Therese Scarpelli Cory (Seattle, WA)

Institut Catholique de Paris 30 May 2012:

“The Footprint of Avicenna’s Flying Man

in Aquinas’s Commentary on the Sentences I.3.4.5”

Abstract

In thirteenth-century Latin discussions of self-knowledge, one of the foremost concerns was the phenomenon of self-familiarity.  In his De Trinitate Augustine had unforgettably described how the mind can never encounter itself as foreign or new, but only as something that had always been familiar.  In Avicenna’s ‘Flying Man’ thought experiment (Liber de anima I.1 and V.7), a number of early thirteenth-century thinkers found an explanation for this phenomenon: The soul is by nature self-thinking, even though it is generally distracted from itself by involvement with sensation.  Although Aquinas initially adopts this Avicennian position at the beginning of his career (In Sent. I), he soon rejects it, arguing instead in Sent. III and De ver. that the intellect perceives itself only in its acts of cognizing other things.  While these two views are strikingly different (Avicenna posits an unconscious realm of natural self-thinking, whereas Aquinas posits only a natural habitual self-knowledge, locating self-awareness in the structure of intentional acts), Aquinas’s mature position turns out to have some interesting debts to Avicenna. This paper will begin by sketching Avicenna’s position on self-knowledge as expressed in the texts available to the Latin medievals.  I will then trace the Avicennian elements in Aquinas’s developing theory of self-knowledge in four early texts texts—In Sent. I.3.4.5, In Sent. III.23.1.2, ad 3, De ver. 1.9, and De ver. 10.8.  Given the prominence of the ‘Flying Man’ in his predecessors’ treatments of self-knowledge and Aquinas’s own strenuous objections to the theory of self-knowledge illustrated therein, however, it is odd that he never refers to it.  I conclude by considering possible explanations for this striking absence.

“The Footprint of Avicenna’s Flying Man

in Aquinas’s Commentary on the Sentences I.3.4.5”

Therese Scarpelli Cory, Seattle University
coryt@seattleu.edu

In his “Flying Man” thought experiment (Liber de anima I.1 and V.7), Avicenna paints a picture of self-knowledge as indissociable from the very essence of the soul.  Even if it had never had any sensory input, he argues, the soul would grasp its own self:

If a man were instantaneously created, with his hands and feet spread out, so that he neither saw nor touched them, nor touched himself nor heard a sound, truly he would not know that they were his own limbs, and nevertheless he would know that he existed (sciret se esse), and that he is some one thing, even though he would not know all those other things (for what he does not know is not [the same as] what he knows).1

The “Flying Man,” and the view of the self-knowledge that it illustrates, appears in a number of prominent Latin Augustinian followers of Avicenna, such as Domenicus Gundissalinus, William of Auvergne, Jean de la Rochelle, and Pseudo-Petrus Hispanus, and, slightly later, Matthew of Aquasparta and Vital du Four.2  And in fact, by the mid-thirteenth-century, there is a widespread assumption that the human soul possesses a constant active self-knowing that remains outside the boundaries of conscious life, serving as the unifying background for all conscious activities (I will call this “supraconscious self-knowing”3).  This Avicennian position resonated nicely with Augustine’s repeated claims in De Trinitate XIV that “the mind always remembers, understands, and loves itself, although it does not always think of itself as something distinct from those things that it is not.”4 Augustine does not clearly explain what this perpetual self-understanding (se intelligere) is, although his medieval readers noticed that he distinguished it from two other kinds of self-knowledge, i.e., the occasional act of thinking about oneself (se cogitare or sometimes se discernere); and a perpetual habitual self-familiarity (se nosse).5  Under the influence of Avicenna, then, Augustine’s medieval readers could characterize this perpetual self-understanding as a supraconscious self-knowing.

Thomas Aquinas, however, breaks with his contemporaries by rejecting the possibility of a supraconscious self-knowing.  He argues that because the human intellect is a mere potency for intelligible form with no native act of its own, all self-knowledge depends on the intellect’s being actualized in the cognition of something else (intelligit se per actum).  Thus for Aquinas, Augustine’s perpetual self-understanding must be taken as merely a perpetual habit for self-knowledge, or else as a kind of self-understanding that occurs “in all acts” rather than “always.”6  Aquinas begins to lay the foundations of this position in Sent. III and IV; he fully develops it in De veritate, q. 10 (1257-58); and from then on he steadfastly maintains it right up into his last discussion of self-knowledge in the Commentary on the Liber de causis (1272).

Consequently, Aquinas’s and Avicenna’s theories of self-knowledge are clearly philosophically at odds.  But in attempting to parse the historical relationship of these two theories, we run into a series of puzzles.  For one thing, as a number of scholars have noted, Aquinas never mentions the Flying Man, not even in passing, in any of his writings.7  In fact, he never attributes to Avicenna a theory of supraconscious self-knowing; instead, when Avicenna appears in Aquinas’s texts on self-knowledge, he always gets a positive mention as an important source for the reflexivity of immaterial acts and the intermittence of self-perception.8  Even if we assume that Aquinas did not read Avicenna’s Liber de anima himself, it is hard to imagine that he had never heard of the “Flying Man,” which appears in the writings of at least two important figures at Paris (William of Auvergne and Jean de la Rochelle).

The more closely one looks at the Thomistic texts on self-knowledge for Avicennian influences, the more complicated the puzzle becomes.  In fact, it turns out that in his very first treatment of human self-knowledge—namely, Sent. I.3.4.5—Aquinas actually attributes to the human soul a supraconscious self-knowing.  But apart from the philosophical similarity to Avicenna’s view, nothing in this text indicates any historical dependence on Avicenna.  Instead, Aquinas’s defense of supraconscious self-knowing derives directly from the parallel text in his teacher Albert’s commentary on the Sentences.  And here is the really interesting part: Avicenna’s “Flying Man” is just as conspicuously absent from Albert as it is from Aquinas9—even though, as Hasse has pointed out, Albert was one of the strongest medieval champions of Avicenna’s theory of the soul.10 

In this paper, then, I want to examine closely the relationship—or lack thereof—between Avicenna’s thought on self-knowledge and Aquinas’s treatment in his commentary on the Sentences, in order to draw some broader conclusions about Avicenna’s impact on Latin medieval discourse on self-knowledge.  The first part of the paper will sketch how Avicenna’s Latin readers incorporated his view of supraconscious self-knowing into their own theories of the soul, and how this development affected their reading of Augustine’s claim that the mind “always understands itself.”  The second part will examine Aquinas’s earliest treatment of self-knowledge in Sent. I d. 3 q. 4 aa. 4-5, where, I argue, we find Aquinas defending the same supraconscious self-knowing theory that he rejects shortly thereafter.  In the third part of the paper, I will sketch some historical conclusions and highlight an important question that remain to be answered.  We will see, I hope, that once set against its historical background, Aquinas’s Sent. I.3.4.5 offers us a fascinating glimpse of the development of one strand in the movement that Gilson called augustinisme avicennisant.

I.  Avicenna’s ‘Flying Man’: How Supraconscious Self-Knowing Entered the Latin Discourse

Supraconscious self-knowing and the Flying Man.  It is generally agreed that Avicenna attributes two different kinds of self-knowledge to the human soul: on the one hand, a supraconsciousness self-knowing (sometimes called “primitive self-awareness” in the secondary literature); and on the other, reflexive second-order acts of cognizing one’s own acts.11  Both kinds of self-knowledge have been the object of several studies already,12 so here I merely summarize Avicenna’s views on the first kind, i.e., supraconscious self-knowing.  Avicenna’s “Flying Man” thought-experiment (which I read at the beginning of the paper) appears in I.1 and in V.7 in the Latin text of the Liber de anima,13 each time as part of a larger argument aimed at distinguishing the soul from the body.  Avicenna claims that the Flying Man, created in a state of total sensory deprivation, lacking any awareness of his own body, would still “know that he existed (sciret se esse), and that he is some one thing.”14  Even though his consciousness is wholly cut off from the body, the Flying Man enjoys a concrete experience of the existing “self” or “I,” showing that the self-knowing soul is neither the body nor a part of the body.15  The same is true, then, for each of us: The “self” or core entity that I consciously grasp is my own soul, and I have direct access to my soul independently of any sort of sensory activity.  In LdA V.7, in the discussion leading up to the second iteration of the “Flying Man,” Avicenna makes clear that this basic supraconscious self-knowing belongs to the soul’s very essence as the basis for all conscious, intentional activity. I experience my acts as belonging to a single self because all these acts are performed by a single self-knowing soul that grasps itself simply by being itself.1617 

The impact of the Flying Man’s supraconscious self-knowing on the Latin medievals.  The introduction of Avicenna’s theory of supraconscious self-knowing in the Latin West effected a significant shift in both the Latin medieval approach to self-knowledge and in the philosophical content of Latin thought on self-knowledge.  The extent of Avicenna’s influence becomes evident when we compare treatments of self-knowledge before and after the translation of Avicenna.  As examples of “pre-Avicennian” treatment of the self-knowing soul, we can use two influential 12th-century Cistercian treatises: the De natura corporis et anima by William of St. Thierry [c. 1140] and the Epistola de anima by Isaac of Stella [c. 1162]).18  For “post-Avicennian” treatments, I will limit myself to authors known to be influenced by Avicenna, such as Gundissalinus, William of Auvergne, Jean de la Rochelle, and Albert the Great.19

In comparing “pre-Avicennian” and “post-Avicennian” Latin treatments of the self-knowing soul, one immediately notices a divergence in approach.  The approach in the Cistercian treatises is largely theological and ethical (as is typical of  pre-Avicennian treatises20), with some admixture of medical knowledge.  For these Cistercian thinkers, the primary goal of exploring the soul’s nature is to emphasize the soul’s distinctness from corporeal reality and its nobility as microcosm and image of God.  There is little attempt to explain self-knowledge as a psychological phenomenon in its own right; rather the treatises themselves are an exercise of purificational self-knowledge, assisting the soul to find God by rediscovering his forgotten image within itself.21  In contrast, in post-Avicennian treatises or texts on the soul, we find a much more philosophical analysis of the soul and its activities, in which self-knowledge is treated as a psychological phenomenon to be explained in its own right.  This trend toward a structured philosophical soul-theory was probably initiated by Avicenna, but it would certainly have been reinforced by the growing body of newly-translated Aristotelian literature.

Further, in comparing the content of pre- and post-Avicennian views on self-knowledge, it seems Avicenna’s “Flying Man” (and the Liber de anima more generally) seems to have contributed three crucial theses that became widely accepted in the late-12th- and early-13th-century discussion of self-knowledge22:

(1) The per essentiam thesis: The soul knows itself by its essence, i.e., simply by being itself.  Avicenna states that the soul itself is “that by which” it grasps its own being.23  In fact, the Flying Man suggested to Latin readers that the soul is an essentially self-knowing being, necessarily and perpetually grasping its own concrete identity24 (for instance, William of Auvergne and Jean de la Rochelle make this point immediately after discussing the Flying Man).  But the question of how the soul knows itself is of little interest to the Cistercian treatises, and does not arise in these treatises at all.

