified in that way in philosophy? […].

He said in response: We have said many times in previous discussions, and in the course of ideas that have presented themselves and become known, that speech intended to cultivate the common people and unify everyone must at times be expansive and at other times concise; at times explicit and clear, and at other times general, indirect, and allusive; at times couched in metaphor and simile, and at other times qualified by arguments and reasons. It must take various forms, which need not be detailed if the intended meaning becomes evident in the course of its presentation or in the midst of it. Once this is firmly understood and clearly shown, it becomes evident that everything the Law delineates in this manner is such that the specialist finds in it an indication that satisfies him, and the layperson an expression that suffices him.

Let me briefly state his formulation of the siddiqin, and then unpack it with a view to motivating the question he poses.

Dashtaki Sr.’s statement of the argument is brief and to the point:

If there were not, in existence, that which is intrinsically necessary of existence, there wouldn’t be any existent at all. But the consequent is manifestly false.

The implicit conclusion: therefore, so is the antecedent. Positively put, this is just to say: there exists some intrinsically necessary existent.

So the bare bones of the argument, then, is a modus tollens syllogism like so:

If there were no necessary existent, then nothing would exist at all
But something does exist
Thus, there exists a necessary existent

The truth of the minor premise is blindingly obvious; not even the most extreme of sophists could deny it. That is, the statement "something exists" is an indubitable fact that cannot be rationally gainsaid. For even the sophist, engaging in clever but fallacious reasoning as he does, must exist in order to make the claim that nothing exists. But this very act of uttering or thinking the denial "nothing exists" proves the existence of the denier, which in turn proves that something exists. So the purported denial of the minor leads to a performative contradiction, where the content of a statement is contradicted by the act of making the statement itself. In sum, the claim is self-refuting because the person making the claim is a counterexample to it.

As for the major premise, it's not obvious and so, for that reason, Dashtaki Sr., devotes most of ch. 1 to its justification and defense. The key insight of the justification is the following stretch of text:

Justification of the entailment: the existence of the contingent depends on a causal factor that existentiates. Thus, it must necessarily exist, since as long as a thing does not exist it will not make [something else] exist. Now on the assumption of the absence of the necessary being, that causal factor will also be contingent and dependent on another contingent causal factor. Hence, the extremity will have the status of the intermediary in the need for a causal factor. And as long as [the situation] remains like that, then from where does a contingent obtain [at all] such that another contingent can obtain from it?

To understand the considerations he advances in its favor, we’ve to first lay down some premises.

Suppose we have a causal series ABC, where A is the cause of B which in turn is the cause of C. A and C are said to be ‘the extremities’ of the series, while B is said to be ‘the intermediary.’ In this causal scenario, each member has a unique property: C is only an effect, A is only a cause, and B is both a cause (e.g., of C) and an effect (e.g., of A). So far so good.

Now suppose we extend this series by adding one more member to it, specifically alongside A, so that we then have: DABC. In this situation, the extremes would now be D and C, while the intermediaries would now be A and B. Thus, adding that additional causal member gave A the status of B i.e., as an intermediary. The same would again be the case if we add a further causal member e.g., E, alongside D. And so on, mutatis mutandis, for any new causal member we add on that side.

Furthermore, with regard to those members, we can ask three distinct questions:

1. Why does this member (e.g. C) exist?

2. Why do all the members ABC exist?

3. Why does any member at all exist?

In 1), the focus is an individual within the chain. The kind of explanation sought is the prior proximate cause e.g., “C exists because B caused it, and B because A.” There’s a limit to this explanation in that this type of question can always be deflected backwards by appeal to another prior cause.

In 2), the focus is the chain itself understood as just all of the members i.e., together. The kind of explanation sought is the cause of the chain taken all together. If it as such could have failed to be, then even if each single member of the chain is explained by the immediately prior member, the entire chain (= all of the members) remains unexplained. And this is so irrespective of whether the chain is finite or not.

In 3), the focus is not this or that member, and not all of the members as a collection, but the very fact of there being any member of that kind at all. The sort of explanation sought is the cause or ground for the possibility of there being at least one member in reality rather than none. Appealing to another member or to all the members together won’t answer this question because the question is not “why this one?” nor “why all of them?” but rather “why any at all?” To say “this member exists because it is due to that member” does not answer the question “why does any member exist at all such that another member is due to it?”

