The Thomist 67 (2003): 505-38 SPIRITUAL COGNITION IN THOMAS AQUINAS PHILIP L. REYNOLDS Emory University Atlanta, Georgia T HOMAS AQUINAS is famous for maintaining that all human knowledge in this life, even of God, begins in the senses. 1 Clearly, such a statement oversimplifies his position. Rapture, prophecy, faith, and self-knowledge surpass or stretch the maxim in various ways. Yet they are not outright exceptions to it. Thomas maintains also that in this life, one knows realities external to the self by means of their likenesses in one's self, such as sensible and intelligible species. But the beatific vision is unmediated, for no likeness in the intellect would be adequate to make known the essence of God. In this article, I shall try to ascertain what Thomas thought about spiritual cognition. In spiritual cognition, one would "see" a superior, immaterial being (the Deity or an angel) by means of an intelligible form that the object itself has impressed directly on one's intellect; and the function of that form would be the same in relation to the knower and to the known as that of intelligible or sensible species in one's direct cognition of created forms. Just as the eye sees the redness of a red apple by means of a sensible species of redness in the eye, so also would the intellect see God by means of an intelligible species that God has impressed on the intellect. 1 I presented an early version of this article at a session on supernatural cognition sponsored by the Midwest Seminar on Ancient and Medieval Philosophy at the 38'h International Congress on Medieval Studies, Kalamazoo, May 2003. I am very grateful to an anonymous reviewer for certain suggestions and criticisms. 505 506 SPIRITUAL COGNITION IN THOMAS AQIBNAS Did Thomas believe that spiritual cognition was possible at all in this life? If one knew God in this way, what would one know? I raise these questions for two reasons. First, Thomas himself posits a purely spiritual way of seeing God, mediated only by infused species, in two early works. 2 But he posits it there only hypothetically, to be excluded as a possible way of knowing the divine essence. Moreover, in one of these texts, he rules out such cognition as a possible way of contemplating God at all in via.3 Second, Thomas attributes spiritual cognition of some sort to Adam before the Fall and even to contemplatives after it. 4 In prelapsarian cognition, God makes God's self known to the mind by means of an interior, spiritual influence, and that influence functions as a mental species by which one knows God. Just as one sees a stone by means of a sensible species of the stone in one's eye, Thomas argues, so Adam knew God by means of an interior, spiritual influence. My purpose in this article is threefold: to consider the hypothetical mediated vision of God; to consider Thomas's account of prelapsarian cognition; and to inquire whether the two modes of cognition are fundamentally the same or different. These are rather arcane topics by modern standards, but inquiry into them highlights some salient features of Thomas's cognitive theory. I shall first outline Thomas's account of cognitive mediation, which is crucial for what follows. 2 IV Sent., d. 49, q. 2, a. l; De Trin., q. 1, a. 2. Editions cited in the notes include the following: Scriptum super libros Sententiarum, ed. P. Mandonnet (books I and II) and M. F. Moos (books III and IV, incomplete), 4 vols. (Paris, 1929-56). For the remainder of the Scriptum (after Book IV, d. 22), I have used the Parma edition of the Opera Omnia (1852-73), repr. New York (1948-50), vol. 7. Compendium Theologiae, in Opuscula, Leonine edition, vol. 42. Quaestiones de Quolibet, Leonine edition, vol. 25. Quaestiones Disputatae De Veritate, Leonine edition, vol. 22. Summa contra Gentiles, Marietti edition, 3 vols. (1961). Summa Theologiae, Ottawa edition, 5 vols. (1941-45). Super Boetium De Trinitate, Leonine edition, vol. 50. Super Epistolas PauliApostoli lectura, Marietti edition, ed. R. Cai, 2 vols. (1953). Super Evangelium s. Ioannis lectura, Marietti edition, ed. R. Cai, (1952). References to Bonaventure are to the Opera Omnia published in Quaracchi (10 vols., 1882-1902). 3 De Trin., q. 1, a. 2. 4 II Sent., d. 23, q. 2, a. 1. De Verit., q. 18, a. 1, ad 1. PHILIP L. REYNOLDS 507 I. COGNITIVE MEDIATION According to Thomas, all knowledge of external realities, apart from the unmediated vision of the divine essence, is obtained through formal representations, or "likenesses" (similitudines), in the knower. Thomas posits four kinds of representative likeness: the sensible species (in a sense organ), the phantasm (a sensory representation of a real or imaginary object in the imagination), the intelligible species (the end result of abstraction), and the concept, or word (a mental definition produced by the possible intellect). 