(2) The supraconsciousness thesis: This self-knowledge is always functioning at the core of the soul’s being, regardless of whether it is consciously thinking about itself or not.  Avicenna insists that the soul cannot ignorant of itself; rather it “always apprehends its essence.”25  To many Latin readers of Avicenna, it seemed obvious that if the soul by its essence is sufficient to know itself (by the per essentiam thesis), then it necessarily does, since it is intelligible, present to itself, and nothing is impeding its self-knowledge.26  In contrast, when the Cistercian Isaac of Stella refers to the knowledge that the soul always has of itself, he is inclined to describe it as an innate self-memory, forgotten because of the distraction of the senses so that the soul becomes ignorant of itself.27 

(3) The incorporeality thesis: This perpetual state of self-knowing is the result of the soul’s incorporeality.28  The “Flying Man” appears in LdA I.1 as part of a discussion of the soul’s distinctness from the body, and in LdA V.7, as part of an argument that the soul is neither the body nor any part of the body.  The link between self-knowledge and incorporeality seems to be reinforced in LdA V.2, where Avicenna describes in detail why a power that uses a corporeal organ cannot know itself or its act.  No wonder, then, as Hasse has pointed out, that Avicenna’s earliest Latin readers were inclined to read the “Flying Man” as an argument for the soul’s incorporeality.29   For the Cistercian treatises, however, incorporeality does not play any special explanatory role in human self-knowledge.  Rather, the goal of these treatises is to convince the soul of its own incorporeality, and incorporeality is used in their arguments only to justify their introspective methodology for self-purification: i.e., one must turn inwards, away from the corporeal senses, in order to behold the pure, incorporeal being of the soul.30

Now one might object: Surely these theses are not uniquely Avicennian!  Did the medievals not find them first in Augustine’s De Trinitate?  For instance, Augustine articulates the per essentiam thesis and the incorporeality thesis together in a single lapidary phrase in De Trin. 9.3.3: “The mind knows itself by itself because it is incorporeal.”31  And Augustine’s claim in De Trin. 14.6.9 and elsewhere that the soul “always understands itself” could reasonably be interpreted as a defense of the supraconsciousness thesis.32

But the attention of Latin authors was not drawn to these texts before their encounter with Avicenna’s theory of supraconscious self-knowledge.  In pre-Avicennian treatises, the relevant Augustinian texts are hardly ever cited, and their role, when they are cited, is negligible.33  In post-Avicennian discussions of self-knowledge, in contrast, we find a burgeoning interest in Augustinian texts that correspond to the above Avicennian theses, both genuinely Augustinian texts like De Trin. 9.3.3 and 14.6.9, and the corrupted versions that appear in the pseudo-Augustinian Liber de spiritu et anima.  The increased citation of these Augustinian texts suggests a scramble among Latin authors to highlight the connections between Avicenna’s theory and Latin authoritative sources.34  In fact, once the Latin followers of Augustine were introduced to a theory of supraconscious self-knowing, it was easy to find the same theory in Augustine, because the parallels between Avicennian and Augustinian thought on the soul are striking (which surely contributed to the Latin acceptance of Avicenna).35  Before the translation of Avicenna, no attempt had been made to read these texts in support of a supraconscious theory of self-knowing.  But after Avicenna, it seemed impossible to read Augustine in any other way. 

II.  Supraconscious Self-Knowing in Aquinas’s Commentary on the Sentences

With this historical background in mind, we can now examine how Avicenna’s theory of supraconscious self-knowing both did and did not affect Thomas Aquinas’s own thought in his commentary on the Sentences, and what their relationship tells us about the discourse on self-knowledge in the 1240s-50s. 

At the very beginning of Aquinas’s academic career, in his first treatment of self-knowledge in his Commentary on the Sentences, book I, d. 3., q. 4, a. 5 (approx. 1252-54), he makes a claim that is puzzling in light of his later theory of self-knowledge: The soul always understands itself (se intelligit), with a “gaze that is nothing other than an intelligible presence to oneself in any way whatsoever”; this self-understanding occurs “naturally” and “does not require an intention.”36  Now this text has usually been read as Aquinas’s earliest description of habitual self-knowledge (namely, the soul’s essential ordering toward self-knowledge).37  But this interpretation cannot be correct.  It is the result of reading the responsio of a. 5 out of context, through the lens of later texts. 

Instead, I argue that in this early text, Aquinas falls in line with his Avicennist contemporaries, attributing to the soul a perpetual supraconscious self-knowing.  When we read a. 5 in context—as I now propose to do— it will become clear that Aquinas considers this perpetual self-understanding to be a non-intentional act of the intellect (corresponding to Augustine’s se intelligere), which is generated by the soul’s essential habit for self-knowledge (corresponding to Augustine’s se nosse), and which must be distinguished from intentional acts of self-knowledge (corresponding to Augustine’s se cogitare).

Aquinas’s discussion of the intellect’s natural self-understanding is set against the background of In Sent. I, d. 3, where he treats the question of how God can be known from creatures.  Q. 4 as a whole examines the vestiges of the Trinity contained in the powers of the mind: memory, intelligence, and will (the first Augustinian mental “trinity”38).  We will begin by looking briefly at a. 4, which provides the essential context for a. 5.  In a. 4, Aquinas explains how different objects allow the mind to express the Trinitarian image more or less perfectly in its acts.  One of the criteria he uses for distinguishing more and less perfect mental expressions of the Trinity is the order of mental acts: namely, the act of memory must generate the act of intellect, producing thence the act of will, just as the Father generates the Son and thence the Holy Spirit.39  This Trinitarian order is not preserved in the case of acquired objects, because for such objects, the act of understanding generates the act of memory (for instance, I must first understand triangularity in order to remember it).  But when the mind itself is its own object,40 then the order of its acts properly expresses the Trinitarian image, with self-remembering generating self-understanding and then self-love.41

Now since memory is a retentive power, it may seem odd to speak of an act of memory generating an act of understanding oneself. Earlier in q. 4, a. 1, however, Aquinas had explained that one can refer loosely to memory’s “holding” something habitually as its “act”: “In place of an act, it belongs to memory to hold (tenere).”42  He defines ‘nosse’ as this state of holding something mentally present: “Nosse is to hold notitia [knowledge or familiarity] of something present with oneself.”43  This habitual “holding” is exercised in the operation of intellect.44  So when Aquinas states in q. 4, a. 4 that the mind preserves the proper Trinitarian order in the acts of memory, intelligence, and will for which the mind itself is the object, he means that the mind is innately, habitually present to itself, and that this “self-memory” generates the act of self-understanding.  “But if these powers are considered with respect to the object that is the soul itself, then [the Trinitarian order of acts] is preserved, because the soul is naturally present to itself.  Thus from familiarity proceeds understanding, and not vice versa (ex notitia procedit intelligere, et non e converso).”45 

Thus in q. 4 a. 4, it is already clear that the mind has a habitual self-knowledge; that this habitual self-knowledge is a sort of innate memory of oneself (the Augustinian se nosse / notitia sui,46 and that this habitual self-knowledge is distinct from the self-understanding that it generates.  So of course the next obvious question is: Does this “natural” self-presence perpetually generate a natural act of self-understanding?47  This is precisely the question that Aquinas goes on to address in a. 5, where he asks “whether the aforementioned powers are always in their acts” of remembering, understanding, and loving oneself and God.48

Throughout a. 5, Aquinas focuses mainly on showing that the power of intellect is always in the act of understanding itself.49  In the body of the text, Aquinas identifies two different solutions to the question of whether the mind is “always understanding itself.”  The first solution, attributed to Augustine, posits a true supraconscious self-knowing.  (I will not discuss the second solution, attributed to “the philosophers”—but for the record, we can note that it provides for the intellect’s cognition of itself “in every act,” a mode of self-knowledge that Aquinas will later elevate to the only way that the soul can actually cognize itself).

In presenting the first solution, Aquinas identifies “self-understanding” (intelligere) as a unique kind of act that differs from and does not compete with attentive, intentional acts of cognition.  As is typical of other Sentences commentaries, he distinguishes among different types of cognition taken from Augustine’s De Trinitate: cogitare (“to consider a thing according to its parts and properties”), discernere (“to know a thing by its difference from others”), and intelligere (“the intellect’s simple vision [intuitus] of the intelligible thing which is present to it”).50 Armed with this distinction, Aquinas explains:

I say therefore, that the soul does not always cogitate or discern God or itself, because in this way anyone would naturally know the whole nature of his soul, which is scarcely attained with great study: for such cognition, the presence of the thing in just any way is not enough; but it is necessary that [the thing] be there [i.e., in the intellect] as an object (in ratione objecti), and an intention of the one cognizing is required.  But insofar as intelligere means nothing more than a vision (intuitus), which is nothing other than an intelligible presence to the intellect in any way whatsoever, the soul always understands itself and God, and a certain indeterminate love follows.51

Now against the background of q.4 a.4, it is evident that this perpetual self-understanding cannot be a habitual self-knowledge.  Rather, it is an act of intellect perpetually generated by the soul’s habitual self-knowledge.52  But Aquinas insists that it is not a “perfected” intellectual operation intentionally determined toward some object; rather, it is a “nonintentional,” “natural” act.53 

Aquinas refines this distinction between nonintentional and intentional intellectual acts in responding to the second and third arguments, both of which assume that all intellectual activity is intentional and that therefore the mind cannot always be in the act of understanding itself.54  According to the second argument, many things cannot be understood at the same time, so that if I am always in the act of understanding myself, I would never be able to think about triangles, chickens, or trees.55  Aquinas agrees with the major premise, but denies its applicability to self-understanding: The mind cannot understand two things at once only if we take “understand” to mean a “perfected operation (operatio completa) of the distinguishing or cogitating intellect,” rather than “in the sense here ascribed to intelligere.”56 According to the third argument, intellectual acts are by definition attentive, requiring an “intention by which the intelligible species is directed toward the thing.”  So since we are not always attentively thinking about ourselves, we are not always in the act of self-understanding.57  In response, Aquinas simply states that intelligere, as defined it here, does not require an “intention by the one understanding.”58 

Here in Sent. I.3.4.5, then, Aquinas is defending a perpetual nonintentional act of self-understanding that precisely parallels the supraconscious self-knowing that Avicenna and his Latin followers endorsed.  Yet he does not mention Avicenna, and in fact, his solution shows no signs of special dependence on Avicenna at all.  His approach and vocabulary are different from those that are standard for Latin Avicennians such as William of Auvergne or Jean de la Rochelle (for instance, he does not connect the soul’s perpetual intelligere to its incorporeality or locate it in the soul’s essence; instead he describes it as a natural act of intellect or a non-intentional self-vision).  As I will argue in section III, there are strong reasons to believe that Aquinas did not derive this theory directly from Avicenna or even associate it specifically with Avicenna. 

Curiously, in the context of self-knowledge Aquinas only cites Avicenna implicitly or explicitly in connection with the intellect’s ability to cognize its own acts (recall that Avicenna posits not only a supraconscious self-knowing, but also a reflexive knowledge of one’s own acts).  For instance, in Sent. I.17.1.5, ad 4, Aquinas appeals to Avicenna to show that the intellect can have a potentially infinite series of thoughts about thoughts.59  In Sent. III.23.1.2, ad 3 (the first text in which we begin to see the outlines of his own mature position of self-knowledge), Aquinas argues that the intellect can cognize its own acts because it is incorporeal, and he repeats Avicenna’s argument from LdA V.2 that powers using corporeal organs cannot cognize their own acts.60  And in Sent. IV.49.3.2 he cites Avicenna’s De medicinis cordialibus to the effect that we only notice acts when there is a change from one act to another, but not when an act is continuing and unchanging.61  In all these texts, Avicenna appears in a positive light and his arguments play a crucial role in justifying the intellect’s ability to cognize its own act. 

