AQUINAS AND ISLAMIC AND JEWISH THINKERS

St. Thomas Aquinas

I. AQUINAS’S ATTITUDES TOWARD AVICENNA, MAIMONIDES, AND AVERROES

The work of Thomas Aquinas may be distinguished from that of many of his contemporaries by his attention to the writings of Moses Maimonides (1135–1204), a Jew, and Ibn Sina (Avicenna) (980–1037), a Muslim. His contemporaries, especially in Paris, were responsive to the work of another Muslim, Ibn Rushd (Averroes) (1126–1198), for his rendition of the philosophical achievements of Aristotle, but Aquinas’s relation to Averroes and to those who took their lead from him was far more ambivalent. Aquinas respected Rabbi Moses and Avicenna as fellow travelers in an arduous intellectual attempt to reconcile the horizons of philosophers of ancient Greece, notably Aristotle, with those reflecting a revelation originating in ancient Israel, articulated initially in the divinely inspired writings of Moses. So while Aquinas would consult “the Commentator” (Averroes) on matters of interpretation of the texts of Aristotle, that very aphorism suggested the limits of his reliance on the philosophical writings of Averroes, the qadi from Cordova. With Maimonides and Avicenna his relationship was more akin to that among interlocutors, and especially so with Rabbi Moses, whose extended dialectical conversation with his student Joseph in his Guide of the Perplexed closely matched Aquinas’s own project: that of using philosophical inquiry to articulate one’s received faith, and in the process extending the horizons of that inquiry to include topics unsuspected by those lacking in divine revelation.

We may wonder at Aquinas’s welcoming assistance from Jewish and Muslim quarters, especially when we reflect on the character of his times: the popular response to the call to arms of the crusades as well as a nearly universal impression on the part of Christians that the new convenant had effectively eclipsed the old. Aquinas may have shared these sentiments, for all we know, yet his overriding concern in reaching out to other thinkers was always to learn from them in his search for the truth of the matters at hand. In this respect, he epitomized the medieval respect for learning, with its conviction that “truth was where one found it.” So he was more inclined to examine the arguments of thinkers than their faith, trusting in the image of the creator in us all to search out traces of the divine handiwork, a theological premise that will prove useful in guiding our explorations into Aquinas’s reliance on Jewish and Christian thinkers, and better than attributing to him an ecumenical or interfaith perspective avant la lettre. Yet it would not be untoward for us to note how other thinkers attempting to employ the inherited philosophy to elaborate their faith-perspective were for that very reason helpful to Aquinas in his vocational task.

It is worth speculating whether the perspective of Aquinas and his contemporaries was not less Eurocentric than our own. What we call “the West” was indeed geopolitically surrounded by Islam, which sat astride the lucrative trade routes to “the East.” Moreover, the cultural heritage embodied in notable achievements in medicine, mathematics, astronomy, as well as the logic, philosophical commentary, translation, and original work in metaphysics begun in tenth-century Baghdad, represented a legacy coveted by western medieval thinkers.1 Marshall Hodgson has called the culture that informed this epoch and extended from India to Andalusia “the Islamicate,” intending thereby to include within its scope Jewish thinkers like Maimonides who enjoyed the protected status of dhimmi and contributed to Muslim civilization.2 Christians like John of Damascus enjoyed a similar status, reserved by Qur’anic authority for “people of the book,” yet the divisions in Christendom saw to it that thinkers in Paris were better acquainted with Muslim and Jewish writers than with their co-religionists in Islamic regions.

Aquinas’s own geographic and social origins could well have predisposed him to a closer relationship with thinkers representative of the Islamicate than his contemporaries could be presumed to have had, in Paris at least. For his provenance from Aquino in the region of Naples, itself part of the kingdom of Sicily, reflected a face of Europe turned to the Islamicate, as evidenced in the first translations commissioned from Arabic: “Latin, Muslim, and Jewish culture mingled freely in Sicily in a unique way that was peculiarly Sicilian.”3 Moreover, in his later years, when his Dominican province asked him to direct a theological studium, Aquinas expressly chose Naples (over Rome or Orvieto) for its location, and that for intellectual reasons: “there was a vitality about Naples that was absent from Rome or any other city in the Roman Province.”4 So it might be surmised that these dimensions of his own personal history led him to be more open to thinkers from the Islamicate than his co-workers from Cologne or Paris might have been. In any case, the number and centrality of his citations from Avicenna and from Moses Maimonides leave no doubt as to their place in his intellectual development. By styling that place as one of “interlocutor” I have tried to finesse the vague historical category of influence in favor of one more familiar to philosophers and theologians of every age, and especially of those consciously working in a tradition of inquiry, who treasure what they learn as a result of contending with their predecessors’ arguments, even when their interlocutors lie beyond the reach of actual conversation.

II. AVICENNA: THE DISTINCTION OF EXISTING FROM ESSENCE

In his early monograph 'De ente et essentia,' composed near the age of thirty when he became Master of Theology at Paris, Aquinas displayed a rare metaphysical acumen in preparing the way for using the philosophy of Aristotle to elucidate a universe created by a sovereign God.5 Presenting a lexicon of key philosophical expressions – ens (“a being”), essence taken in itself and in its relation to genus, species, and differentia, as well as essence in separate substances and in accidents – Aquinas takes the opportunity to introduce a new level of “composition” in created things beyond that established by Aristotle of matter and form. His guide here is Avicenna, whose notion of “essence in itself” gave him the key premise in the argument to a new level of composition: “every essence or quiddity can be understood without knowing anything about its existing (esse)” (DEE 4.6). This fact is utilized as a sign that “existing itself cannot be caused by the form or quiddity of the thing” (4.7), which then “must be potential with regard to the existing it receives from God, and this existing is received as an actuality” (4.8). As form is actualization with respect to matter for Aristotle, so existing will be with respect to essence for Aquinas. The import of this famous “distinction” in his overall project will not appear until his treatment of creation, where existing will be identified as “the proper effect of the first and most universal cause, which is God” (ST Ia.45.5), but its role in Aquinas’s reception of Aristotelian metaphysics has been comprehensively canvassed by Edward Booth.6 To appreciate the part that Ibn Sina played in the unfolding of that drama, let us try to understand its terms and something of the cast of players confronting Aquinas in his task of adapting a classical ontology to articulate a freely created universe.