5 It is chiefly with reference to species (sensible and intelligible) that Thomas articulates the modes of cognitive mediation. His account presupposes that it makes sense to speak of the intellect's knowing things, as well as its knowing that propositions are true. Indeed, that is its primary operation. What the sense organs know are simple accidental forms, such as external colors. And what the intellect knows in the first place are material quiddities, considered in abstraction from their material conditions and thus universally. Such simple apprehension is not a judgment about any subject, but, according to Thomas, it does 5 For summary accounts of Thomas's theory of cognition, see Georges Van Riet, "Le theorie thomiste de la sensation externe," Revue philosophique de Louvain 51 (1953): 374-408; idem, "La theorie thomiste de !'abstraction," Revue philosophique de Louvain 50 (1952): 353-93; and Edward P. Mahoney, "Sense, Intellect, and Imagination in Albert, Thomas, and Siger," in N. Kretzmann, A. Kenny, and Jan Pinborg, eds., Cambridge History of Later Medieval Philosophy (Cambridge: Cambridge University Press, 1982), 602-22, at pp. 605-11. See also Eleonore Stump, "Aquinas's Account of the Mechanisms of Intellective Cognition," Revue intemationale de philosophie 25 (1998): 287-307. For fuller accounts, see Robert Pasnau, Theories of Cognition in the Later Middle Ages (New York: Cambridge University Press, 1997); and idem, Thomas Aquinas on Human Nature: A Philosophical Study of 'Summa theologiae' Ia 75-89 (Cambridge: Cambridge University Press, 2002), chaps. 6, 9, and 10. John F. Peifer, The Concept in Thomism (New York: Bookman, 1952), reprinted as The Mystery of Knowledge (Albany, N.Y.: Magi, 1984), remains a useful account, albeit one reflecting developments in John of St. Thomas and later Thomism. Peifer focuses on the concept, or mental word. On intelligible species, see Leen Spruit, "Species intelligibilis": From Perception to Knowledge, vol. 1, Classical Roots and Medieval Discussions (Leiden: Brill, 1994), especially pp. 156-74 (on Thomas); and Katherine H. Tachau, "Some Aspects of the Notion of Intentional Existence at Paris, 1250-1320," in Sten Ebbesen and Russell L. Friedman, eds., Medieval Analyses in Language and Cognition (Copenhagen: Royal Danish Academy of Science and Letters, 1999), 331-53 (an important article, notwithstanding some polemics). 508 SPIRITUAL COGNITION IN THOMAS AQUINAS have a certain truth value. Indeed, it is infallible. 6 In what sense such quiddities are external to the inteHect is a question that need not detain us here. In the disputed questions De Veritate and the seventh Quodlibet, which date from his first Parisian regency (1256-59), Thomas distinguishes three distinct modes of cognitive mediation. As usual, sight is the paradigm and the master metaphor that he uses to analyze cognition. 7 First, there is the means "under which" (medium sub quo) one sees. This causes what is potentially knowable to become actually knowable. Physical light is the means under which the eye sees colors, and the light of the agent intellect is the means under which the inteHect knows quiddities. Second, there is the means «by which" (medium quo) one sees, such as a sensible or inteHigible species. Third, there may be a means "in which" one knows something, or "from which" one receives knowledge (medium in quo, medium a quo), which Thomas likens to a mirror, or mirror image (speculum). 8 The distinction between the second and third forms of mediation, which one might characterize respectively as "formal" and "objective," is crucial here. 9 Thomas explains the distinction in the De Veritate as foHows: [I]n corporeal vision ... the medium by which [quo] the object is seen is the species itself of the sensible thing present in the eye, which, as the form of the one who sees inasmuch as he sees, is the principle of the visual operation. An example of the medium from which [a quo] one receives cognition of the seen object is the mirror from which the species of some visible thing, such as a stone, may sometimes come to the eye, rather than immediately from the stone itself. And these ... are found also in intellectual vision .... The intelligible species, STh I, q. 85, a. 6. On the philosophical implications of visual paradigms in theology, see Reijo Tyorinoja, "Fides et visio: On Visual Metaphysics of Knowledge and Religious Belief in the Middle Ages," in Arch iv fur mittelalterliche Philosophie und Kultur, vol. 6, ed. T. Boiadjiev et al. (Sofia: UK, 2000), 115-31. 8 Quodl. 7, q. 1, a. 1. De Verit., q. 18, a. 1, ad 1. Thomas characterizes the medium as that in quo in Quodlibet 7 and as that a quo in the De Veritate. The word speculum can mean either "mirror" or "mirror image," and it is often impossible to know in a given context which is the appropriate translation. 9 Thomas says that the intellect uses intelligible species formaliter in Quodl. 5, q. 5, a. 2, ad 1. 