III.  The Missing Footprint of the Flying Man in Aquinas

So what conclusions can we draw from all this, regarding the historical influence of Avicenna on Aquinas’s early thought on self-knowledge?  In attempting to put the pieces of this historical puzzle together, we can get an important clue by turning back from Aquinas to his teacher, Albert the Great.  Significantly, no trace of Avicenna’s ‘Flying Man’ has yet been found in Albert either—not even so much as a footprint!  This absence is even more puzzling in Albert than in Aquinas, because Albert is an enthusiastic proponent of the soul’s supraconscious self-knowing and he is known as a champion of Avicenna in the mid-thirteenth century.62  Yet his treatment of self-knowledge does not show any signs of special dependence on Avicenna—instead, he characterizes the soul’s supraconscious self-knowing in a way that Avicenna never would have done, i.e., as the result of the light of the agent intellect perpetually shining on the possible intellect.  The influence here is Averroist—and this gives us a clue to Avicenna’s absence. 

Supraconscious self-knowing in Albert’s Sentences. An important piece of evidence appears in Albert’s treatment of the problem of perpetual self-understanding63 in his commentary on the Sentences I.3.H.29, which dates from approximately 1243.64  When one places Albert’s Sent. I.3.H.29 side-by-side with Aquinas’s Sent. I.3.4.5, it is clear that the two solutions in Aquinas’s respondeo, as well as nearly all of his pro and sed contra arguments, are simply paraphrases of Albert (see Appendix 1).65  Albert presents his two solutions in much more detail.  He distinguishes the solution of “a certain” Aristotelian philosopher, i.e. Alexander of Aphrodisias,66 from Augustine’s solution, which holds to a supraconscious self-knowing). 

A certain philosopher explicating Aristotle in De intellectu et intelligibili holds that the intellect always sees, that is, understands itself.  And he holds that this is the same as to say that the intellect understands itself in every intelligible. . . . But even though this solution appropriately satisfies the Philosopher’s dictum,67 in no way does it satisfy the dictum of Augustine, who wants to hold that behind every other intelligible (sub omni intelligibili alio), intellect always understands itself and the other two powers: and the memory always remembers itself and the other two powers: and will always wills itself and the other two.68   

Albert then goes on to sketch the “Augustinian” solution by distinguishing among nosse, intelligere, cogitare, and discernere as different kinds of cognitions, in order to identify intelligere as a unique kind of intellectual gaze, neither a habitual “holding-present” (nosse) nor an act of thinking about the intellect as a distinct intentional object (cogitare / discernere).  We already saw this solution in Aquinas, and it appears in other contemporary commentaries on the Sentences at the time, such as those of Alexander of Hales and Bonaventure. But Albert goes a step further, struggling to explain how exactly this perpetual self-understanding takes place.  For an explanation, he appeals to the mechanics of Aristotelian psychology:

But this intellectual gaze (intuitus) is completed by two things, namely by the emitted intelligible light and by the object: and to behold oneself (se intueri) is nothing other than to be presented (objici) to oneself in that light.  Therefore because the light of the agent intellect always shines upon the possible intellect inasmuch as the agent intellect is the form and act of the possible intellect, the possible intellect is always presented (objicitur) to itself in its light: and because this is what “understanding oneself” is according to Augustine, it always understands itself.69

Where Avicennian-inspired explanations of self-understanding tend to appeal to the soul’s incorporeality (an incorporeal being must be intrinsically self-knowing), Albert casts his solution in terms of the natural relationship between the possible and agent intellects.  For Albert, the light of the agent intellect is sheer undifferentiated intelligibility-in-act.  When this actual intelligibility is “specified” to the particular species ‘catness’, this actually-intelligible species informs the possible intellect, resulting in an intellectual act that is intentionally directed toward “catness” (and now I am thinking about what it is to be a cat!).70 

But for Albert, even when I am not thinking about anything specific, the light of the agent intellect is still always “on” as “the form and act of the possible intellect,” resulting in the intellect’s sheer actual intelligibility.  Because this act is not determined or specified by a species, it is non-intentional, i.e., it is not “about” anything in particular, which is why this self-understanding does not allow the intellect to distinguish itself from other things.  In responding to the 4th argument, Albert identifies this self-understanding as something that the intellect has “by nature” (secundum naturam), just as Aquinas will later describe it as “natural.”71

Now certainly this account of supraconscious self-knowing does not come from Avicenna, for whom, as Albert well knows, the agent intellect is a separate intelligence, not a power inhering in the individual human soul.  Instead, I believe Albert’s view is somehow linked to his misreading of Averroes’s discussion of the continuatio or conjunction between the material and agent intellects in the Long Commentary on De anima.  The details must be set aside for another paper, but in short, the story is as follows: As Richard Taylor has recently pointed out, Albert mistakenly took Averroes’ position to be that the agent and possible intellect inhere in the individual soul as its powers.72  Under the influence of this misinterpretation, Averroes’ discussion in LCDA 3.36 of the continuatio of the material and agent intellects in the act of thinking may have given Albert the idea that even when there is no determinate act of thinking, the light of my agent intellect inheres in my possible intellect as the undetermined form of the possible intellect.73  Incidentally this interpretation would have also fit nicely with the extromissive theories of vision74 and theories of “indeterminate light” in earlier Latin illuminationists.75  This account of supraconscious self-knowing seems to be original with Albert, and initiates a whole tradition of thought on the ever-shining agent intellect that includes Dieterich of Freiburg and later Nicholas of Cusa.76 

A Historical Narrative.  From this discussion, I think we can draw two tentative historical conclusions about Avicenna’s role in the early Scholastic discourse about self-knowledge and in the early Aquinas in particular, as well as one methodological conclusion about reading Aquinas’s Commentary on the Sentences. 

The first conclusion is this: While Avicenna’s “Flying Man” introduced the medieval Latins to the concept of supraconscious self-knowing, it became the victim of its own historical success in the Latin West.  Paradoxically, it is the absence of Avicenna in Albert’s and Aquinas’s discussions of self-knowledge that tells us just how pervasively Avicenna’s theory of the supraconsciously self-knowing soul had soaked into the Latin discourse on the soul since Gundissalinus first retold the tale of the Flying Man in his own De anima.  A thinker cannot become the “poster child” for a theory unless his view stands out in striking opposition to those of his contemporaries.  But Avicenna’s paradigm of the supraconsciously self-knowing soul fit so naturally with preexisting discourse, and found so many supporting texts in Augustine, that it quickly became the standard “Augustinian” theory of self-knowledge and was no longer uniquely associated with Avicenna.77  This historical narrative is consistent with the textual evidence: namely, the shift from an ethical approach to self-knowledge in pre-Avicennian treatises to a philosophical analysis of self-knowledge in post-Avicennian treatises; the widespread adoption of the ‘Flying Man’ and a theory of supraconscious self-knowing ; and the new focus on Augustinian texts that manifest Avicennian theses.

Now we can see that there is nothing particularly suspicious in the absence of Avicenna and his “Flying Man” from Albert’s discussions of self-knowledge in De homine and the Commentary on the Sentences.  As a careful reader of Avicenna, Albert was probably quite familiar with the “Flying Man,” but he does not omit Avicenna from his treatment of self-knowledge by some deliberate desire to marginalize his view.  Rather, it simply does not occur to Albert to single Avicenna out from the many voices supporting the soul’s supraconscious self-knowing; as far as Albert is concerned, this is a bona fide Augustinian view, and he only needs to deal here with Augustine.  Moreover, Avicenna’s Liber de anima does not explain how this supraconscious self-knowing actually works in terms of Aristotelian faculty psychology.78  So when Albert seeks a detailed account of Augustine’s supraconscious self-knowing in terms of Aristotelian psychology, it would have been natural for him to look for answers in Averroes and other Aristotelian commentators, especially since he had already taken other key components of his broader theory of self-knowledge from the commentary tradition.79

The second historical conclusion that I want to propose is this: The early Aquinas’s way of construing supraconscious self-knowing is distinctively Albertist, both when he defends supraconscious self-knowing in Sent. I.3.4.5 and when he rejects it in DV 10.8.80  In light of my first conclusion, a dependence on Albert’s way of framing the issue explains the absence of distinctively Avicennian themes in Aquinas’s early defense of supraconscious self-knowing in Sent. I.3.4.5.  It also explains Aquinas’s approach when he rejects supraconscious self-knowing in DV 10.8 and beyond.  Albert presents supraconscious self-knowing as an Augustinian concept that can be explicated by applying Aristotle’s view of the ever-active agent intellect.  Not surprisingly, then, in DV 10.8 and subsequent texts, Aquinas takes pains to show that neither Augustine nor Aristotle endorse this view: rather, Augustine means that the mind  essentially has a habitual self-knowledge, and Aristotle holds that the intellect knows itself by its acts.81 

Now it is tempting to press this conclusion further, to say that Aquinas may not have even read Avicenna’s Liber de anima himself at the time of writing the Commentary on the Sentences, relying instead on Albert for indirect transmission of Avicennian ideas.  Certainly the historical narrative I sketched above is consistent with an indirect transmission hypothesis; but the same narrative is equally consistent with Aquinas’s having directly read the Liber de anima under Albert’s influence, without seeing in Avicenna’s discussion of self-knowledge anything unusual or interesting enough to warrant mention.  To settle this question, more work on Aquinas’s Sentences is needed.82 

I want to end with a general methodological conclusion about reading Aquinas’s commentary on the Sentences. Since this work is not as widely read as later texts such the Summae, one typically approaches the text already familiar with Aquinas’s mature position, looking for further discussion of a particular theme.  It is particularly tempting to appeal to the later texts to clear up the ambiguities or unusual phrasing in this earlier work.  As we can see with Sent. I.3.4.5, however, this methodology could easily lead one astray.  A safer way of obtaining exegetical clarity, in my view, is to place Aquinas’s commentary on the Sentences back into its historical background, tracing it to its sources and comparing its doctrine with that of other Sentences commentaries, such as those of Bonaventure, Albert, and Alexander of Hales.  In this way, we can get a sense of what traditional distinctions, questions, and terminology are already part of the conversation that Aquinas joined; we can recognize instances in which his earlier and later positions diverge; and we can thus become more sensitive to the philosophical and theological concerns driving these shifts.

1 AviLat LdA V.7: “Repetamus autem id quod praediximus, scilicet quod si subito crearetur homo, expansis eius manibus et pedibus, quae ipse nec videret nec contingeret nec ipsa se contingerent nec audiret sonum, nesciret quidem esse aliquid membrorum suorum et tamen sciret se esse, et quia unum aliquid est, quamvis non sciret illa omnia (quod autem nescitur non est ipsum quod scitur).”  See also the first iteration of the ‘Flying Man’ in I.1 I.1 [Van Riet 1:36–37]: “[N]on enim dubitabit affirmare se esse, nec tamen affirmabit exteriora suorum membrorum, nec occulta suorum interiorum nec animum nec cerebrum, nec aliquid aliud extrinsecus, sed affirmabit se esse, cuius non affirmabit longitudinem nec latitudinem nec spissitudine.” 

2 As Gilson has pointed out, Avicenna’s Latin followers do not take themselves to be primarily “Avicennists”; rather, they adopt Avicenna’s idea inasmuch as they took him to be confirming and clarifying ideas they already had from Augustine: “Are we not witnessing the grafting of Augustinianism on the trunk of Avicenna?” (Gilson, “Introduction” to “The Treatise De anima of Dominicus Gundissalinus,” ed. J.T. Muckle, Medieval Studies 2 [1940]: 26).  For analysis of the “Flying Man” in these authors, see Hasse, Avicenna’s De anima in the Latin West, 80-92, 236.