The persistent problem bequeathed to posterity from Aristotle’s Metaphysics concerns the relationship of existing individuals to their “intelligible natures” (or rationes, as the medievals identified them). It is clear that Aristotle meant substance to be exemplified paradigmatically by existing individuals, yet equally clear that “what makes something to be what it is,” its essence (or “secondary substance”), comprises what is knowable about it. Which of the two has ontological primacy, and why? It is fair to say that the Metaphysics left this as a radical problem (or aporia), as Booth’s assemblage of commentators on that seminal text will testify. What focused an otherwise abstruse metaphysical issue, however, was the avowal of a created universe. In Charles Kahn’s admirable summary:

"existence in the modern sense becomes a central concept in philosophy only in the period when Greek ontology is radically revised in the light of a metaphysics of creation. … As far as I can see, [the early Christian theologians] remained under the sway of classical ontology. The new metaphysics seems to have taken shape in Islamic philosophy, in the form of a radical distinction between necessary and contingent existence: between the existence of God, on the one hand, and that of the created world on the other."7

The “sway of classical ontology” was confirmed and stamped by three figures spanning the third to the fifth centuries: Plotinus (205–270), his pupil and publicist Porphyry (232–305), and Proclus (410–485). (What is more, two books attributed to Aristotle and vastly influential among Arab and western thinkers – The Theology of Aristotle and Liber de causis – were in fact editions of Plotinus and Proclus, respectively.) Their neoplatonic tendencies neatly reversed the primacy of “first/second substance” in Aristotle, as they yoked ontology to logic with the less general serving the more universal. The most inclusive category of all will then be being (eis to einai), itself an emanation (for Plotinus) from the One. Of immense fascination to religious minds, whether pagan, Jewish, or Christian (and later Muslim), this systemic explanation of all things by an emanation that turns the increasing generality of substantial predication into a causal efflux can hardly be said to reproduce Aristotle’s focal concern for individuals. The publication of the “Theology” under his name can only have been an act of pious deference, whose effect was to create the hybrid that the Arabs later encountered as “philosophy.”

The western witness to the ensuing attempt to contain Aristotle within a neoplatonic scheme of emanation, while deferring to his concern for individuals, was Boethius (c. 480–524). His logical works tend to reproduce the Porphyrian tree in a manner reminiscent of Proclus (Liber de causis), yet he also comments on Porphyry’s hesitation regarding the status of universals by considering them to be abstracted from experience as a means of giving experience an intelligible form.8 In general, however, it seems that Boethius avoided judging “between Plato’s separate ideas and Aristotle’s universals,”9 utilizing the realist conception of universals prior to things (ante res) when needing to express their containing priority, and the conceptualist view of them as dependent on things (post res) when deferring to Aristotle’s insistence. When he does bring them together, it is to assert that an individual subject can be taken at once particularly and universally, although the being (esse) is clearly that of the subject. Aquinas commented on two of Boethius’s works that exhibit a greater affinity to pseudo-Dionysius’s monotheistic correction of Plotinus and Proclus: De trinitate and De hebdomadibus. In the latter Boethius identifies God with ipsum esse, carefully distinguishing “between the esse which makes God ipsum esse, and the ipsum esse of things which flow from him.”10 It is a notion on which Aquinas capitalized, but only after clearly discriminating esse from essentia, as we have seen.

Aristotle’s central aporia will not admit of resolution, then, and even returns to threaten the urge to reduce the tension between species and individual by subsuming both in a larger emanation scheme. One maneuver, however, had not yet been attempted: distinguishing what constitutes the individual, namely its existing, from what makes it to be the kind of thing it is. As Kahn suggests, this move does not emerge until the Islamic philosophers, arguably first with al-Farabi (?870–?950), and clearly (though not yet coherently) with Ibn Sina. And the pressure to do so comes from the need to distinguish the “first being” (in al-Farabi) from all that is not first and derives from it. While not yet a coherent notion of creation, the concern to make a clear hiatus in the emanation scheme that he adopted made al-Farabi separate “a principle which has no essence as apart from existing (huwiyya) [from everything else that] must have [its existing] from something else” – namely the principle.11 What will be required to keep the principle from being identified simply with the first in a logical scheme-in short, to secure a notion of creation – will be a way of clearly distinguishing existing from essence (mahiyya). So we are brought to Ibn Sina’s wrestling with that task. Although it does not appear as clearly in him as it does later in Aquinas, that same distinction will allow one to overcome Aristotle’s central aporia. Through a notion of creation, the difference of creator from creation will also mark what distinguishes the individual existent from its essential explanations. But that is to anticipate the story’s final point.

Ibn Sina’s discussions of existing (mawjûd or anniyya) as distinct from essence (mahiyya) are all in the context of distinguishing necessary being (wajîb al-wujûd) from possible being (mumkin al-wujûd).12 And the consideration of universals-in-themselves, which might be said to prepare the way for the distinction, reminds us that “the providence of God accounts for something’s being in so far as it is an animal.”13 Here he has in view the essence – the haqîqa – usually rendered by the specific term “man” or “animal.” Ibn Sina, in short, is less preoccupied with Aristotle’s quandary regarding the proper way to characterize existing individuals so as to secure their exemplary status, than he is concerned to find a way of characterizing essences so that their existence in things may properly be explained.

But that does not mean that such essences can exist apart; explicitly not, in fact (5.1, 204:14–17). As the essence of what may possibly exist, something other than itself must explain this animal’s existing. For the essence as such is neither universal nor particular, one nor many; all it can explain is the thing’s being an animal. (And, as we have seen, that is all that Aristotle seemed directly concerned to account for.) As for the individual animal’s coming to be and passing away, as well as its continuing to exist as long as it does, it is this fact which, Ibn Sina insists, cannot be accounted for by the essence itself. Why not? Because all essences are essences of possible beings, and the “proper character [of such beings] is that they necessarily require some other thing to make them be actually (bil-fi’l mawjûdan)” (1.8, 47:10–11). There is only one whose existence is necessary, and that one, “the first, has no essence (mahiyya) except its existence (anniyya)” (8.4, 344:10). “Necessary being has no essence (mahiyya) except that it be necessary being, and this is its existence (anniyya)” (8.4, 346:11).