6 7 PHILIP L REYNOLDS 509 by which [qua] the possible intellect is caused actually to understand, corresponds to the visible species. And an effect from which [a quo] one comes to know a cause is comparable to the medium from which one receives knowledge of a visible object, as from a mirror. In such cases, the likeness of a cause is impressed on our understanding not immediately by the cause, but by the effect, in which the likeness of the cause is reflected. Hence cognition of this sort is said to be specular, because of its similarity to vision that comes through a mirror. 10 In the seventh Quodlibet, Thomas points out that it is only when there is a medium a quo-that is, an intermediate objectthat cognition is said to be indirect (mediata). Someone who is looking at a stone is said to see it directly even though such vision requires not only light as the medium sub quo but also the received sensible species of the stone as the medium quo. 11 Thomas does not regard the medium quo as such as an object of knowledge. Indeed, it is only by reflecting upon the process of cognition that one becomes aware of its existence. 12 When I see the redness of an apple, what I see is the external redness, and not the sensible species by which I see it. Because no likeness in the intellect would be sufficient to make known the essence of God, Thomas argues, one cannot know that essence unless God joins God's self to the intellect as an intelligible form. 13 In the beatific vision, therefore, there is neither a medium a quo nor a medium quo, but only a medium sub quo, De Verit., q. 18, a. 1, ad 1: "in visione corporali ... medium vero quo videtur est ipsa species rei sensibilis in oculo existens quae sicut forma videntis in quantum est videns principium est visivae operationis, medium autem a quo accipitur cognitio rei visae est sicut speculum a quo interdum species alicuius visibilis, ut puta lapidis, fit in oculo non immediate ab ipso lapide. Et haec ... etiarn in visione intellectuali inveniuntur .... speciei vero visibili [respondeat] species intelligibilis qua intellectus possibilis fit actu intelligens, rnedio vero a quo accipitur visi cognitio sicut speculo cornparatur effectus a quo in cognitionern causae devenirnus, ita enirn similitudo causae nostro intellectui imprimitur non immediate ex causa sed ex effectu in quo similitudo causae resplendet: uncle huiusmodi cognitio dicitur specularis propter similitudinem quam habet ad visionem quae fit per speculum." I have eliminated treatment of the first mode of mediation. 11 Quodl. 7, q. 1, a. 1. 12 STh I, q. 85, a. 2. See Leen Spruit, Species Intelligibilis, 1:159-60 (with the references there given). 13 STh I, q. 12, a. 5. See also STh I, q. 12, a. 9; STh I-II, q. 3, a. 8; IV Sent., d. 49, q. 2, a. 1; ScG III, c. 51. 10 510 SPIRITIJAL COGNITION IN THOMAS AQUINAS namely, an infused light that enhances or perhaps takes the place of the agent intellect and enables the soul to be united with God and thereby to see God's essence. 14 According to Thomas, it is precisely because that light, while supernaturally bestowed, is a created form that even the beatific vision is not comprehensive. 15 The blessed do not know God to the extent that God is knowable, and therefore they apprehend rather than comprehend God. Nevertheless, they do see the divine essence. The medium quo and the medium sub quo limit cognition in different ways. Rapture, too, according to Thomas, is an unmediated vision of the divine essence. Rapture is not an exception to the rule that one cannot know God's essence in this life because the rapt person is temporarily removed from this life. 16 Even dimly specular knowledge of God can be construed as vision of a sort: one sees God through a mirror and obscurely. And persons who have some special, revelatory insight that is mediated externally or by phantasms may properly be said to see God. 17 Nevertheless, the term "vision" in its strictest sense characterizes direct rather than inferred knowledge. One may infer the presence of a fox in the woods by its signs, such as its footprints, or one may actually see the fox. But even vision in the strictest sense is usually mediated by species, for "that thing whose likeness exists in the intellect is known to the intellect by way of vision, just as a likeness of something seen corporeally is in the sense of the one who sees. " 18 Knowledge of God achieved simply through infused species would amount to vision precisely inasmuch as the mediation would be formal rather than objective (although, as we shall see, there is a sense in which such vision would be indirect, for what one would see would not be the divine essence). STh I, q. 12, a. 2; I, q. 12, a. 5. STh I, q. 12, a. 7. 16 See appendix, below. 17 In loan., c. 1, lect. 11. 18 ScG III, c. 41: "res enirn ilia per intellecturn visionis rnodo cognoscitur, cuius sirnilitudo in intellectu existit, sicut et sirnilitudo rei corporaliter visae est in sensu videntis." 14 15 PHILIP L. REYNOLDS 511 II. COGNITIVE MIRRORS Thomas characterizes causally inferred, externally mediated knowledge of God as specular. 