3 I use the term “subconscious” rather than “unconscious” because these authors universally held that this ever-running self-knowing can be raised to a conscious level by turning one’s attention away from the sensed universe and into the soul.

4 DeTrin 14.8.9 [CCSL 50A.432]: “Sed quoniam mentem semper sui meminisse semperque se ipsam intellegere et amare, quamuis non semper se cogitare discretam ab eis quae non sunt quod ipsa est, quaerendum est quonam modo ad cogitationem pertineat intellectus, notitia uero cuiusque rei quae inest menti etiam quando non de ipsa cogitatur ad solam dicatur memoriam pertinere”; DeTrin 14.6.9 [CCSL 50.332]: “Mentem quippe ipsam in memoria et intellegentia et uoluntate suimetipsius talem reperiebamus ut quoniam semper se nosse semperque se ipsam uelle comprehendebatur, simul etiam semper sui meminisse semperque se ipsam intellegere et amare comprehenderetur, quamuis non semper se cogitare discretam ab eis quae non sunt quod ipsa est”; DeTrin 10.5.7, 10.12.19, and 15.15.25 (using scire rather than nosse).  See Booth, “Augustine’s notitia sui.”

5 See texts in note 2 below.  These types of self-knowledge were compiled into a standard list, which appear (with some variation) in a number of mid-thirteenth-century commentaries on the Sentences, usually in Book I d. 3 and Book III d. 23.  Discernere typically found its way into the list, perhaps because of DeTrin 10.9.12 [CCSL 50.325]: “Non itaque uelut absentem se quaerat cernere, sed praesentem se cures discernere. Nec se quasi non norit cognoscat, sed ab eo quod alterum nouit dinoscat. Ipsum enim quod audit: ‘Cognosce te ipsam,’ quomodo agere curabit si nescit aut quid sit cognosce aut quid sit te ipsam? Si autem utrumque nouit, nouit et se ipsam . . . Sed cum dicitur menti: ‘Cognosce te ipsam, eo dictu quo intellegit quod dictum est te ipsam cognoscit se ipsam, nec ob aliud quam eo quod sibi praesens est.”  Among medieval followers of Augustine, there is some variation in how the terms were defined.  See Alexander of Hales, Glossa Sent. I.3, no. 44 [Quar. 1.62], distinguishing noscere, intellegere, cogitare (equated with discernere—which is replaced by cernere and defined somewhat differently in his Summa theologica, Tract. Int., q. 2, no. 15 [Quar. 1.25]); Albert’s Sent.  I.3.G.27 [Borgnet 25.126], distinguishing cogitare, discernere, nosse, and velle; as well as Albert’s De homine, “De his mot.,” 6.2.2.1, ad 6 [Col. 27/2.547-550], distinguishing nosse, intelligere, and cogitare (equated with discernere); and Aquinas, Sent. I.3.4.5 [Mand. 1.122], distinguishing cogitare, intelligere, discernere with definitions basically identical to those in Albert’s Sent., cited above (Mandonnet mistakenly refers these distinctions to Augustine’s De utilitate credendi, cap. 11).  Interestingly, in the parallel text from his own commentary on the Sentences (I.3.2.2.1), Bonaventure jettisons these traditional distinctions in favor of a distinction between actual and habitual self-knowledge; but he does later return to a distinction between se nosse, se cogitare, and se discernere/distinguere in his Sent. III.23, dub. 4.  Along similar lines, William of Auvergne uses a distinction between “thinking about the soul” (cogitare) and a self-knowledge that guarantees the impossibility of self-ignorance (Summa de anima 3.12–13 [Paris 2.102-4]).

6 See for instance Ia.93.7, ad 4.

Aquinas is not alone in arguing that the soul has a habit of self-knowledge and no supraconscious self-knowing; I have found one other, namely Bonaventure, Sent. I.3.2.2.1, ad 1 [Quar. 1.89–90]: “Quendam enim habitum habet animae potentia ab acquisitione, quendam ab innata dispositione, tertium habet a sui ipsius origine . . . [Anima est facilis] ad diligendum se ipsum per sui naturalem originem. . . . Similiter, cum intellectus noster semper sit sibi praesens, semper est habilis sibi ad se cognoscendum”;  Bonaventure thus collapses Augustine’s semper se intelligere into se nosse (see Sent. III.27, dub. 3), as the 12th-century Cistercians had done (see note the pure, incorporeal being of the soul.30).  But Bonaventure and Aquinas’s theories diverge in that Bonaventure holds that actual self-knowledge is not sense-dependent: the soul can turn its attention inward to think about itself whenever it pleases, without any assistance from the senses. See Bonaventure’s Sent. III.33.1: “Et si tu obiicias, quod concomitantia non sufficit, quia non conceditur, quod musica sit alba; dicendum, quod concomitantia non sufficit simpliciter, sed concomitantia cum simplicitate subiecti, per quam subiectum potest reflectere se super se et super omne illud, quod est in se; sicut patet, quod anima intellectiva simpliciter potest se convertere super se per actum intelligendi et similiter per actum diligendi.; Bonaventure, Sent. II.39.1.2 [Quar. 2.904]: “Si qua autem sunt cognoscibilia, quae quidem cognoscantur per sui essentiam, non per speciem, respectu talium poterit dici conscientia esse habitus simpliciter innatus, utpote respectu huius quod est Deum amare et Deum timere. . . . Quid autem sit amor et timor, non cognoscit homo per similitudinem exterius acceptam, sed per essentiam; huiusmodi enim affectus essentialiter sunt in anima.  Ex his patet responsio ad illam quaestionem, qua quaeritur, utrum omnis cognitio sit a sensu.  Dicendum est, quod non.  Necessario enim oportet ponere, quod anima novit Deum et se ipsam et quae sunt in se ipsa, sine adminiculo sensuum exteriorum.”

7 See for instance Deborah Brown, “Aquinas’s Missing Flying Man,” Sophia 40 (2001): 17-18.

8 See DV 1.9 (obliquely referenced in Sent. III.23.1.2, ad 3) and Sent. IV.49.3.2.

9 Note that not all of Albert’s texts have been indexed

10 See Hasse, Avicenna’s De anima in the Latin West, 60–69.  Hasse notes that explicit references to Avicenna are more common in Albert’s early writings such as De homine (written approx. 1241-1242), while later writings such as De anima (1254–57) are still pervaded by Avicennian influence, but with fewer explicit references; later in life, too, Albert is more willing to critique Avicenna instead of trying to harmonize him with Aristotle.

11 See Black and Kaukua.

12 For secondary literature on the ‘Flying Man’, see M.E. Marmura, “Avicenna’s ‘Flying Man’ in Context,” Monist 69 (1986): 383–95; T.-A. Druart, “The Soul and Body Problem: Avicenna and Descartes,” in Arabic Philosophy and the West: Continuity and Interaction, ed. id. (Washington, D.C.: Center for Contemporary Arab Studies, Georgetown Univ., 1998), 27–48; S. Pines, “La conception de la conscience de soi chez Avicenne et chez Abū’l-Barakāt al-Baghdādī,” Archives d’histoire doctrinale et littéraire du moyen âge 29 (1954): 21–56; Jari Kaukua, “Avicenna on Subjectivity: A Philosophical Study” (Ph.D. Diss., University of Jyväskylä, 2007); Deborah Black, “Avicenna on Self-Awareness And Knowing that One Knows,” in The Unity of Science in the Arabic Tradition, ed. S. Rahman, T. Hassan, T. Street (Dordrecht: Springer Science, 2008), 63–87; Luis Xavier López Farjeat, “Self-Awareness (Al-Shu‘Ūr Bi-Al-Dhāt) in Human and Non-Human Animals in Avicenna’s Psychological Writings,” in Oikeiosis and the Natural Bases of Morality: From Classical Stoicism to Modern Philosophy, ed. Alexander Vigo (New York: Georg Olms Verlag, 2012), 121 – 40; A. Hasnawi, “La conscience de soi chez Avicenne et Descartes,” in Descartes et le moyen âge, ed. J. Biard and R. Rashed (Paris: Vrin, 1997): 283-91; and Dag Nikolaus Hasse, Avicenna’s De anima in the Latin West (London: The Warburg Institute, 2000).

13 Note that his theory of self-knowledge is much more explicitly developed in other works such as the Notes, the Directives, and the Investigations, but I will not refer to these here, as the Latin medievals had no access to them.

14 AviLat LdA V.7: “Repetamus autem id quod praediximus, scilicet quod si subito crearetur homo, expansis eius manibus et pedibus, quae ipse nec videret nec contingeret nec ipsa se contingerent nec audiret sonum, nesciret quidem esse aliquid membrorum suorum et tamen sciret se esse, et quia unum aliquid est, quamvis non sciret illa omnia (quod autem nescitur non est ipsum quod scitur).”  See also the first iteration of the ‘Flying Man’ in I.1[Van Riet 1:36–37]: “[N]on enim dubitabit affirmare se esse, nec tamen affirmabit exteriora suorum membrorum, nec occulta suorum interiorum nec animum nec cerebrum, nec aliquid aliud extrinsecus, sed affirmabit se esse, cuius non affirmabit longitudinem nec latitudinem nec spissitudine.” 

15 Avicenna’s reasoning is as follows: the “self” cannot be the whole body or a single body part, because then self-awareness would just be an experience of the existing body or that existing body part.  And thus if I lost sensory access to that body part or to the body as a whole, then I could not be aware of my own existence.  But the Flying Man has no sensory access to any part of his body—and yet he is still certain of his own existence.  Hence the “self” that the Flying Man experientially accesses cannot be the whole body or any part of the body. 

16 AviLat LdA V.7: “Unde verum est dici quod ‘quia sensimus, concupivimus’ et ‘quia vidimus hoc et hoc, offensi sumus.’  Illud autem unum in quo coniunguntur hae virtutes est id per quod cognoscit unusquisque quae sit sua essentia, ita ut sit verum dicere quod ‘quia sensimus, concupivimus.’”

17 AviLat LdA V.7 [Van Riet, 164]: “Intentio autem de eo quod cognosco quod sit ego est id quod designo mea dictione cum dico “sensi,” “cognovi,” “feci”: quae proprietates coniunctae sunt in uno quod est ego.”

18 These treatises are representative of the kind of approach to the soul prior to Avicenna, as documented in Avicenna’s De anima in the Latin West, 9-12.  For See the dating for these works given in Bernard McGinn’s introduction to Three Treatises on Man: A Cistercian Anthropology, ed. id. (Kalamazoo, Mich.: Cistercian Publications, 1977), 28 and 48.  One could also add the pseudo-Augustinian Liber de spiritu et anima (possibly 1170-90; see McGinn, 63-67), but its dating is uncertain, and not enough work has been done in this area to know whether the author may have been already influenced by Avicennian themes (although this seems unlikely).  See ibid. for the scholarly disagreements about authorship and dating of the Liber de spiritu et anima, which was attributed to Augustine and others by its medieval readers.  It is often now assumed to be the work of Alcher of Clairvaux, but this attribution has been challenged by Raciti, who believes it to be the work of Peter Comestor (see his Ailred of Rivaulx: De Anima). 

19 The history is more complicated than this, though, because of the intermingling of the Liber de spiritu et anima, as well as the Liber de causis.  Perhaps the bridge figures here are Liber de spiritu et anima (popularized by Philip the Chancellor) and Avicenna, who are close enough to link up these two traditions and effect a shift.  This point needs further study.

20 For a brief overview of earlier medieval psychological literature on the soul (a generally under-studied topic), see Hasse, Avicenna’s De anima in the Latin West¸ 9-12.