By insisting that the necessary being’s essence (dhat) can be characterized only by its very existing (anniyya), Ibn Sina wants to avoid a misunderstanding that could jeopardize his entire enterprise: taking existence (wujûd) as a property contingently held by everything but the first being, who possesses it necessarily.14 Such a reading would jeopardize his entire project, for it would make the distinction of necessary from possible being explicable by an independent understanding of modalities. (It would also require understanding wujud as a property, a point that will emerge for comment.) Ibn Sina seeks rather for an independent way of characterizing “the first,” which will then clarify his use of necessary/possible being. That is to present it as “sheer being, with the condition of negating anything understood as [adding] properties to it” (8.4, 347:10). The result is that such a one alone is utterly without potentiality, and “a unity, while everything else is a composite duality” (1.7, 47:18).

The statement just cited actually uses the ordinary Arabic word fard, or “individual,” but it is better translated “unity” here since the entire chapter is concerned to show the radical unity attending necessary being. In the process of so distinguishing necessary from possible being, Ibn Sina succeeds in identifying a new mode of composition in everything that is not necessary. It is a “composite duality” – not that of matter and form, which he presumes throughout, but one of essence (mahiyya) and of some other factor that causes the individual thing to be. That factor is never identified as such, although it would be tempting to identify it as anniyya. The pair mahiyya/anniyya would then sound like essence/existence. Yet that factor is never isolated; anniyya expresses “the real existence of a particular individual” rather than identifying what it is that makes the individual exist.15

Moreover, the term that Ibn Sina consistently prefers, mawjûd, which is the participial form of wujûd, is probably best rendered “existing,” as we do when we look for a participial form for “being” – once that term has been fixed as a noun! So we say that a being is an existing thing. Yet we have not thereby isolated a distinct factor, existence, which is why I have usually rendered wujûd as “being” rather than “existence” and mawjûd as “existing,” saving “existence” (or “very existing”) for anniyya.l6 Yet even here, following Frank, “existent” would render the usage more accurately. So once again, existence eludes us, yet we are on the track. It will continue to elude us in Ibn Sina, for he begins, as Anawati notes, “with essence in such a way as to arrive at the existing (esse) which affects it as though it were an accident.”17 It is in fact his treatment of the universal-in-itself that affords him the leverage to consider being (wujûd) as something that “comes to” the essence, while also guaranteeing that it not be considered as an accident properly so-called, that is, a property. His discussion (5.1) quickly leaves behind the general term “universal” (kullî) and concentrates on man or animal: “animal insofar as it is animal, and man insofar as it is man, that is in terms of their definition and meaning, without reference to other things accompanying them – nothing but man or animal” (5.1, 201:1–3). One cannot help but find this a congenial rendering of Aristotle’s “secondary substance,” or the formula. Universality, or predicability of many, belongs to the essence only upon further reflection regarding its role in discourse; hence it is an accompanying feature of “animal as animal.”18

What Ibn Sina is reaching for is an essence prior to universality or particularity with no conditions, not even, he insists, one with the expressed condition not to attribute particularity or universality to it (5.1, 203:18). It is the essence taken by itself, without regard to existence, and hence short of the Platonic status of separateness. Such a one, he avers, can and indeed does “exist in reality,” while the Platonic one – considered as separate – can exist only in the mind (5.1, 204:5–10). But how can we say it “exists in reality” if not separately? The Latin translation, which formed the basis for western interpretation of Avicenna, answers that unequivocally by translating “in reality (fi’l-a’yân) [as] in sensibilibus (in things which can be sensed).”19 Such a rendering would leave no doubt as to Ibn Sina’s Aristotelian commitments, and it is as plausible as any in rendering the Arabic expression fi’l-a’yân, which carries the original meaning of “upon observation.” What is essential, after all, is that we arrest our consideration at the essence-itself – “animal insofar as animal” – which, if it were to exist, would exist in particulars. Yet the strategy of stopping short of that is to show that in itself no such essence can explain the presence of animals.

If we ask why it cannot do so, the question fairly answers itself. For the essence of all that is not necessary being is itself indifferent to existence or to non-existence; indeed, that is what it is to be possible being (1.6, 38:12–17). To have no cause is not to exist, and to exist such an essence “demands another thing which will make it be in actuality” (1.7, 47:12). There is no further question remaining beyond that implicitly put by Anawati: why select such a starting point? Nor can we expect Ibn Sina to answer that question; the best we can do is point to the neoplatonic manner of resolving Aristotle’s quandary, and note the predilection of that tradition (and of much of philosophy) to focus on the formula itself.20

Standing in such a tradition, yet unwilling to give ontological primacy to what is more general, Ibn Sina sought a reason for giving primacy to existing individuals. Although the Aristotelian aporia did not structure his inquiry, it could not help but motivate it. Since that reason could not come from the formal side, it had to come from elsewhere. With matter a mere repository of possibility, it could come only from “the first” being whose very essence would be to exist. The image that comes to mind is of the Copernican system before Newton. As Bellarmine rightly saw, it remained a likely mathematical story without an account of the origin of movement.21 The Plotinian emanation scheme remained a logico-aesthetic theory without an ontologico-kinetic source. Aristotle’s prime mover accounted for the activity of the spheres governing generation and corruption; Ibn Sina’s “first being” would account for the scheme’s actually existing. No wonder Kahn insisted on the newness of this “notion of radical contingency, not simply the old Aristotelian idea that things might have been other than they in fact are …, but that the whole world of nature might not have been created at all: that it might not have existed.”22

The prospect of a metaphysical rendition of that new situation must have directed Aquinas to Avicenna’s “distinction” and led him to transform it the way he did. For Ibn Sina did not himself succeed in formulating a notion of creation corresponding to so radical a contingency, any more than he was able to identify what it was that united with essence to yield the composite dualities called substances.23 Whatever it was, it had to “happen to” or “come to” essence or possible being (1.7, 47:12). And since the Arabic verb for “happen/come to,” like the Latin accidere, in its noun form had translated Aristotle’s “accident,” Ibn Sina was said to have made of existing an accident. Kahn describes the new situation neatly: “for the contingent being of the created world (which was originally present only as a ‘possibility’ in the divine mind) the property of ‘real existence’ emerges as a new attribute or ‘accident’, a kind of added benefit bestowed by God upon possible being in the act of creation.”24