19 The notion comes from 1 Corinthians 13:12: "Now we see through a mirror obscurely fper speculum in aenigmate ], but then face to face." The term "specular" usually implies that the cognition is dim ("enigmatic") as well as indirect. Yet the metaphor of mirrors is ambiguous. Its use in 1 Corinthians suggests cognition that is dim and indirect, while the function of actual mirrors suggests cognition that is virtually immediate, and in which ideally the medium is hardly noticeable. 20 Thus in the Summa contra Gentiles, discussing angelic cognition, Thomas distinguishes between (a) seeing something as if in a mirror and (b) discursively inferring its existence through its effects. In discursive causal inference, there are two acts of understanding (cognitiones): that by which one knows the effect, and that by which one knows the cause. A reasoned inference separates the two. But when one sees something in a mirror, there is only a single act of cognition, for one grasps both the intermediate object and the ultimate object at once. It is by simple mirror vision of this sort, Thomas argues, and not through discursive inference, that one angel knows God through another angel. 21 In simple mirror vision, according to Thomas, one knows yet need not notice the medium. It becomes what one might call an «unnoticed mirror." In the De Veritate, discussing how the angels know God, Thomas argues that whenever something is visible in its image, one may consider the image either as a thing in itself or precisely as an image; and in the latter case, the motion of the cognitive power toward the image is the same as its motion In loan,, c. 1, lect. 11. See also De Verit., q. 18, a. 1, ad 1. Perhaps medieval mirrors were less transparent than modern ones, but cf. Bonaventure, Collationes in Hexaemeron 5.25, where Bonaventure likens the pure mind to a dear, "polished" mirror, and moreover distinguishes between natural mirrors, made by polishing steel, and artificial mirrors, made by precipitating lead on glass. 21 ScG III, c. 49. 19 20 512 SPIRITUAL COGNITION IN THOMAS AQUINAS toward the object. Likewise, when one knows a cause through an effect, the motion of the cognitive power can proceed to the cause "immediately," so that one does not think about anything else. "And in this way," Thomas concludes, "the intellect of a wayfarer is able to think about God without thinking about any creature. " 22 Such considerations may blur but they do not negate the distinction between a medium quo and a medium a quo. III. THE AVICENNAN MODEL: KNOWLEDGE THROUGH IMPRESSION In two early works-his commentaries on book 4 of Peter Lombard's Sentences and on Boethius's De Trinitate-Thomas considers how one might know God not through species abstracted from sense data or by the mediation of external creatures, but through intelligible species that God has impressed directly on the intellect. In both passages, Thomas cites Avicenna. In his commentary on book 4 of the Sentences, Thomas is inquiring as to how the human intellect can know God's essence. 23 His response is largely a rehearsal of discussions and debates among philosophers about knowledge of separate substances. 24 Thomas explains that the problems met by the theologians regarding knowledge of the divine essence are parallel to problems met by the philosophers regarding quidditative knowledge of separate substances (i.e., of intelligences, which Christians call "angels"). Having shown that no species derived from our sensible experience of material things is sufficient as a 22 De Verit., q. 8, a. 3, ad 18: "Et hoc modo intellectus viatoris potest cogitare de Deo non cogitando de aliqua creatura. » 23 IV Sent., d. 49, q. 2, a. 1. The Scriptum summarizes Thomas's teaching as a sententiary bachelor (1252-56), although he continued to edit it after he became a master of theology in 1256. 24 Apart from the reference to Avicenna, Thomas's response follows a discussion from Averroes' commentary on Aristotle's De anima, book III, regarding knowledge of separate substances, where the Commentator outlines and criticizes the views of Avempace and Alexander of Aphrodisias as well as presenting his own. The point of the discussion is that through such knowledge, the philosopher can attain ultimate happiness. Thomas returns (ca. 1260-64) to Averroes' treatment in ScG III, cc. 41-44, where he devotes a lengthy polemical critique to it, refuting the theories of Alexander and Averroes as well as that of Avernpace. PHILIP L. REYNOLDS 513 means for quidditative knowledge even of created separate substances, let alone of God, Thomas turns to consider knowledge by impressed species. He notes that, according to Avicenna in the Metaphysics, one can understand separate substances through "the intentions of their quiddities," that is, through likenesses that are not abstracted from them but are rather impressed by them. For one cannot intellectually abstract anything from a being that is already immaterial. But Thomas argues that even if one could know God through impression, such knowledge would still fall short of the vision of the divine essence. 