21 See Isaac of Stella 20, cited in note so that the soul becomes ignorant of itself.27 below.

22 These theses may not provide the most accurate representation of Avicenna’s own theory of supraconscious self-knowledge, but one can see how they could be easily derived from a reading of the Liber de anima.

23 See AviLat LdA V.7, in note .16 above.

24 Here one must be careful to distinguish the immediate impression that the “Flying Man” gives, from the more precise interpretation of Avicenna’s thought that is available in connection with his other texts that were not translated into Latin.  For instance, I suspect that the “Flying Man” gave medieval readers such as William of Auvergne the impression that in stripping away sensory distractions, the soul is able to bring its pre-existing supraconscious self-knowing into the sphere of conscious thought, making explicit what is implicit is all knowing.  According to the most recent studies on Avicenna’s theory of self-knowledge, however (Black; Kaukua, 118ff.; Farjeat), I cannot make my supraconscious self-knowing into a conscious self-knowing, because it is by definition non-intentional and non-conscious.  The mental purification achieved by working through the “Flying Man” scenario allows me only to notice that supraconscious self-knowing in a second-order reflexive act (perhaps loosely analogous to the way in which I might notice myself engaged in a non-conscious act of growing).  The Flying Man’s bare self-knowledge is not a conscious recovery of the soul’s basic self-knowing (by definition, one cannot experience what it is like to have non-conscious self-knowing) but a second-order reflexive self-knowing directed toward the naturally self-knowing core of his being.

25 AviLat LdA V.2 [Van Riet 95]: “Nihil autem horum debet obiici ad hoc ut anima apprehendat seipsam : anima enim apprehendit essentiam suam semper, quamvis plerumque apprehendit eam separatam a corporibus cum quibus est ipsa, sicut iam ostendimus.”  Van Riet notes that the “separatam a” in the Latin translation is the reverse of the meaning given in the Arabic, which actually says “conjoined to” (note 90). See also V.7 [164]: “Si quis autem dixerit: nescis quia hoc anima est, dicam me semper scire, et quia haec est intentio quam voco animam, sed fortassis nescio illam appellari animam; cum autem intellexero illud vocari animam, intelligam illud esse hoc, et quia ipsum est regens instrumenta moventia et apprehendentia: unde nescio quod illud ego sit anima, dum nesciero quid sit anima.” As Black, “Avicenna on Self-Awareness,” 23, puts it: “If the soul is essentially immaterial and rational, then there can be no point in its existence at which it is not in some sense actually cognitive. To the extent that the human soul is truly an intellective soul, it must have the characteristic property of all subsistent intellects, that of being actually intelligible to itself.”

26 See Jean de la Rochelle, Summa de anima, I, c.1 [Bougerol, 51]; William of Auvergne, De anima 1.4 [Paris 2.68] and 2.13 and 3.12-13 [Paris 2.102-104]; Albert, In De anima III.2.17 [Col. 203].  Discussions of the impossibility of self-ignorance in the first three cases follows a discussion of the “Flying Man”, accompanied by references to the pseudo-Augustinian Liber de spiritu et anima (see note 34 below).  This argument is rejected in Bonaventure, Sent. I.3.2.1 and I.3.2.2 ad 1; and Aquinas.

27 Isaac of Stella 20: “The eye of intellectus and intelligentia [have been weakened by iniquity] so that it sees almost nothing.  For who sees himself?  Who knows himself?  I refer to the terms ‘image and likeness of God’ and ‘according to the image and likeness of God.’  He who knows one or the other of similar terms, knows both; and he who does not know one or the other, knows neither.  Therefore, the soul which through itself ought to know God abo ve itself, has lost the capacity to know itself in itself, and the angel on a par with it”; Lib de spir et an 44 [PL 812]: “Mens tamen rationalis, quoniam non se semper cognitat, sicut sui semper meminit, liquet quia cum se cogitat, verbum ejus nascitur de memoria.  Unde apparet quia si semper se cogitaret, semper verbum ejus de memoria nasceretur”; see also cc. 31 [PL 801]: “Idcirco ut homio incognitus cognoscat se, magna opus habet consuetudine recedendi a sensibus, ut animum ad se colligat, et in se ipso retineat.  His siquidem sensibus impeditur anima, ne cernere semetipsam valet et Creatorem suum, quem sola et simplex sine istis oculis intueri debet”; and 32 [801-2].

28 This thesis is probably not one that Avicenna holds himself, however.  Kaukua and Farjeat have pointed out recently that for Avicenna, the souls of animals also have this supraconscious self-knowing, although they cannot become aware of it or reflect on their acts of knowing since such souls are not incorporeal.  If this is correct, then Avicenna does not hold that incorporeality as such is the necessary condition for enjoying supraconscious self-knowing (though it may be a sufficient condition).

29 Hasse, Avicenna’s De anima in the Latin West, 88-90, with reference to Gundissalinus and William of Auvergne.  There has been some discussion recently about whether Avicenna intended the “Flying Man” as an indication of the soul’s incorporeality or merely as an indication of its distinctness from the body, or as an indication of its mode of self-knowing: see Hasse (who holds that the “Flying Man” indicates the soul’s distinctness from the body) and Druart and Black. 

30  See William of St-Thierry, De natura corporis et anima 11: “The soul realizes that it sees incorporeal things of itself, corporeal things through the body.”  A more thorough discussion appears in Liber de spiritu et anima 2 [PL 781]: “Visibilia per corpus videt, invisibilia per se, et in eo se videt, quod invisibilem se esse videt. . . Animus corporis dominator, rector, habitator videt se per se: per se ipsum semetipsum videt.  Non quaerit auxilium corporalium oculorum, imo vero ab omnibus corporis sensibus tanquam impedientibus et perstrepentibus abstrahit se ad se, ut videat se in se, ut noverit se apud se.”  See also the often-cited text from c. 4 [PL 781-2]: “Cognoscit siquidem Deum supra se, et se in se, et angelum juxta se, et quidquid coeli ambitu continetur infra se.”  It is typical to find these authors arguing for a special cognitive “power” in the soul, i.e., intellectus, whereby it cognizes purely incorporeal creatures as itself and angels (contrasted with sensu, imaginatio, and ratio for cognizing material objects and their essences, and intelligentia for cognizing God); see Isaac of Stella18: “By means of sense knowledge then, as has been said, the soul perceives bodies; by imagination, that which is scarcely a body, by reason, the scarcely incorporeal; by discernment, the truly incorporeal”; Lib de spir et an 11 [PL 787]: “Sic igitur anima sensu percipit corpora, imaginatione corporum similitudines, ratione corporum naturas, intellectu spiritum creatum, intelligentia spiritum creatum.

The view that because the soul is incorporeal, it only needs to turn its attention inward in order to cognize itself, is probably traceable to Augustine’s DeTrin 9.3.3 [CCSL 50.296]: “Mens ergo ipsa sicut corporearum rerum notitias per sensus corporis colligit sic incorporearum per semetipsam.  Ergo et se ipsam per se ipsam nouit quoniam est incorporea.”  This link between incorporeality and self-knowledge would be handled quite differently by Latin Avicennians.  But in the 13th century, it is Bonaventure who retains the integrity of the earlier Cistercian tradition; he does not appear to endorse an innate self-knowing, but he does insist on the soul’s ability to think about itself “sine adminiculo sensuum”  (see texts cited above in note 6).

31 Augustine, De Trin. 9.3.3 [CCSL 50.296]: “Mens ergo ipsa sicut corporearum rerum notitias per sensus corporis colligit sic incorporearum per semetipsam.  Ergo et se ipsam per se ipsam nouit quoniam est incorporea.  Nam si non se nouit, non se amat.”

32 See the texts cited in note 4 above.

33 The maxim that the soul “see incorporeal things by itself” (part of De Trin. 9.3.3) appears in William of St. Thierry, 11, p. 143, but not as part of a discussion of self-knowledge; a spinoff from this text (“The soul which through itself ought to know God above itself has lost the capacity to know itself in itself, and the angel on a par with itself”) appears in Isaac of Stella 20, p. 173, but in order to show how the soul ought to know itself, angels, and God, but does not because of sin; the same idea appears for the opposite reason in the Liber de spiritu et anima 4 [PL 782].  The Liber de spiritu et anima additionally quotes parts of De Trin. 8-10 on the soul’s self-presence and lack of self-ignorance (32): “nec sicut absentem se quaerat, sed velut praesentem se curet cernere et discernere, et intentionem voluntatis . . . Ita videbit quod nunquam se non amaverit, nunquam nescierit”; see also 44), but does not really work up these texts into a developed theory of self-knowledge (this is typical of the text as a whole, which is basically a pastiche of quotes from Augustine and other authors).  Earlier texts seem to have been more interested in Augustinian texts related to the circumincession of powers and the identity of the mental powers with the soul’s essence [Augustine: All of it knows, remembers, and loves and all these together are one mind, possibly De Trin 10.3.6; see William of St. Thierry, De natura corporis et anima 11, p. 144, as well as Peter Lombard, Sentences I.3.2, pp. 72-73].  It is also worth noting that the questions about Augustinian self-knowledge that traditionally arise at specific points in Sentences commentaries in the first half of the 13th-century (namely, I.3, I.17, III.23) are not raised in the corresponding distinctions of Peter Lombard’s original Sentences.  Even the discussion of the imago Trinitatis in Peter’s Sent. I.3 focuses on the circumincession of the powers and their equality with the soul’s essence.

34 De Trin. 9.3.3 (note 31 above) is cited by Alexander of Hales, Sum. Theol. Inq. II, Tract. III, Sect. II, q. 1, tit. II, memb. II, cap. 2, arg 1 contra [Quar. 2.176]; see also Bonaventure, Sent. II, d. 3, p. 2, a. 2, q. 2, arg. 3 (citing other similar Augustinian texts).  De Trin. 10.9.12 (note [Referenced content is missing.] above) is cited by Alexander of Hales., Glossa in Sent. I.3, no. 44, c., resp. [Quar. 62]; Bonaventure, Sent. III, d. 23 “dubia”.  De Trin. 14.6.9 [?? i.e., the “semper texts”] is cited by Albert, De homine 6.2.2.1, ad 6 and 9 (Col. 27/2.547-550); Bonaventure, Sent. III, d. 27, dub. 3; Peter of Tarantasia, Sent. I.3.5.3; and of course Aquinas in multiple places ; also Alexander of Hales?  Bonaventure Sent. I, d. 3, p. 2, a. 2?  And then there are the “naturally inserted habits” texts: see Bonaventure QDMT, q. 1, a. 1 and others.

Note that authors who take the Liber de spiritu et anima to be the work of Augustine frequently cite its bastardized version of De Trin. 9.3.3: “Anima cognoscit Deus supra se, se in se, angelus juxta se…”  See Jean de la Rochelle, Summa de anima II, c. 113; id., Tract. De div., pars II, XV, pp. 82-83; Tract. De div., pars II, XX, pp. 90-91;  They also cite the LDSA 32’s claim that the soul cannot be ignorant of itself; see Jean de la Rochelle, Summa de anima, I, c.1 (p. 51, where the reference to the impossibility of self-ignorance is cited immediately after his description of the “Flying Man”); William of Auvergne, De anima 1.4 [Paris 2.68] and 2.13 and 3.12-13 [Paris 2.102-104, both times following a discussion of the “Flying Man”].  This is just a sampling; I have not completed a review of all the relevant texts.