It requires no exceptional philosophical acumen to show that existing cannot in any proper sense be an accident. For the grammar of that category – “what exists in another” – presupposes primary existents of which it can be an accident. If existing is taken to be that which enters into composition with essence to make a primary existent, then it could not itself be of such a sort as to presuppose itself. And if the contrast term for “existing” is not “substance” but “essence taken by itself,” then Ibn Sina could well say that existing must come to such an item for it to exist as an individual, but would have no right to call what “came to” it an accident of it. Ibn Rushd (Averroes) belabored this point, intending it as a criticism of Ibn Sina; in our time Fazlur Rahman and Alexander Altmann have cleared the record.25 Yet it took Aquinas’s radically new metaphysical step in his early De ente (DEE) to fashion an adequate response by removing existing (esse) from the entire slate of Aristotelian categories, proposing that it be understood in terms of the master-analogy of actuality/potentiality. Its formal ontological status will have to await his treatment of creation, where he identifies the “esse received from God” (DEE 4.8) as “a relation to the creator as the origin of its existence” (ST Ia.45.3). Relation-to-a-transcendent-agent is the only possible way one can identify the act within each thing that is the expression of the activity whereby God “produces a thing without motion”: creation. In one fell swoop, Aquinas has succeeded in restoring the primacy Aristotle intended for individual existing things, by linking them directly to their creator and by granting Avicenna’s “distinction” an unequivocal ontological status. Yet as should be clear, this is more than a development of Avicenna; it is a fresh start requiring a conception of existing that could no longer be confused with an accident, and which has the capacity to link each creature to the gratuitous activity of a free creator. Only in such a way can the radical newness (hudûth) of the created universe find coherent expression, for the existing “received from God” will be the source of all perfections and need not presume anything at all – be it matter or “possibles.”

III. MAIMONIDES: STRATEGIES OF CONCILIATION

If Avicenna gave the impetus to Aquinas’s project of adapting the received philosophy of the Greeks to the larger project of elaborating a universe created and redeemed by the one God, it is fair to say that Moses Maimonides gave that project its critical shape. For the assistance that Aquinas most needed had to do with the respective criteria of reason and of revelation, and Maimonides’s ongoing conversation with his student Joseph focused precisely on this point: the interaction of reason and of revelation in determining what one might responsibly hold.26 In Aquinas’s milieu, the translation of the Guide of the Perplexed must have been a boon, for the goal of his own project was questioned from two sides: the conservative Augustinians, who pretended to be invoking a pure tradition of faith against the “new learning,” and the Latin Averroists, who were so enamored of Aristotle as to make of his teaching a virtual revelation for the philosophically minded.27 Although Maimonides wrote his dialectical inquiry with Joseph in Judaeo-Arabic, it was translated into Hebrew in 1204, and thence into Latin by the 1220s. As Louis Gardet has shown, Aquinas gleaned all that he knew about the Muslim “theologians,” the mutakallimûn, from the expositions available to him from the Guide.28 But we are less concerned about information than strategy. Taking the central issue of creation as our focus, we will be able to see how Aquinas took his cue from Rabbi Moses precisely in the delicate domain of reconciling the deliverances of revelation with conclusions of reason. This will allow us to trace this tutelage in three critical areas: the “eternity” or temporal limitedness of the cosmos, the meaning-structure of divine names, and God’s knowledge of singulars and the range of providence.

A beginningless world? Creation and “the distinction”

Since the free creation of the universe marked the divide separating medievals from the ancients, the task of reconciling biblical faith with Greek metaphysics found its natural focus there. The role of intermediaries is also crucial here, since the pagan thinker Plotinus provided the philosophers of a fledging Islamic tradition with what looked like a promising pattern for articulating creation: the scheme of necessary emanation modeled on logical deduction. This allowed “al-Farabi and Avicenna [to] reconstruct the traditional notion of creation ex nihilo in terms of a Plotinian metaphysics that, given its compatibility with Aristotle’s physics, they attribute or read into Aristotle’s metaphysics. Thus for al-Farabi and Avicenna the eternal creation theory is the theory of Aristotle, although in its historical development Plotinus and Proclus are its real progenitors.”29 And the same can be said for Maimonides, who catalogues his “third theory,” that of eternal emanation, as Aristotle’s (2.13). This conflation is particularly significant, since it helps us to understand the appeal of the neoplatonic emanation scheme, when identified with Aristotle, to flesh out the lacuna in explanation which bedeviled his Metaphysics.30 Moreover, linking this theory with Aristotle would pose a formidable obstacle to the program of Maimonides and Aquinas. Neither regarded the efforts of their companions in faith to prove the creation of the universe de novo to be very helpful. Quite the contrary, in fact, as both complain that the arguments adduced are of such poor quality that they could “furnish infidels with an occasion for scoffing, as they would think that we assent to truths of faith on such grounds” (ST Ia.46.2).

Maimonides is thinking of the mutakallimûn, the Muslim religious thinkers who “did not conform in their premises to the appearance of that which exists [as Aristotle had], but considered how being ought to be in order that it should furnish a proof for the correctness of a particular opinion” (1.71). For this purpose they elaborated an atomistic metaphysics running “counter to the nature of existence that is perceived so that they resort to the affirmation that nothing has a nature in any respect.” Maimonides is referring specifically to the Ash’arite thesis that was intended to open the world of creatures to the direct action of God by withdrawing any intermediary structures like natures, whereas Aquinas focuses on their Christian counterparts, notably Bonaventure, who did not dally with an occasionalist thesis but nonetheless insisted that creation ex nihilo had also to be de novo.31 That the arguments, good or bad, could be so easily shared testifies to the accuracy of Maimonides’s contention that the “affirmation of the temporal creation of the world [is] common to all three of us, I mean the Jews, the Christians, and the Muslims” (1.71). So the divisions emerged within these communities rather than between them, and their origins were at once metaphysical and semantic: clarifying the precise import of the expressions used when the philosophical background favored certain understandings over others. Yet such confusions, as Wittgenstein has reminded us, can often illuminate the contours of the issues at stake. In this case, the terms “eternal,” “exnihilo,” “de novo,” and even “creation” itself are at issue.