25 His reasoning turns on the principle that "what is received by something is present there according to the manner of the recipient." Since the created intellect falls far short of being perfectly like God, any species received by it will be insufficiently like God for quidditative knowledge. Thomas uses Avicenna's opinion as the premise of an objection in the same article. According to Thomas, Avicenna argues that when we know a separate substance, what is in our intellect cannot be the very essence of that substance but is rather an impression of it. But one could not know the divine essence in this way, the argument proceeds, for God is more different from us than "any angel or intelligence." Therefore if this were how we knew God, we would not know the divine essence. 26 Thomas begins his reply by stating that he does not agree with Avicenna, and that other philosophers disagree with him as well. Thomas probably refers here to Alexander of Aphrodisias and Averroes, whose positions on knowledge of separate substances he has summarized, with qualified approval, in the body of the article. For just as they maintain that one knows separate substances when the (separate) agent intellect is united with the human soul as its form, so also, mutatis mutandis, Thomas maintains that in the beatific vision, God will be united to the human intellect as its form. Thomas rejects the position that he attributes here to Avicenna: that the intellect cannot know 25 26 IV Sent., d. 49, q. 2, a. 1. Ibid., obj. 9. 514 SPIRITUAL COGNITION IN lHOMAS AQUINAS another essence except by means of some likeness in itself. But Thomas adds, "unless perhaps we wish to say that Avicenna understands the knowledge of separate substances insofar as they are known by the habits of the speculative sciences and of the likenesses of things. " 27 On this view, any mental representation by which one knew something about God, even if derived from sense data, would be an impression of God, merely because it is less simple and less spiritual than God. Thomas probably has in mind here a position that he attributes to Aristotle: that through abstraction from sense data and syllogistic reasoning, one can arrive at a certain refined (but natural) knowledge of separate substances, the most sublime and felicitous knowledge possible in this life. 28 But that is not how Thomas interprets Avicenna in the body of the article. Thomas cites Avicenna's opinion also in his commentary on Boethius's De Trinitate. Reviewing the various ways in which one might know something, Thomas notes that one might see something through a form that does not come from the object by abstraction, but rather is impressed by the object on the knower, "as Avicenna says that we know the intelligences through their impressions in us." In that case, "the thing is simpler than the likeness through which it is known. " 29 But Thomas argues that no likeness of God impressed on the human intellect would suffice for quidditative knowledge of God, since God "infinitely surpasses every created form. "30 To what passage in Avicenna does Thomas refer? In both critical editions of the commentary on Boethius's De TrinitateBruno Decker's and the Leonine edition-the editors cite here a 27 Ibid., ad 9: "Ad nonum dicendum, quod dictum Avicennae quantum ad hoc non sustinemus, quia ei etiam ab aliis philosophis in hoc contradicitur; nisi forte velimus dicere, quod Avicenna intelligit de cognitione substantiarum separatarum, secundum quod cognoscuntur per habitus scientiarum speculativarum, et similitudinum rerum." 28 Cf. ScG III, c. 44. 29 De Trin., q. 1, a. 2: "siue sit impressa intelligenti ab eo, utpote quando res est simplicior quam similitudo per quam cognoscitur, sicut Auicenna dicit quod intelligentias cognoscimus per impressiones earum in nobis." Thomas wrote the commentary 1257-58, during his first Parisian regency. 30 Ibid. PHILIP L. REYNOLDS 515 passage from book 5 of Avicenna's De anima regarding the agent intellect. 31 But in the Scriptum, Thomas refers to Avicenna's Metaphysics. Moreover, he gives no indication that he links the idea of knowledge of separate substances through impression with Avicenna's theory about the agent intellect. 32 The reference is rather to a passage from book 3 of Avicenna's Metaphysics regarding the difference between cognition of material forms and cognition of separate substances. 