35 There is even there a version of Avicenna’s thought-experiment in Augustine’s DeTrin 10.10.14: “Sed quoniam de natura mentis agitur, remoueamus a consideratione nostra omnes notitias quae capiuntur extrinsecus per sensus corporis, et ea quae posuimus omnes mentes de se ipsis nosse certasque esse diligentius attendamus. Utrum enim aeris sit uis uiuendi, reminiscendi, intellegendi, uolendi, cogitandi, sciendi, iudicandi; an ignis, an cerebri, an sanguinis, an atomorum, an praeter usitata quattuor elementa quinti nescio cuius corporis, an ipsius carnis nostrae compago uel temperamentum haec efficere ualeat dubitauerunt homines, et alius hoc, alius illud affirmare conatus est. Viuere se tamen et meminisse et intellegere et uelle et cogitare et scire et iudicare quis dubitet?”  This text appears as part of an extended reflection that begins from 10.10.12 (where, incidentally, the discussion of the Delphic precept “Know thyself” appears—this discussion contains many of the Augustinian texts on perpetual self-understanding that are most frequently cited in support of supraconscious self-knowing by post-Avicennian Latin medievals).

These parallels have led to speculation about a common source.  Both authors are certainly rooted in the Greek Neoplatonic tradition, and many common elements in their thought can be traced back to Plotinus, but it remains unknown whether they share another more specific common source.  For discussion, see Gilson, “Les sources gréco-arabe de l’augustinisme avicennisant,” Archives d’histoire doctrinale et littéraire du moyen âge 4 (1929–30): 5–149.  Sorabji has proposed Porphyry (The Philosophy of the Commentators, 168–69).  Regarding the medieval acceptance of Avicenna as a clarifier of Augustine, since the dating of Liber de spiritu et anima is uncertain, I also wonder about the possibility of a feedback loop, if this text turns out to have been influenced by Avicennian ideas itself.  If not, the LDSA is so close to many Avicennian themes that it shows a medieval theory that is ripe for conversation with Avicenna.

36 Text discussed below.

37 See Gardeil, “La perception expérimentale,” 225, n. 3; De Finance 31-34; Picard 55-56;  Romeyer 52-53; Lonergan 92 (mentioned in Lambert). The only author I know who has recognized that this is not just an early, unclear endorsement of habitual self-knowledge is D. Juvenal Merriell, To the Image of the Trinity: A Study in the Development of Aquinas’ Teaching (Toronto: Pontifical Institute of Mediaeval Studies, 1990), 73–76.  Lambert expresses uncertainty about how this text should be read, and gives plausible arguments on both sides in Self-Knowledge in Thomas Aquinas: The Angelic Doctor on the Soul’s Knowledge of Itself (Bloomington, Ind.: AuthorHouse, 2007), 111-12.

38 Note that in Sent. I.3.4, Aquinas distinguishes the powers of memoria and intelligentia to preserve the Trinitarian image in the mind.  Responding to an objection that Aristotle only identifies two powers, intellect and will, he argues: “Philosophi accipiebant potentias illas tantum quae ordinantur ad aliquem actum. Proprietas autem retentiva ipsius animae non habet aliquem actum; sed loco actus habet hoc ipsum quod est tenere; et ideo de memoria sic dicta non fecerunt mentionem inter potentias animae” (Sent. I.3.4.1, ad 3).  In DV 10.2, Aquinas will state that intellectual memory is not a separate power of the soul, but rather ‘memory’ refers to intellect under its retentive aspect.  Perhaps this is why in Sent. I.3.4 he tends to use the term intelligentia rather than intellectus in this context, emphasizing the Augustinian terminology.  For background on the two Augustinian mental trinities, see Sent. I.3.5.1: “Et ita haec assignatio [=mens, notitia, amor] sumitur secundum essentiam et habitus consubstantiales; praedicta [=memoria, intelligentia, voluntas] autem secundum potentias. Unde in ista non est tanta conformitas sicut in praedicta, nec ita propria assignatio.”

39 Sent. I.3.4.4: “Expressa autem repraesentatio est in ipsis potentiis propter quinque. Quorum duo se tenent ex parte ipsius animae, scilicet consubstantialitas et distinctio potentiarum, et ideo se habent indifferenter respectu quorumlibet objectorum; alia vero tria, scilicet aequalitas, et ordo, et actualis imitatio respiciunt objecta, unde se habent diversimode respectu diversorum objectorum. Potest autem attendi in potentiis animae duplex aequalitas, scilicet potentiae ad potentiam et potentiae ad objectum. Et haec secunda aequalitas salvatur hic diversimode respectu diversorum objectorum.”

40 Note that for Aquinas, the Trinitarian order is also preserved when God is the object of memory, intelligence, and will, but I here set aside the problem of the soul’s natural knowledge of God.

41 Sent. I.3.4.4 [Mand. 1.120]: “In illis enim quae per habitum acquisitum discuntur, non servatur ordo, ut dictum est supra, qu. 2, art. 3, quia intelligendi actus praecedit actum memorandi; et ideo non est ibi actualis repraesentatio ipsius Trinitatis, secundum quod intendit illis objectis quae non exprimunt Trinitatem. Servatur autem ibi aequalitas quaedam, scilicet potentiae ad potentiam: quia quaecumque comprehenduntur una potentia, comprehenduntur alia: non quod quidquid intelligimus, simpliciter velimus; sed aliquo modo in voluntate sunt, inquantum volumus nos ea intelligere: sed non servatur aequalitas potentiae ad objectum: quia res corporales sunt in anima nobiliori modo quam in seipsis, cum anima sit nobilior eis, ut dicit Augustinus.  Si autem considerentur istae potentiae respectu hujus objecti quod est anima, sic salvatur ordo, cum ipsa anima naturaliter sit sibi praesens; unde ex notitia procedit intelligere, et non e converso.”

42 Sent. I.3.4.1, ad 3: “Philosophi accipiebant potentias illas tantum quae ordinantur ad aliquem actum. Proprietas autem retentiva ipsius animae non habet aliquem actum; sed loco actus habet hoc ipsum quod est tenere; et ideo de memoria sic dicta non fecerunt mentionem inter potentias animae.” 

43 See Sent. I.3.4.1, ad 5: “[I]ntelligere et nosse differunt: nosse enim est notitiam rei apud se tenere; intelligere autem dicit intueri.”

44 Sent. I.3.5.1, ad 5: “Ad quintum dicendum, quod habitus est principium elicitivum operationis. Unde, quia memoria non habet per se actum qui sit simpliciter operatio, non respondet sibi aliquis habitus, sed eodem habitu notitia, memoria et intelligentia reducuntur in unam operationem.”  So although there are three mental powers, there are only two mental operations, i.e., the operations of intelligence and will. 

45 Sent. I.3.4.4 [Mand. I.120]: “Si autem considerentur istae potentiae respectu hujus objecti quod est anima, sic salvatur ordo, cum ipsa anima naturaliter sit sibi praesens; unde ex notitia procedit intelligere, et non e converso.  Servatur etiam aequalitas potentiae ad potentiam simpliciter: quia quantum se intelligit, tantum se vult et diligit: non sicut in aliis, quod velit se tantum intelligere, sed simpliciter.”  He goes on to add that it is only when the mind is its own object that the Trinitarian equality of the powers to the object is preserved: “Servatur etiam ibi aequalitas potentiae ad objectum. Servatur etiam ibi actualis imitatio ipsius Trinitatis, inquantum scilicet ipsa anima est imago expresse ducens in Deum.”  The reason is that in remembering, understanding, willing other objects, the mind preserves only “a certain equality” because of a disparity between the aspects under which the intellect and will comprehend such objects.  For instance, in understanding murder, I will to understand murder—but I do not necessarily will murder itself. 

46 Notice that in Sent. I.3.4.4 (cited in note 45 above), Aquinas is treating notitia as the habitual self-presence that is proper to memory.  Aquinas’s reference to “natural presence” echoes Albert’s description of memory as a storehouse wherein is kept the natural habit for knowing God and oneself.  See Albert, De homine, “De vir. mot.,” 6.2.2.2 [Col. 27/2.549:34–37].  And in fact in a later article he calls notitia a “habit” that is the principle for exercising the intellectual operation.See Sent. I.3.5.1, especially ad 5: “Habitus est principium elicitivum operationis. Unde, quia memoria non habet per se actum qui sit simpliciter operatio, non respondet sibi aliquis habitus, sed eodem habitu notitia, memoria et intelligentia reducuntur in unam operationem.”

47 This precise question appears in Sent. I.3.4.5, s.c. 2: “Praeterea, dicit Augustinus quod quidquid est in memoria mea, illud memini. Sed anima et Deus semper est praesens memoriae. Ergo memoria semper est in actu eorum, et similiter est in aliis.”

48 See the divisio quaestionis of q. 4: “quaeritur . . . 5: de ipsis per comparationem ad actum, utrum scilicet semper sint in suis actibus dictae potentiae” and the introduction to the arguments in q. 4, a. 5: “Videtur quod istae potentiae non semper sint in suis actibus respectu horum objectorum, in quibus praecipue attenditur imago.”

49 This only makes sense, because the act of intellect is the missing link in establishing whether all three powers are always in the act of remembering, loving, and willing the mind itself.  Aquinas has already determined in the previous articles that:

(1) The mind is always in the “act” of self-memory, aka it naturally holds itself habitually present to itself (q. 4, a. 4);

(2) The operation of willing and desiring oneself follows necessarily on the operation of understanding oneself (q. 4, a. 4). 

So if Aquinas wants to hold that the mind’s three powers are “always in their act,” he needs to show that the mind’s perpetual self-presence in memory necessarily generates a perpetual act of intellect, i.e., self-understanding; if there is a perpetual self-understanding, self-love will perpetually follow.

50 Sent. I.3.4.5 [Mand. 1.122]: “Respondeo dicendum, quod, secundum Augustinum De util. credendi cap. XI, differunt cogitare, discernere et intelligere. Discernere est cognoscere rem per differentiam sui ab aliis. Cogitare autem est considerare rem secundum partes et proprietates suas: unde cogitare dicitur quasi coagitare. Intelligere autem dicit nihil aliud quam simplicem intuitum intellectus in id quod sibi est praesens intelligibile.”  Mandonnet seems to have inserted the reference to De util. cred. (apparently referring to 11.25 [CSEL 25/1.31–32]), but this is incorrect.  The tradition of commentaries on the Sentences is concerned with an entirely different distinction here.  See Alexander of Hales, Glossa Sent. I.3, no. 44 [Quar. 1.62], distinguishing between noscere, intellegere, and cogitare; and Albert’s De homine, “De his mot.,” 6.2.2.1, ad 6 [Quar. 27/2.547-550], distinguishing between nosse, intelligere, and cogitare; both Alexander and Albert equate cogitare and discernere.  Earlier, however, in his own commentary on Sentences I.3.G.27 [Borgnet 25.126], Albert had distinguished cogitare and discernere, using definitions that Aquinas repeats nearly verbatim here: “Sed tamen aliud est discernere se, et cogitare se, quam nosse se vel intelligere: quia dicit Augustinus quod discernere se, idem est quod ab aliis rebus ostendere se differre, et hoc non semper facit anima: cogitare autem se, est quasi coagitare se, id est, converti super se, comparando naturam mentis ad partes et vires et passiones ipsius, et sic non semper facit mens: et sic patet differentia horum quinque, scilicet, nosse se, intelligere se, velle se, discernere se, cogitare se.”  Interestingly, in the parallel text from his own commentary on the Sentences (I.3.2.2.1), Bonaventure introduces the notion of a habitual self-knowledge, instead of appealing to these traditional sets of distinctions; he does, however, mention a similar distinction between se nosse and se discernere in his Sent. III.23, dub. 4.