Aquinas’s monograph on this subject is entitled “On the Eternity of the World,” yet he himself insists that “eternity, in the true and proper sense, belongs to God alone” (ST Ia.10.3). Like Maimonides, who does not hesitate to presume “the doctrine of the eternity of the world” in his demonstration of God’s existence, not because he believes it but because he wishes “to establish … the existence of God … through a demonstrative method as to which there is no disagreement” (1.71), Aquinas will adopt the current parlance while explicitly noting, toward the end of the work, “that nothing can be co-eternal with God, because nothing can be immutable save God alone” (DAM 11, also 10). Yet the thrust of his argument is to show that there is no contradiction in asserting that “something has always existed, understanding that it was caused by God with regard to all the reality found in it” (DAM 1). So “eternal” in this discussion will mean “always existed,” a predicate logically compatible with “the universe,” even when we acknowledge its total derivation from God. The argument proceeds to clarify at each step the grammar of the other terms involved in explicating such a transcendent relation. Ex nihilo will then be parsed not according to its surface grammar, which would turn nothing into that something “out of which” the universe was made, but as a façon de parler: “the creature is made ‘from nothing’, that is, it is made ‘after nothing’, [where] the term ‘after’ unquestionably connotes order: … it is nothing before it is a being” (DAM 7, also 6). Similarly, de novo cannot mean “in time,” as we often express this alternative – “that the world was created in time,” but rather must intend “that its initial moment initiate time as well” (cf. ST Ia.46.1, ad 6).

Moreover, none of these grammatical clarifications could be attempted if creating were a process that takes time. Indeed, the claim that “creation cannot be called change except metaphorically, so far as a created thing is regarded as having existence after non-existence” (SCG II.37) formulates the nub of Aquinas’s argument that one could conceive of the universe as always existing yet totally dependent upon its creator: if the activity of creating is inherently instantaneous, then there is no need that God temporally precede the universe to be its creator. So the press of controversy helps Aquinas to clarify a notion that had escaped al-Ghazali as well as Maimonides: that one could speak properly of a free creation without insisting that it had to be de novo (2.21).32 Yet the prevailing philosophical climate, drawn so powerfully to the logical model of intellectual emanation as an elegant articulation of the origin of the universe, would have supported Maimonides’s prima facie insistence that “in conceiving the world as created, we see it exhibiting purpose and design, expressing the will of a freely creating agent; [while] in understanding the world as eternal, we see it as displaying determinate and fixed laws that govern all natural phenomena, [so that] these two conceptual frameworks [would be] mutually exclusive models in terms of which nature is made intelligible to [us].”33 So the notion of an “eternal creation” has the ring of an oxymoron, if “creation” contains the note of “free origination” – so powerful is the grip of the necessary emanation scheme.34 Yet if Aquinas was able to escape that grip, it was no doubt due to the fact that Rabbi Moses had effectively prepared the way for negotiating the pitfalls in this debate, by mapping a way of distinguishing what is proper to the respective domains of demonstration (reason) and of revelation (faith).

Maimonides’s strategy for dealing with this vexing question is disarmingly simple – one more testimony to his skill as a teacher and expositor. He introduces the extant views as three “opinions,” thereby presaging the crucial step in his argument: none has the status of a philosophical demonstration. This ploy demands that he reduce the biblical view to “the opinion of all who believe in the Law of Moses, [namely] that the world as a whole … was brought ir ‘o existence by God after having been purely and absolutely nonexistent, and that God [who] had existed alone … through His will and His volition … brought into existence out of nothing all the beings as they are, time itself being one of the created things” (2.13). But such a dialectical move, like the presumption of the world’s eternity in proving God’s existence, in no way impugns his faith-commitment. The second opinion, said to be “the belief of Plato,” has God creating from “a certain matter that is eternal as the deity is eternal,” while the third, identified as “that of Aristotle,” insists that the impassibility of deity entails that the universe “was not produced after having been in a state of nonexistence.” He then focuses on Aristotle, “for it is his opinions that ought to be considered” (2.14), “to make it clear that Aristotle possesses no demonstration for the world being eternal, as he understands this” (2.15). What follows is a close reading of Aristotle to convince his “latter-day followers [who] believe that Aristotle has demonstrated the eternity of the world” that the master was quite aware of not having done so: that his form of argument corroborates his explicit statement: “that this doctrine was an opinion and that his proofs in favor of it were mere arguments [and] can Aristotle have been ignorant of the difference between mere [dialectical] arguments and demonstrations?” (2.15).

I have already suggested that such a picture of Aristotle, as espousing the eternal origination of the world, was really the work of al-Farabi and Avicenna, relying on Plotinus and Proclus. But never mind; that it was believed to be Aristotle’s is part of what gave the scheme its authority. So all that Maimonides needed to do was to detach the scheme from that authority and so reduce it to one opinion among others. Then he could discuss the relative merits of the other two – those of Moses and Plato. It is only at this point that he has recourse to the kalam arguments of “purpose and particularization” (2.20): not, however, as demonstrations of anything, but solely as “arguments” in favor of the opinion of Moses that can better account for the fact that what we observe in the movement of the heavens does not follow the ideal requirements of Aristotle’s scheme. Furthermore, since “the belief in eternity the way Aristotle sees it – that is, the belief according to which the world exists in virtue of necessity … destroys the Law in its principle, necessarily gives the lie to every miracle, and reduces to inanity all the hopes and threats that the law held out” (2.25), it would be absurd for believers to feel they had to espouse it short of its having been demonstrated. Had that been the case, Maimonides insists “we could interpret [the texts of the Torah] as figurative, as we have done when denying [God’s] corporeality” (cf. 1.1–48), so “our shunning of the affirmation of the eternity of the world is not due to a text figuring in the Torah” (1.25), but to our freedom to affirm what we believe to be true in the absence of proof to the contrary. The dangler in his treatment is the “opinion of Plato, [which] would not destroy the foundations of the Law,” and in accordance with which “it would also be possible to interpret figuratively the texts [of the Torah].” But given the fact that this too lacks demonstration, we are free to “take the texts according to their external sense and shall say: The Law has given us knowledge of a matter the grasp of which is not within our power, and the miracle [of the Torah itself] attests to the correctness of our claims” (2.25).35

That such matters are beyond our ken, since they have to do with the free activity of a divinity whose attributes we cannot know, is Maimonidean doctrine (1.51–60). What we can know are the results of divine action, revealed through nature as well as through the Torah. That these both reveal something of the divinity we know, however, requires “refuting the proofs of the philosophers bearing on the eternity of the world” (1.71), which would deliver to us a God who ruled by necessity; so Maimonides considers this to be the foremost task “of one who adheres to a Law.” Aquinas certainly agreed, as his efforts testify, but he seemed in this matter more consistent than Rabbi Moses, for he eschewed employing any kalam arguments, leaving the matter solely to faith, and not wishing to reduce the affirmation of faith to an opinion among others, even for dialectical purposes. Yet his transformation of Avicenna’s distinction of existing from essence allowed him a more positive characterization of this “matter the grasp of which is not in our power,” as the “continual influx of existing (esse)” (ST Ia.104.3) on the part of an eternal God acting freely.