33 The chapter from which this passage comes concerns the similitude between quiddities in reality and in the mind. Since there are both accidental and substantial forms in reality, must there be both accidental and substantial forms in the intellect to represent them? By insisting on the "intentional" character of mental forms, Avicenna can show that all mental forms, as such, are accidents, and that knowledge itself (scientia) is accidental. The agent intellect and other separate substances seem to present a special case. Since they are simple essences, not composites, how can one distinguish a quiddity from the thing itself? Must the mind become united with their very essence to know them? Avicenna argues that the mind understands them not directly, by their essence, but rather by "the intentions of their quiddities." The intellect must abstract material forms from matter to know them, but separate substances impress their own forms upon the intellect, as material forms would do too if they existed separately (which, needless to say, they do not). 34 What Thomas posits in both of these passages, I submit, is a purely spiritual cognition that would result in a mediated vision of God. But it is important to note that he posits it only 31 Expositio super librum Boethii De Trinitate, ed. Bruno Decker, editio altera (Leiden: Brill, 1959), 65 n. 1. Leonine edition, p. 84, note to line 61. Cf. Avicenna, De anima V, c. 5, in Avicenna Latinus, Liber de anima seu sextus de naturalibus N-V, ed. S. Van Riet (Louvain: Editions orientalistes; Leiden: Brill, 1968), pp. 126-27). 32 On that theory, see STh I, q. 84, a. 4. See also ScG II, c. 76. 33 As indicated in the apparatus fontium to the parallel text in the Ottawa edition of STh Suppl., q. 92, a. 1, obj. 9 and resp. 34 Avicenna, Metaplrysics III, c. 8. See Avicenna Latinus, Liber de philosophia prima sive scientia divina I-N, ed. S. Van Riet (Louvain: Peeters; Leiden: Brill, 1977), p. 161, lines 14-16. Ibid., p. 162, lines 26-32. 516 SPIRITUAL COGNITION IN THOMAS AQUINAS hypothetically. Moreover, in the passage from his commentary on Boethius, Thomas adds that such knowledge does not occur at all in via: Nor is God known by us in this life through purely intelligible forms that have some resemblance to him, because of the connaturality of our intellect toward phantasms, as has been said. Hence it remains that God is known only through the form of his effects.35 There is another passage in the Scriptum in which Thomas refers to knowing God through impression rather than abstraction, and in this case his use of the notion is affirmative. The passage occurs in book 1, distinction 3, where Thomas asks whether any created intellect can know God. 36 He explains in the body of the article that his question is not whether created intellects can have an unmediated knowledge of the divine essence (which he postpones until book 4), but whether God can be known in any way. Thomas's treatment of knowledge in this article is therefore broad and nonspecific. In the remainder of the response, he simply affirms that God is knowable, and goes on to argue that created intellects can never know God to the extent that God is knowable per se, and therefore can never comprehend the divine essence, since knowledge of something is always proportionate to the knower rather than to the known. 37 35 De Trin., q. 1, a. 2: "Nee etiam in statu huius uie cognoscitur Deus a nobis per formas pure intelligibiles, que sint aliqua similitudo ipsius, propter connaturalitatem intellectus nostri ad phantasmata, ut dictum est. Vnde relinquitur quod solummodo per effectus formam cognoscatur." 36 I Sent., d. 3, q. 1, a. 1. 37 Ibid. That statement leaves the beatific vision within the scope of the article, since no created intellect, according Thomas, even in the next life, can ever comprehend God: see IV Sent., d. 49, a. 2, q. 3; and STh I, q. 12, a. 7; I, q. 12, a. 1, ad 1. Here I disagree with John F. Wippel, in The Metaphysical Thought of Thomas Aquinas (Washington, D.C.: The Catholic University of America Press, 2000), 386: "Thomas's answer as he sets it forth here is that God can be known by us, but not in such a fashion that we can grasp or comprehend his essence. Here we see foreshadowed a position Thomas will often defend in subsequent discussions: we can know that God is, and what God is not, but not what God is" (emphasis added). But that is how Thomas characterizes our indirect, a posteriori, sense-based knowledge of God in this life, while even the vision of the divine essence in the next life, in his view, is not comprehensive but only apprehensive. PHILIP L. REYNOLDS 517 The five objections are designed to show that God cannot be known in any way. According to the third, no intellect can know God because things are known only through species, by which the intellect is assimilated to its objects. But intelligible species presuppose abstraction. Since God is entirely simple, no species can be abstracted from God; and therefore we cannot know God. 38 In reply, Thomas argues that we can know both God and the angels "not through abstraction, but through their impression on our understanding. "39 Since any species exists in the knower in a manner that befits the knower, Thomas explains, abstracted species are simpler than their objects, while the impressed species through which we know immaterial substances are less simple than what they represent. 