51 Sent. I.3.4.5 [Mand. 1.122]: “Dico ergo, quod anima non semper cogitat et discernit de Deo, nec de se, quia sic quilibet sciret naturaliter totam naturam animae suae, ad quod vix magno studio pervenitur: ad talem enim cognitionem non sufficit praesentia rei quolibet modo; sed oportet ut sit ibi in ratione objecti, et exigitur intentio cognoscentis. Sed secundum quod intelligere nihil aliud dicit quam intuitum, qui nihil aliud est quam praesentia intelligibilis ad intellectum quocumque modo, sic anima semper intelligit se et Deum <indeterminate add. Parma [6.43]>, et consequitur quidam amor indeterminatus.” 

52 Note too the distinct definitions of nosse and intelligere in Sent. I.3.4.1, ad 5: “[I]ntelligere et nosse differunt: nosse enim est notitiam rei apud se tenere; intelligere autem dicit intueri.”

53 A similar distinction appears in Sent. I.17.1.4, ad 4: “[Q]uod ad hoc quod aliquid cognoscatur ab anima, non sufficit quod sit sibi praesens quocumque modo, sed in ratione objecti. Intellectui autem nostro nihil est secundum statum viae praesens ut objectum, nisi per aliquam similitudinem ipsius, vel ab ipso effectu acceptam: quia per effectus devenimus in causas. Et ideo ipsam animam et potentias ejus et habitus ejus non cognoscimus nisi per actus, qui cognoscuntur per objecta. Nisi largo modo velimus loqui de cognitione, ut Augustinus loquitur, secundum quod intelligere nihil aliud est quam praesentialiter intellectui quocumque modo adesse.”  One might be tempted to ask whether the early Aquinas might have erroneously viewed habitual knowledge as a sort of nonintentional knowing, but Aquinas is evidently very aware of the distinction between the two, as we can see from the very next article: Sent. I.3.5.1, ad 1: “Ad primum ergo dicendum, quod ad esse habitus intellectivi duo concurrunt: scilicet species intelligibilis, et lumen intellectus agentis, quod facit eam intelligibilem in actu: unde si aliqua species esset quae in se haberet lumen, illud haberet rationem habitus, quantum pertinet ad hoc quod esset principium actus. Ita dico, quod quando ab anima cognoscitur aliquid quod est in ipsa non per sui similitudinem, sed per suam essentiam, ipsa essentia rei cognitae est loco habitus. Unde dico, quod ipsa essentia animae, prout est mota a seipsa, habet rationem habitus. Et sumitur hic notitia materialiter pro re nota; et similiter est dicendum de amore.”

54 And think of how different his responses would be, if he were assuming that the intellect’s perpetual self-understanding were merely a habit for self-knowledge! 

55 Sent. I.3.4.5, arg 2: “Item, philosophus dicit quod non contingit multa simul intelligere. Sed anima quandoque intelligit quaedam alia. Ergo tunc non intelligit simul seipsam et Deum.” [NB: the medievals also quote Avicenna to this effect, but here the source is given as Aristotle]

56 Sent. I.3.4.5, ad 2: “[P]hilosophus loquitur de intelligere, secundum quod est operatio intellectus completa distinguentis vel cogitantis, et non secundum quod hic sumitur intelligere.”

57 Sent. I.3.4.5, arg. 3: “Item, ad hoc quod anima intelligat vel videat, secundum Augustinum requiritur intentio cognoscentis, per quam species cognoscibilis in rem deducatur. Sed quandoque anima intelligit ex intentione se intelligere. Cum igitur non percipiamus nos intelligere semper animam et Deum, videtur quod intellectus noster non semper sit in actu, respectu horum objectorum.”

58 Sent. I.3.4.5, ad 3: “[I]ntentio intelligentis non requiritur ad tale intelligere, sicut dictum est.”  Note that as Aquinas is using it here, intentio has the twofold meaning of voluntariness and consciousness, and it is required for the “specification” of the act.  See Albert’s Sent. I.3.H.29, ad 3 [Borgnet 25.129]: “Ad illud intelligere quod tantum est objici sibi in lumine intellectus agentis, non exigitur voluntas nisi naturaliter consequens, et non secundum actum: sed ad illud quod est discernere utrum intelligibilium ab alio, vel in propria natura, et partibus et differentiis considerare, ad hoc exigitur actualis voluntas utrumque conjungens,” responding to an argument that cites a contradiction in Augustine: “Præterea, Quid est quod dicit Augustinus, quod mens intelligit se, et hoc semper, quia semper sibi adest? Videtur enim Augustinus sibi ipsi esse contrarius: quia supra habitum est, quod in Iibro XI de Trinitate dicit, quod ad visum tria exiguntur, scilicet visum, et videns, et utrumque conjungens intentio voluntatis” (Sent. I.3.H.29, arg 2 [Borgnet 25.128]).  The same point appears in Peter of Tarantasia, who is clearly borrowing from both Albert and Aquinas.

59 Sent. I.17.1.5, ad 4: “Cum enim actus distinguantur per objecta, oportet dicere diversos actus qui terminantur ad objecta diversa. Unde sicut sunt diversi actus quibus intellectus intelligit equum et hominem, ita sunt diversi actus in numero, quo intelligit equum et quo intelligit actum illius sub ratione actus. Nec est inconveniens quod in actibus animae eatur in infinitum in potentia, dummodo actus non sint infiniti in actu. Unde etiam Avicenna concedit non esse impossibile quin relationes consequentes actum animae, multiplicentur in infinitum”; the same dismissal of the infinite regress problem appears, without attribution to Avicenna, in ST Ia.87.3, ad 2.  But could Aquinas have gotten this reference in the Liber de anima?  It appears in the Notes (161), but that wasn’t supposed to have been translated. 

60 Sent. III.23.1.2, ad 3: “Alio modo anima reflectitur super actus suos cognoscendo illos actus esse. Hoc autem non potest esse ita quod aliqua potentia utens organo corporali reflectatur super proprium actum, quia oportet quod instrumentum quo cognoscit se cognoscere, caderet medium inter ipsam potentiam et instrumentum quo primo cognoscebat. Sed una potentia utens organo corporali potest cognoscere actum alterius potentiae, inquantum impressio inferioris potentiae redundat in superiorem, sicut sensu communi cognoscimus visum videre.”  There is no correlate to this argument in the parallel text from Albert’s Sent.

Aquinas uses this argument in a few other places in the Commentary on the Sentences as well as in DV 1.9, often also citing the Liber de causis 15.  It is important to note that the Liber de causis 15 merely provides Aquinas with an authoritative statement that incorporeal beings are capable of self-knowledge.  The argument that a corporeal organ interferes with reflexivity by standing “between” the power and the act is typically Avicennian (see LdA V.2).  An quite different argument based on the incompatibility of reflexivity and 3-dimensional extension appears in Liber de causis 7, but Aquinas does not use this argument here.

For other uses of this Avicennian argument, see Aquinas’s Sent. II.19.1.1, on the immortality of the human soul: “Tertio, quia intellectus intelligit se; quod non contingit in aliqua virtute cujus operatio sit per organum corporale; cujus ratio est, quia secundum Avicennam, cujuslibet virtutis operantis per organum corporale, oportet ut organum sit medium inter ipsam et objectum ejus.  Visus enim nihil cognoscit nisi illud cujus species potest fieri in pupilla. Unde cum non sit possibile ut organum corporale cadat medium inter virtutem aliquam et ipsam essentiam virtutis, non erit possibile ut aliqua virtus operans mediante organo corporali cognoscat seipsam. Et haec probatio tangitur in libro de causis in illa propositione 15: omnis sciens qui scit essentiam suam, est rediens ad essentiam suam reditione completa” (and compare to Albert’s quite different argument in his parallel text, which argues that the soul is separable from the body as a sailor is separable from a ship, and which praises Avicenna’s superior arguments about the immortality of the soul, but without mentioning reflexivity).  See also DV 1.9: “Cognoscitur autem ab intellectu secundum quod intellectus reflectitur supra actum suum, non solum secundum quod cognoscit actum suum, sed secundum quod cognoscit proportionem eius ad rem . . . Cuius ratio est, quia illa quae sunt perfectissima in entibus, ut substantiae intellectuales, redeunt ad essentiam suam reditione completa: in hoc enim quod cognoscunt aliquid extra se positum, quodammodo extra se procedunt; secundum vero quod cognoscunt se cognoscere, iam ad se redire incipiunt, quia actus cognitionis est medius inter cognoscentem et cognitum. Sed reditus iste completur secundum quod cognoscunt essentias proprias: unde dicitur in Lib. de causis, quod omnis sciens essentiam suam, est rediens ad essentiam suam reditione completa. Sensus autem, qui inter cetera est propinquior intellectuali substantiae, redire quidem incipit ad essentiam suam, quia non solum cognoscit sensibile, sed etiam cognoscit se sentire; non tamen completur eius reditio, quia sensus non cognoscit essentiam suam. Cuius hanc rationem Avicenna assignat, quia sensus nihil cognoscit nisi per organum corporale. Non est autem possibile ut organum corporale medium cadat inter potentiam sensitivam et seipsam. Sed potentiae insensibiles nullo modo redeunt super seipsas, quia non cognoscunt se agere, sicut ignis non cognoscit se calefacere.” 

A similar position appears in Sent. I.17.1.5, ad 3, but supported by a different argument, i.e., that powers using corporeal objects are determined to a particular range of objects that does not include their own act, and thus that such powers cannot cognize their own acts.  This particular version of the argument seems to come from Albert, Sent. I.17.C.3, ad 1, who connects it to the necessity of postulating the common sense (he explicitly mentions Avicenna in support of the claim that corporeal powers cannot cognize their own acts, so that the common sense must be postulated as the power that perceives the acts of the five external senses). 

61  Sent. IV.49.3.2: “Secundo, quia, cum delectatio sit in appetitu, et omnis passio vel operatio appetitus praeexigat apprehensionem; oportet quod bonum conjunctum quod delectationem causat, sit apprehensum. Dispositio autem quae est jam inhaerens, a nobis non apprehenditur ita sicut dum est in fieri, ut Avicenna dicit in 6 de naturalibus; unde hectici minus sentiunt calorem febrilem quam alii febricitantes, quamvis in eis febris sit intensior: quia jam calor ille est infusus membris quasi complexio illorum; et hoc ideo quia nostrum sentire et intelligere sunt ex aliqua permutatione intellectus et sensus a suis objectis; ab eo autem quod jam in se quiescit, nihil immutatur. Unde cum esse nostrum et vivere nostrum, et omnes actus proprii insint nobis ut in nobis quiescentes, sola autem operatio insit in nobis ut in fieri existens; in ipsa operatione percipimus et esse nostrum et vivere nostrum; et inde est quod in operando delectamur, et in vivendo et sentiendo, et aliis hujusmodi, in quibus indicatur quodammodo nostrum esse et vivere, ut patet per philosophum in 10 [rather, 9] Ethic.”  See Avicenna, De medicinis cordialibus (appendix to LdA, Van Riet 192-93). There is no parallel to this argument in the corresponding text of Albert’s Commentary on the Sentences, but it may be fruitful to compare to Albert’s Commentary on the Nicomachean Ethics IX, which Aquinas had a hand in transcribing.

62 See Hasse, Avicenna’s De anima in the Latin West.  One should note, though, that not all of Albert’s writings have been indexed, and it may be that there is a reference in one of the unindexed works.