Naming God

It should be clear by now that both Maimonides and Aquinas were concerned primarily about a proper understanding of divinity, although the context was the way to characterize creation. And they each safeguarded God and the ways of God from the reductionist solvents of philosophy by recourse to a via negativa, yet always a disciplined one. Nowhere does this appear more clearly than in treating the “names of God,” a topic on which Aquinas explicitly takes issue with Rabbi Moses, although, in practice, he espouses his critical concerns.36 Their respective treatments presume the medieval parallelism between semantics and ontology: what may be said of God cannot be thought as a modification of the divine substance, which must be “One by virtue of a true Oneness” (1.50): “supremely one … because [it is] subsistent existence itself [and] altogether simple, not divided in any way” (ST Ia.11.4). Maimonides’s prophetic insistence initiates his treatment of the multiple “names of God,” while Aquinas’s statement is the culmination of his metaphysical elaboration of the “formal features” attendant upon divine simpleness to show how our philosophical acumen must return us to a biblical faith.37

The context was set for Maimonides by the longstanding debate in the Islamicate on divine attributes, a discussion at once Qur’anic and philosophical in nature. It is customary for the Qur’an to punctuate its paranesis by recalling us to the feature of God apposite to the point being made – “He is the Wise, the Aware” (34:1), “the Merciful, the Forgiving” (34:2) – yet the overriding revelation respected the divine unity (tawhîd). So the earliest religious thinkers, the Mu’tazilites, exploited Greek philosophy to insist (according to Maimonides’s summary) that “there is no oneness at all except in believing that there is one simple essence in which there is no complexity or multiplication of notions” (1.51). The later followers of al-Ash’ari objected to the results of this teaching, which in effect reduced the Qur’anic statements to metaphorical expressions. So they settled for acknowledging the reality of such attributes in one God, yet “without saying how (bi-lâ kayf)” that could be possible.38 Maimonides mocks this intermediate position as inherently unstable: “some people engaged in speculation have ended by saying that His attributes … are neither His essence nor a thing external to His essence – these are things which are merely said” (1.51). So he will propose an entirely different approach to such expressions, familiar to him from the Psalms.

Since we can attribute nothing to a simple divinity, and since the prevailing context of God’s Scriptures (as well as the Qur’an) speaks of God’s deeds on behalf of the people, every such attributive statement is to be interpreted as expressing “an attribute of His action and not an attribute of His essence” (1.53), as “all His different acts [are] carried out by means of His essence and not… by means of a superadded notion” (1.52). Even the “attributes essential to” God, without apparent reference to divine activity on our behalf – “life, power, wisdom, and will” – “are not to be considered in reference to His essence, but in reference to the things that are created” (1.53). Indeed, what Moses himself was given to know (Exodus 34:6–7) “were simple pure attributes of action: merciful and gracious, long-suffering” (1.54). If this be all that God could communicate to Moses through a direct revelation, what more can we expect to glean of the divinity from the words of Scripture or the deliverances of reason? This does not keep us, of course, from regarding some of these expressions as “attributes indicative of a perfection likened to our perfections” (1.53), but we should realize that “the attributes ascribed to Him are attributes of His actions and that they do not mean that He possesses qualities” (1.54). What we can find significant in Maimonides’s resolution of this question is the way he finesses the tangle of issues in the Islamic debate in favor of the spirit of the Hebrew scriptures’ rendition of God.

Yet Aquinas finds Maimonides’s stated position unstable and so takes him on directly, focusing on a corollary of his treatment: that all such terms, when applied to God, “are purely equivocal, so that their meaning when they are predicated of Him is in no way like their meaning in other applications” (1.56). Beyond that, since “there is no composition in [God], He cannot have an affirmative attribute in any respect” (1.58). The best we can do with such statements is to interpret “every attribute that we predicate of Him as an attribute of action, or, if the attribute is intended for the apprehension of His essence and not of His action, it signifies the negation of the privation of the attribute in question.” Aquinas takes this to be “the view of Rabbi Moses: … that sentences like ‘God is good’, although they sound like affirmations are in fact used to deny something of God rather than to assert anything” (ST Ia.13.2). Aquinas does not object to the metaphysics implicit in Rabbi Moses’ semantics, for he treats the practice of naming God immediately after securing the divine simplicity; he is rather concerned that “this is not what people want to say when they talk about God.” The elaborate translation scheme that Maimonides proposes cuts against the grain of religious practice – presumably Jewish, Christian, and Muslim!

So Aquinas shifts the grounds of the discussion one step beyond those of his predecessor, employing some semantic tools unavailable to his interlocutors in the Islamicate, distinguishing between “the perfections themselves that are signified (res significata) – goodness, life, and the like – and the way in which they are signified (modus significandi)” (ST Ia.13.3). This distinction, gleaned from the twelfth-century explorations in the West of the various senses of Scripture, allowed him to insist that “so far as the perfections signified are concerned (res significata) the words are used literally (proprie) of God, and in fact more appropriately than they are used of creatures, for these perfections belong primarily to God and only secondarily to others. But so far as the way of signifying (modus significandi) these perfections is concerned the words are used inappropriately, for they have a way of signifying that is appropriate to creatures.” The distinction regards the adaptations one must make in surface grammar to construct valid syllogistic arguments, but Aquinas had already put it to metaphysical use, reminding us that “in talking about simple things we have to use as models the composite things from which our knowledge derives. Thus when God is being referred to as a subsistent thing, we use concrete terms [‘God is just’] (since the subsistent things with which we are familiar are composite); but to express God’s simplicity we use abstract terms [‘God is justice’]” (ST Ia.3.3, ad 1).