40 This reply has puzzled scholars such as Ferdnand Van Steenberghen and John F. Wippel. 41 Its flavor seems uncharacteristic and too reminiscent of Platonic-Augustinian illumination. What does Thomas mean by "impression"? And is he referring here (as he does in reply to the fifth argument) to a supernatural mode of cognition? Augustinian ideas of illumination are surely in the background. As Wippel points out, Augustine says in the De libero arbitrio that "notions" of happiness and of wisdom have been "impressed" by God on our minds. 42 Moreover, Bonaventure refers to knowledge of God by an impressed likeness when commenting, like Thomas, 38 I Sent., d. 3, q. 1, a. 1, obj. 3. 39 Ibid., ad 3: "Unde non dicimur cognoscere ea per abstractionem, sed per impressionem ipsorum in intelligentias nostras." 40 Species are said to be abstracted from known objects, but perhaps one should interpret this "intentionally," such that what is known by means of the species (i.e., the quiddity itself) is abstracted from and simpler than the object itself. But even the reception of a species by a sense organ is a kind of abstraction, inasmuch as the matter of the form is left behind. In any case, the Scholastics sometimes conflate the quod of intellectual cognition (the quiddity) with the quo (the intelligible species). 41 Fernand Van Steenberghen, Le probleme de /'existence de Dieu dans /es ecrits des. Thomas d'Aquin, Philosophes Medievaux 23 (Louvain-la-Neuve: L'Institut superieur de Philosophie, 1980), 20. Wippel, Metaphysical Thought, 386-87 n. 20. 42 Augustine, De libero arbitrio 2.9 .26.103 (CCL 29:254 ). Wippel, Metaphysical Thought, 387 n. 20. See also Augustine, De Trinitate 8.3.4 (CCL 50:272), where Augustine speaks of an impressed notion of the good. 518 SPIRITUAL COGNITION IN THOMAS AQUINAS on book 1, distinction 3, of the Sentences, and here Bonaventure appeals to the authority of Augustine. Bonaventure concedes that God cannot be known through an abstracted likeness, since such likenesses are more spiritual than the objects from which they are abstracted. But he argues that "the intellect is informed by a certain knowledge [of God] that is a kind of likeness that is not abstracted but impressed, being inferior to God because it is in an inferior nature, yet superior to the soul insofar as it makes the soul better." 43 Here Bonaventure cites a text from De Trinitate in which Augustine argues that we know God, just as we know material objects, through some likeness in the soul (a position analogous to Avicenna's regarding knowledge of separate substances). 44 It is clear from another passage in Bonaventure's commentary that he has in mind an innate idea of God inscribed on the human intellect. 45 But Thomas says that humans know angels, as well as God, through impression, a reference that points rather to Avicenna's influence than to Augustine's. It is highly unlikely that Thomas is proposing a Platonic doctrine of illumination. He may have in mind a purely spiritual, supernatural knowledge of God that is mediated by infused species. But why go to such lengths to defend the nonspecific position that God is knowable in some way? Since the problem posed in the objection is that no species can be abstracted from God because God is entirely simple, it is likely that when Thomas refers to our knowing immaterial things not through abstraction but "through their impression on our understanding," he is referring to any knowledge of an object attained by means of a mental species that is less simple than the object. 46 In that case, even quite ordinary, natural, sense-based knowledge of God (if 43 Bonaventure, I Sent., d. 3, p. 1, a. un., q. 1, obj. 5. Ibid., ad 5: "nihilominus tamen, dum cognoscitur ab intellectu, intellectus informatur quadam notitia, quae est velut similitudo quaedam non abstracta, sed impressa, inferior Deo, quia in natura inferiori est, superior tamen anima, quia facit ipsam meliorem." 44 Augustine, De Trinitate 9.11.16 (CCL 50:307). 45 Bonaventure, II Sent., d. 39, a. 1, q. 2, dictum post resp. Bonaventure quotes Augustine here too, but the text is inauthentic. See also ltinerarium mentis in Deum, c. 3, n. 4, where Bonaventure states that counsel presupposes "an impressed notion of the supreme good." 46 I owe this suggestion to an anonymous reviewer. PHILIP L. REYNOLDS 519 any such there be) would be attained "through an impression" rather than through abstraction, and any species, phantasm, or concept by means of which one understood God or an angel would necessarily be an impression. This is indeed the alternative (and improbable) reading of Avicenna's theory that Thomas proposes in book 4 of the Scriptum. 