If the above is clear to you, we can proceed to parse the support Dashtaki Sr. offers to the major premise in the text above. He thinking seems to be this:

If, per the initial assumption, there’s no intrinsically necessary existent, and given that there are existents, it follows that the only existents there are are contingent ones (i.e., beings of the sort that may or may not exist). Now for any contingent existent, there exists a cause which makes it exist. This is obvious; for nothing can existentiate another unless it itself first exists. Dashtaki Sr. here is presupposing (a version of) the PSR or tarjih bi la murajjih. In the present scenario, where all and only contingents exist, the cause of any contingent will have to be just another contingent, which itself will have been made to exist by yet a further contingent, and so on without end. This implies that each successive cause in the series of contingents will have the property of being an intermediary: a cause of one thing and an effect of another. If so, then Dashtaki Sr. poses the following crucial question: why is there any contingent, as it were, ‘up and running’ to begin with such that another contingent is explicable by it? What explains that?

Appealing to some further contingent being just will not do. For, as noted, it’s one thing to ask, “where does this one come from?” or “why does this particular contingent exist?” and another to ask, “where does any one at all come from?” or “why does a contingent exist at all in the first place?”. The first question can be answered by appealing to a particular contingent prior to the one being asked about. But the second question cannot be answered in the same manner. And note that the question Dashtaki Sr. is posing is neither an intra-series explanatory question (type (1) above) nor, as Avicenna himself had done, a meta-series explanatory question (type (2) above) about the explanation for the entire (infinite) contingent series. Nor, moreover, is Dashtaki Sr. saying that any intra-series explanatory account will fail to be sufficient or even adequate to answer the meta-series question (even though that’s true). Rather, what the Shirazi philosopher is asking about (at the end of the passage) is the very fact of contingent existence. In other words, he’s asking a question of type (3): why does it obtain at all i.e., in that some contingent being or other exists such that it serves as the explanation of another contingent? This is not your standard “series-of-contingents” objection but a more fundamental modal-explanatory question i.e., “why does any contingent exist at all, even one?”

If so, then Dashtaki Sr. is not taking the whole regress itself as the problem, but probing something deeper: from whence did any contingent entity come to be in the first place? That “first place” question is different in kind; unlike “why does this contingent exist?”, which can get answered by “because of this other contingent”, asking “why does any contingent exist at all?” cannot be answered in that way, since that move just cites the very sort of thing for which an explanation is being sought - namely, the fact that there’s a contingent being at all. Compare: why is this candle lit? Answer: because this one ignited it. Okay, but why is there any lit candle at all? Here, appealing to just another lit candle fails as an answer, since any particular lit candle one invokes will be among the sort of things for which an explanation is being sought. Hence, the explanatory gap Dashtaki Sr. is pointing to doesn’t concern any particular member of the series, nor the series as a whole, but contingent existence as such - that is, how contingent being as such gets instantiated at all. The force of his question lies in precisely that - namely, why, if there’s no necessary being, does the category of ‘contingent being’ have any instance at all i.e., such that other instances can come about from it?

At this point, the opponent only has three options before him to answer this more fundamental question: either (A) accept there is an explanation by appealing to a non-contingent i.e., necessary, being, which is just the conclusion sought, or (B) reject that there’s an explanation in saying something like ‘it is just a brute fact that contingent being is instantiated at all’. This amounts to a denial of the PSR. Or, (C) offer an explanation in terms of the truth of the following proposition:

Necessarily, some contingent exists

That is, (C) answers the question “why is there any contingent at all?” by saying: “because there must be some contingent”.

If so, the (C)-response fails just as much as the (B) one, and for at least two reasons. First, that proposition by itself is consistent with there being a necessary being which, of necessity, existentiates some contingent. If so, then (C) is not an alternative explanation to positing an intrinsically necessary being i.e., (A).

And second, you know for certain that of each contingent it is true that it may fail to exist. If so and if, per our assumption, there’s no intrinsically necessary being, then it follows that all of them may fail to exist. For there’s nothing that guarantees that, necessarily, at least one should exist. Compare: imagine a universe with only (Mawlid) lightbulbs, all contingent, and no power grid that necessarily keeps at least one on. Since each bulb could go out, and nothing guarantees that at least one must stay on, it follows that:

‘Possibly, they all go out’

But if all of the contingents may fail to be, then it’s false that:

‘Necessarily, some contingent exists’

Hence, on analysis it turns out that (C) and ‘all and only contingents exist’ form an inconsistent dyad.

Now given that some contingent does in fact exist, why does it exist at all if it could have failed to? We’re right back to (B): wild bruteness. Option (A) is the only rational alternative.