This observation addresses one of Maimonides’s recurrent concerns: that the very form of predication will mislead us into presuming that God “possesses qualities” (1.54). It also presupposes that in using the language appropriate to our condition we will be aware when the topic under discussion outstrips that language, and in what specific directions. Hence the focus on perfections, which Aquinas can suppose to represent traces of the creator, and whose semantic structure should reflect that fact. For not every term is susceptible of being distinguished in the way required: only those whose meaning can and must be said to outstrip their customary use, yet in a direction already intimated by that use.39 This highly articulate grasp of analogous features of language differentiates Aquinas’s treatment of “divine names” from that of Maimonides, who not only betrayed a rudimentary (and to Aquinas’s mind, false) grasp of “terms used amphibolously” (1.56) but also emphasized the differences between scriptural and philosophical usage by the rabbinic adage: “the Torah speaks in the language of the sons of men” (1.53), that is, in ordinary parlance.40 No doubt cutting the ties with our ordinary language reinforced his own convictions that the “names of God” that purport to signify essential attributes “are purely equivocal” (1.56). Yet Aquinas has shown how a speaker sensitive to distinctions already imbedded in our living language can use certain privileged terms of that language (“perfections”) to point beyond our “manner of signifying” to “intend to signify” their source in God. So “the language of the sons of men” need not be misleading about the utter simplicity of the One, nor need one be a philosopher to use such terms correctly of God. For our native grasp of perfection terms demands that they outstrip their current descriptive sense if we are to use them properly – whenever we use them.

The upshot of this nearly direct exchange between the two is that Aquinas’s resolution of the matter retains a generous dose of the “unknowing” that Maimonides sought to secure. Not, to be sure, in cutting all threads of meaning between, say, “knowledge” in our usage, and “knowledge” said of God as its object; yet neither does he demand that there be a shared “likeness in respect to some notion” (1.56). We may truly assert that God knows (res significata) without any sense of how it is that such is the case (modus significandi). Analogous usage for Aquinas is not to be explicated in terms of concepts but according to use, which could explain why many philosophers seem to have found the strategy elusive. In our reading of the two religious thinkers intent upon finding a way to speak of the utter oneness of God without distortion, however, it appears as though Aquinas’s strategy responds to his predecessor’s concerns without having to have recourse to the Rabbi’s self-defeating insistence on “pure equivocity.” For the grasp that we “composite beings” might be able to have of the “perfections signified” as they are in God’s own self will ever be a tenuous one; indeed, it will comprise a via negativa every bit as taxing as that sketched by Maimonides in his culminating chapter on “divine names” (1.59).

God’s knowledge of singulars and the range of providence

Maimonides’s fourteenth-century commentator, Levi ben Gershon (Gersonides), located the decisive reason for his master’s extreme agnosticism regarding divine attributes in the dilemma regarding God’s foreknowledge and human freedom.41 Maimonides insists that “it is in accordance with our Law that God’s knowledge does not bring about the actualization of one of the two possibilities even though He knows perfectly how one of them will come about,” and attributes all confusion on these matters to forgetting that “between our knowledge and His knowledge there is nothing in common” (3.20, emphasis added). So customary ways of generating the dilemma, “if God knows that something will take place, then it must occur,” are rendered nugatory if the formula “knows that” does not function in connection with God the way the argument-form presumes it should. Insistent as he is about the matter, however, Maimonides does not leave it at that, but suggests a specific difference (inspired by Avicenna): “A great disparity subsists between the knowledge an artificer has of the thing he has made and the knowledge someone else has of the artifact in question” (3.21). In short, if we insist on comparing God’s knowledge with ours, we ought to look not at our “knowings that” so much as the knowing that directs and issues in doing or making: “for in knowing the true reality of His own immutable essence, He also knows the totality of what necessarily derives from all His acts.”42 And while “it is impossible for us to know in any way this kind of apprehension, [it] is something extraordinary and a true opinion; … no mistake or distortion will be found in it.”

So Maimonides accepts Ibn Sina’s cue regarding the reversal of direction in knowing: from us who derive knowledge from existent things to the One whose knowing makes things exist. Yet the very necessity of Ibn Sina’s scheme keeps that same One from knowing individuals “except in so far as they are universal” (8.6, 360:3). This will not suffice for a follower of the Torah, who must insist that “divine providence watches … over the individuals belonging to the human species” (3.17). Aquinas transforms the assertion into a theorem: “God’s knowledge is the cause of things” (ST Ia.14.8) and “has the same extension as [God’s] causality, [so] his knowledge must necessarily extend to individuals” (ST Ia.14.11). There is no hint of how, of course, for that regards the utterly basic relation of free creation; but we can know that whatever exists does so by participation in God’s very existence, and only individuals exist. We are also freed from the specific bind from which Maimonides sought to escape, namely that God’s knowing that something was to occur would determine its occurrence, since the practical knowing associated with the creator of things’ very being is contemporaneous with the event itself and so must not be thought of as “foreknowledge.”

At this point, Aquinas will take exception to a restriction that Rabbi Moses presents as “my own belief” – that for all species below the human, individuals are simply subject to chance: “I do not by any means believe that… this spider has devoured this fly because God has now decreed and willed something concerning individuals” (3.17) The argument here is with Ash’arite Islamic thinkers, whom he depicts as claiming that “every leaf falls through an ordinance and decree from God” (3.17). Aquinas does not have to deal with such a view, and so focuses on the immediate context of Maimonides’s restriction: “intelligent creatures, because they have control over their own actions through free decision, come under providence in a special manner: blame or merit is imputed to them.” But that is not to be taken “in the sense of Maimonides, who thought that God’s providence has no concern for individual non-rational creatures” (ST Ia.22.2, ad 5). But in what sense does God have concern for them? In the measure that they fulfill their natures, which is (according to Aristotle) to contribute to the preservation of the species. So Aquinas’s assertion seems to come to little more than Maimonides’s denial, yet their statements differ according to the context of their concerns.