63 In Albert, the text arises from the question of circumincession, i.e., whether all the Trinitarian mental powers are all equally “in” each other (i.e., whether they are all in act with respect to each other); but as in the text (and probably for the same reasons) the solutions focus on showing that the intellect is always understanding itself.

64 Note that a very similar treatment of self-knowledge appears also in Albert’s slightly earlier De homine, dating from 1241-42; see “De anima rationi / De viribus apprehensivis suis,” 6.2.2 [Cologne 27/2.419-22].

65 Albert is the only commentator on the Sentences that I know of who pairs the traditionally “Augustinian” way of answering the question (by defining intelligere as a non-attentive kind of cognition) with a quite different solution, which he attributes to “a certain philosopher.”  Aquinas’s summarizing of both solutions thus provides unmistakeable evidence of his reliance here on Albert.  Note too that Albert’s args 1 and 2 are Aquinas’s args 1 and 3, respectively; Albert’s arg 2 s.c. is Aquinas’s arg 1 s.c. 

66 Albert names Alexander as the source of the same view in De hom. 2.3.3, ad 4 (De gradibus potentiae intellectus possibilis), p. 425-427.

67 Probably Aristotle’s De anima III; see Albert, arg 2 s.c.: “Item, Aristoteles in III de Anima dicit, quod intellcetus agens semper in actu est: et cum sit agens respectu alicujus intelligibilis, erit aliquod intelligibile semper terrninans actionem suam: hoc autem non est nisi quod semper præsens est intellectui, et hoc est intelligibilo quod est ipse intellectus: ergo intellectus possibilis semper illustratur luce intellectus agentis: et hoc facit intelligere possibilem secundum actum: ergo intellectus semper intelligit se.”

68 Albert, Sent. I.3.H.29 [Borgnet 25.130]: “[Augustinus] vult, quod sub omni intelligibili alio intellectus semper intelligat se, et alia duo: et memoria semper meminerit sui, et aliorum duorum : et voluntas semper vult se, et ali duo. Unde sine præjudicio dico secundum prædicta, quod meminisse nihil aliud est quam notitiam rei apud se retinere. Intelligere vero nihil aliud quam notitiæ specie vel essentia notitiæ se intueri simpliciter sine consideratione sui et discretione suæ naturæ, quia hoc vocat Augustinus intelligere se: et discernere se et cogitare se plus secundum eurn dicunt quam intelligere se. Cum igitur intelligere se non ponat discretionem sui, sed simplicem intuitum sine discretione et cogitatione, dico quod non ponit conversionem intellectus ad aliquid quod sit extra ipsum ad quod sui ponat convenientiam vel differentiam, nec aliquam proprietatem suam quæ cogitetur ei inesse vel non inesse.” 

69 Albert, Sent. I.3.H.29 [Borgnet 25.130]: “Hic autem intuitus perficitur duobus scilicet lumine intelligibli emisso, et objecto: et intueri se nihil aliud est quam in lumine illo sibi objici. Cum igitur lumen intellectus agentis semper splendeat super possibile, eo quod agens sit forma et actus possibilis, intellectus possibilis semper objicitur sibi in lumine suo: et cum hoc sit intelligere se secundum Augustinum, semper intelligit se.” 

70 Note that he had already laid out this view of the relationship between agent intellect, possible intellect, and intelligible object in the first half of the respondeo, in explaining the “philosopher’s” claim that the intellect cognizes itself in every intelligible.

71 Albert, Sent. I.3.H.29, ad 4 [Borgnet 25:130-31]: “Ad aliud dicendum, quod nosse se est notitiam apud se habere, vel per speciem vel per præsentiam, et intelligere se est sibi objici in notitia illa et luce intellectus agentis: et divisio quam ponit ratio illa, tenet in discretiva scientia, et non in ista quae secundum naturam talis est, ut diximus”; and note Aquinas’s corresponding assumption that intelligere is a natural knowledge: “Dico ergo, quod anima non semper cogitat et discernit de Deo, nec de se, quia sic quilibet sciret naturaliter totam naturam animae suae” (Sent. I.3.4.5 [Mand. 1.122]).  Also notice Albert’s definitions of nosse and intelligere, which are obviously the source for Aquinas’s; see especially Sent. I.3.G.27 [Borgnet 25.125]: “Unde secundum Augustinum, nihil aliud est nosse quam notitiam rei apud se tenere: cum ergo ex superiori constet hoc convenire memoriæ, patet quod memoria novit se et Deum hae notitia. Intelligere autem se secundum Augustinum nihil ahud est, quam se in notitia illa vel Deum intueri et videre: non enim idem est notitiam rei apud se tenere, et intueri rem cujus est species illa vel notitia: et ideo aliud est intelligere se, et aliud nosse se”; and compare to Aquinas’s functionally identical definitions in Sent. I.3.4.1, ad 5, in note 43 above.

72 Taylor, “Albert the Great and the Development of Thomas Aquinas’ Natural Epistemology,” paper at an Aquinas and ‘the Arabs’ session of the American Catholic Philosophical Association, 2011.  Taylor draws our attention to Albert, De homine [Col. 411:46-53]: “Item, Averroes: ‘Omnis intellectus in nobis existens habet duas actiones. Quarum una est de genere passionis, et est intelligere; alia de genere actionis, et est abstrahere eas a materia, quod nihil aliud est quam facere eas intellectas in actu postquam erant intellectae in potentia’. Cum igitur unum horum sit intellectus agens et alterum possibilis, uterque istorum intellectuum erit in nobis existens et non separata substantia.”

73 It is not clear to me yet whether Albert thought this was in fact Averroes’ view, or the view of one of the authors he reports; since he chronically misread Averroes on the inherence of the agent/possible intellects, it is hard to know exactly what he thought this text meant.  See Averroes, LCDA 3.36 (484): “Sed cum aliquis intuebitur omnes sermones istius viri [viz. Alexander of Aphrodisias] et congregabit eos, videbit ipsum opinari quod, quando intellectus qui est in potentia fuerit perfectus, tunc intelligentia agens copulabitur nobiscum, per quam intelligemus res abstractas, et per quam faciemus res sensibiles esse intellectas in actu, secundum quod ipse efficitur forma in nobis. Et quasi intendit per hunc sermonem quod, intellectus qui est in potentia quando fuerit perfectus et completus, tunc copulabitur cum eo iste intellectus et fiet forma in eo, et tunc intelligemus per ipsum alias res; non ita quod intellectus materialis intelligat ipsum et propter illud intelligere fiat continuatio cum hoc intellectu, sed continuatio istius intellectus nobiscum est causa eius quod intelligit ipsum et intelligimus per ipsum alias res abstractas.” 3.36 (500): “ Quoniam hoc posito, continget necessario ut intellectus qui est in nobis in actu sit compositus ex intellectis speculativis et intellectu agenti ita quod intellectus agens sit quasi forma intellectorum speculativorum et intellect speculativa sint quasi materia.”  And (503): “ Et est etiam manifestum quod, cum posuerimus intellectum materialem esse generabilem et corruptibilem, tunc nullam viam inveniemus secundum quam intellectus agens copuletur cum intellectu qui est in habitu copulatione propria, scilicet copulatione simili continuationi formarum cum materiis.” Also note the language of “conjunction” in the De homine.

Notice the bit about motion that may have inspired the way he thinks about intentionality: Averroes, LCDA 3.5 (401): “Quemadmodum enim subiectum visus movens ipsum, quod est color, non movet ipsum nisi quando per presentiam lucis efficitur color in actu postquam erat in potentia.”

74 See John Blund, Grosseteste.

75 See Gundissalinus, John Blund, Jean de la Rochelle.

76 Dieterich of Freiburg’s adherence to this point was discussed in Jean-Luc Solère’s paper “Intentionality and the activity of the subject in medieval theories of cognition,” American Philosophical Association Congress, Society for Medieval and Renaissance Philosophy session, New York, Dec. 29, 2009.

77 It would also have been reinforced by the newly-translated Liber de causis.

78 Deliberately so, since as noted earlier, Avicenna does not think supraconscious self-knowing is an intellectual activity, but the very being of the soul (but compare Albert, Super Ethica IX.11).

79 Namely that the intellect intermittently cognizes itself in its acts.

80 He may well have developed a more Avicennian concept of supraconscious self-knowing later in life, but that is another story.

81 Notice the setup of sed contra arguments in DV 10.8 (those on behalf of supraconscious self-knowing), which show the audience’s familiarity with Albertist language and an attempt to trace supraconscious self-knowing back to Augustine and Aristotle (not Avicenna).  The first of the sed contra arguments in DV 10.8 is a quote from Augustine’s De Trin. 9.3.3; the second speaks of “intellectual vision”; the fourth argues from intelligible presence to actual understanding; and the last two are citations from Aristotle’s De anima concerning the agent intellect. And see the conclusion of the respondeo: “Sic ergo patet quod mens nostra cognoscit seipsam quodammodo [i.e., habitualiter] per essentiam suam, ut Augustinus dicit: quodam vero modo per intentionem, sive per speciem, ut philosophus et Commentator dicunt; quodam vero intuendo inviolabilem veritatem, ut item Augustinus dicit.”

One may speculate that he was already uncomfortable with Albert’s explanation for supraconscious self-knowing in Sent. I.3.4.5—there, although he presents the first half of Albert’s “Augustinian” solution, he does not repeat Albert’s Aristotelian explanation in terms of the ever-shining light of the agent intellect.  Is this omission the casualty of a quick summary?  Or was Aquinas uncomfortable with Albert’s line of reasoning?  There is no way to know for sure.

82 My suspicion is that Aquinas had not read Avicenna’s Liber de anima himself when he started work on the Commentary on the Sentences, but that he did read it before he had finished, perhaps when addressing human nature in Book II.  In the process of researching this paper, for instance, the following clues emerged, which could point in this direction (although none are conclusive enough to make a final determination).  1) Beginning in Sent. II.19.1.1, Aquinas explicitly cites Avicenna’s argument that a corporeal organ interferes with the reflexivity of the power because the organ falls “between” the power and its act, an argument he repeats without naming Avicenna in Sent. III.23.1.2, ad 3, and again in DV 1.9, this time again naming Avicenna.  2) So far, I have only found in texts of Albert with which Aquinas was probably familiar a named reference to the conclusion that Avicenna draws from this argument (i.e., therefore the senses cannot sense their own acts, hence justifying the existence of a common sense); see Albert, Sent. I.17.C.3 ad 1, and Super Ethica IX.11. 3) In De homine [Col. 460:22-33], Albert does sketch out this Avicennian argument, but without naming a source.  The reference in the apparatus criticus, incidentally, is to Algazel’s Metaphysics.  4) In Sent. IV.49.3.2, Aquinas cites Avicenna’s De medicinis corporibus to explain why the soul only notices its own intentional acts and not, say, its vegetative or nutritive acts.  I have not found this argument mentioned anywhere in Albert.  5) There is also the entirely separate problem of where Aquinas got his knowledge of Avicenna’s dismissal of an infinite regress in acts of knowing that I know.  For some of the relevant texts see note 60 above.

One way to interpret this evidence is that at some point by Sent. II.19, Aquinas had read Avicenna directly and he is now taking from him arguments that Albert had not employed.  But so far I have not reviewed all the texts, and I cannot exclude another equally plausible interpretation: namely, since Aquinas was Albert’s student, it would not be unreasonable for him to have learned from Albert in verbal discussion about the Avicennian arguments that I cannot locate explicitly in Albert’s writings.  Moreover, note that the Avicennian corporeal-organ argument mentioned above appears (with the source left unnamed) in Jean de la Rochelle, Summa de anima II, c. 112.