Where they differ considerably is in their respective characterizations of divine providence, which, for Maimonides, “is consequent upon the divine overflow” (3.17). That is, as he puts it in the concluding chapters of the Guide, “providence watches over everyone endowed with intellect proportionately to the measure of his intellect, [which allows such a one to attain] the perfection of the intelligibles that lead to passionate love of Him” (3.51). Yet while he seems to evoke the model of emanation to countenance a highly elitist providence, he will go on to note “that the end of the actions prescribed by the whole Law is to bring about the passion” (3.52) otherwise attained by “pure thought” (3.51). Where Rabbi Moses modifies an emanationist model to evoke something of divine grace, yet leaves the initiative with the individual (as he also does with prophecy [2.37]), Aquinas will carry the model of practical knowing to the level of God’s interaction with individuals as well, in his elaboration of an elevation by God of each person to a new level of sharing in the divine life – a sharing communicated by the death and resurrection of Jesus. But that carries us into a new paradigm introduced by the Christian tradition’s elaboration of the person of Jesus. It should suffice for us to have noted the considerable parallels between these exemplary Christian and Jewish thinkers; their differences are more readily apparent.

NOTES

1 Kraemer 1986, pp. 1–30.
2 Hodgson 1974, pp. 58–60.
3 Weisheipl 1983, p. 15.
4 Ibid., p. 296.
5 Thomas Aquinas 1983. For Maurer’s rendering of esse as “being” (in this translation) I have substituted “existing.” For an annotated edition of the original text, see Roland-Gosselin 1926.
6 Booth 1983. See also Burrell 1986a. In citations I have regularly altered his rendition of the Arabic huwiyya [lit., “thisness”] from “being” to “existing”. See also Owens’s Chapter 2.
7 Kahn 1982, p. 15. See Burrell 1986b, on which the following pages depend.
8 Booth 1983, p. 68.
9 Ibid., p. 66, n. 48.
10 Ibid., p. 74.
11 Ibid., p. 100.
12 For an account of the various terms the early Islamic philosophers adopted to translate key notions from Aristotle, see Shehadi 1982 passim.
13 References to Ibn Sina will be from Avicenna 1960a and Avicenna 1960b with citations by book and chapter, followed by page: lines. The French translation of Georges Anawati, Avicenna 1978 and Avicenna 1985, contains the Arabic pagination. The sections relevant to this inquiry are 1.6, 5.1, 8.4; the citation is from 5.1, 205:3.
14 The distinction between mahiyya and dhat is that between quiddity and essence, where quiddity answers the question “what [quid, ma] is it?” (in proper terms: genus plus differentia), while essence will tolerate a less precise answer. So in answering “what is necessary being?” one will not be able to respond with a genus and a differentia, but can say something: it is anniyya. On this point (and several others) I have been assisted by Shehadi 1982, esp. p. 84.
15 On anniyya, see Frank 1956.
16 The abundance of terms can be confusing because the Islamic philosophers were refining them as they went along; al-Farabi’s preferred huwiyya (or “thisness”) is gradually displaced by a set of terms that assume the more “abstract” cast of the original Greek. See Shehadi 1982, esp. p. 88.
17 Avicenna 1978, p. 78.
18 One is reminded here of Charles Sanders Peirce’s preference for “general” rather than “universal,” lest one pre-judge the issue by speaking of the “problem of universals.” See Peirce 1960, 1.27, 1.422, 1.165, 5–429, 5.453, 5.503.
19 Avicenna 1980, pp. 236-37.
20 For a useful discussion of the background to Avicenna’s notion, see Verbeke 1980, pp. 2*–19*.
21 Bellarmine’s letter can be found in Drake 1957, pp. 162–64.
22 Kahn 1982, p. 8.
23 Verbeke 1980, pp. 30*–36*, 51*–68*.
24 Kahn 1982, p. 8.
25 For the objections of Ibn Rushd (Averroes), see Averroes 1954, I.236. See also Rahman 1958 and Altman 1969.
26 The Dalâlat al-hâ’irîn, or Guide of the Perplexed, is available as Maimonides 1956 and the more recent Maimonides 1963.1 shall use Pines, amended as needed from Maimonides 1974.
27 For the Augustinians and their implicit philosophical commitments, see Gilson 1986; for the Latin Averroists, as well as later western and Islamic receptions, see Leaman 1988, “Averroism,” pp. 163–78.
28 Gardet 1974.
29 See Seymour Feldman’s clear summary of the positions in Feldman 1980, p. 293.
30 For an illuminating reading of Plotinus in this regard, see Gerson 1990, pp. 203–20; for a similar judgment on Aristotle, p. 140.
31 Aquinas’s discussion of this matter has been collated in Vollert 1984. For the Ash’arite thesis, see Gimaret 1990.
32 Ghazali provides background for our discussion here, since it appears that Aquinas knew only his Maqâsid al-falâsifâ, intended to introduce the positions of the Islamic “philosophers” before refuting them in his Tahâfût; cf. Hanley 1982.
33 Feldman 1980, p. 294, whose expression “model” is utterly true to Maimonides’s conception of our capacity to understand such matters.
34 For a study of confusions in Maimonidean scholarship on this matter, see Dunphy 1983.
35 Much has been made of Maimonides’s maneuvers here and whether his stated preference for the “opinion” of Moses reflects his true position. For an assessment, see Dunphy 1989.
36 This is the consensus of recent writing: see Broadie 1987; Burrell 1986c, ch. 4: “Names of God: Attributes of Divine Nature”; and Miller 1977.
37 See my extended treatment of this issue in Burrell 1979, ch. 2: “The Unknown;” also Jordan 1983.
38 For a comprehensive treatment of this discussion, see Gimaret 1988; for an example of a classical commentary, see Burrell 1991.
39 On analogous terms, see Ross 1981, as well as Burrell 1973. For critiques of different accounts of analogy, see Sherry 1976a, Sherry 1976b, and Burrell 1985. See also Ashworth 1991.
40 For a review of the state of such semantics in the Islamicate, see Wolfson 1973.
41 Samuelson 1977, pp. 182–224.
42 This is Ibn Sina’s contention: “it is not possible for the necessary existent to understand things from the things themselves…; rather, because it is the principle of all existence, it understands from its own essence that of which it is the principle” (8.6, 358:13; 359:1).

Written by David B. Burrell in "The Cambridge Companion to Aquinas", Cambridge University Press, UK, 1993, edited by Norman Kretzmann and Eleonore Stump, excerpts chapter 3. Digitized, adapted and illustrated to be posted by Leopoldo Costa.