47 If the last interpretation is correct, Thomas must have borrowed the vocabulary and conceptual apparatus of knowledge through "impression" from its Augustinian and Avicennan settings but adapted it to fit his own Aristotelian empiricism (and in so doing, deprived it of most of its explanatory force). The fifth argument is based on an analogical syllogism whose major premise is from Aristotle: the intellect is to phantasms as vision is to colors. But one cannot see anything without colors. Therefore the intellect cannot know anything without phantasms. But there can be no phantasms of God. For proof of this premise (the minor of a second syllogism), Thomas quotes Isaiah 40:18: "What image will you make of him?" 48 In reply, Thomas begins by saying that Aristotle was talking about such cognition as is connatural to us in this life. 49 It is true that one cannot know God in this way except through phantasms, and phantasms not of God himself but of the effects of God. But this does not preclude a higher, supernatural way of knowing God "through the influence of divine light," for which phantasms are not necessary. 50 Thomas may be referring here to some infused, supernatural knowledge of God that is possible in this life. Or he may be referring to the beatific vision, which is supernatural too but not possible in this life. IV. THE LIMITATIONS OF MEDIATED VISION Why should knowledge by impressed species be insufficient for quidditative knowledge of God? And if such cognition were IV Sent., d. 49, q. 2, a. 1, ad 9. I Sent., d. 3, q. 1, a. 1, obj. 5. 49 Ibid., ad 5. so Ibid. 47 48 520 SPIRITUAL COGNITION IN THOMAS AQUINAS possible, what would its content be? What would one see by mediated vision? The only medium in the beatific vision is the medium sub quo: God illumines the intellect, and thus disposes it for union. 51 No created form can adequately represent the divine essence, whether as a medium a quo or as a medium quo. Thomas uses three kinds of argument to show that one cannot know the divine essence in this life. First, there is a line of argument pertaining chiefly to specular, externally mediated cognition. Thomas maintains that while God is manifest in creatures as a cause is manifest in its effects, creatures do not reveal the essence of the Creator because they do not "equal the power of their cause" (non adaequanturvirtuti suae causae).52 This idea deserves detailed exposition, but the point is that a form reveals what it is by what it does: the external efficacy of an agent manifests the agent's intrinsic power (virtus), which in turn reflects its quiddity. 53 But God does not reveal the divine power in creation to this extent, for as creator, God is an equivocal cause. 54 Therefore causally inferred cognition shows only that God is, not what God is. Second, Thomas argues that no species that the mind has abstracted from material things is adequate for quidditative knowledge of any immaterial form, whether uncreated or created. For however mentally separated from its material conditions a material form may be, it is still a material form. 55 In the De Veritate, Thomas mentions that the essence of angels, unlike that of God, can be known "through certain intelligible species that differ from their essence," although not through species abstracted STh I, q. 12, a. 2, resp., ad finem; I, q. 12, a. 5. q. 1, a. 2. In De Trin., q. 6, a. 3, Thomas applies this argument to our knowledge of separate substances (angels), since their sensible effects too do not equal the power of the cause. 53 De Trin., q. 1, a. 2. 54 That is, a cause that produces a likeness inferior to itself, such that the resemblance involves no common nature. The standard example of an equivocal cause in Scholasticism is the sun as the source of sublunary generation, light, heat, etc. Thomas points out in STh I, q. 13, a. 5, ad 1 that God should really be called an analogical rather than an equivocal cause. 55 IV Sent., d. 49, q. 2, a. 1. ScG ill, c. 41. 51 52 De Trin., PHILIP L. REYNOLDS 521 from phantasms. 56 Clearly, Thomas has in mind here the Avicennan model of vision by impression, but he may be referring to how angels understand other angels, and not to how human beings might understand angels. Third, Thomas argues that no representative likeness in the intellect, whatever its source, can ever be sufficient (as a medium quo) for quidditative knowledge of God. This line of argument, which is the most comprehensive in scope and renders the other two strictly redundant, depends on the principle that the representation or species by which one knows an object must be a good likeness of it. Thomas assumes that the species and the external essence that it represents are comparable forms, and that one can know the latter only if the former exactly resembles it. 57 He notes that one could not see white by means of a sensible species of yellow. But he argues also that the two things, while formally the same, need not (and usually do not) have the same mode of being. 58 The form by which one sees the redness of a red apple has a quite different mode of being in the eye from that which the external sensible form has in the apple. Hence the alteration (immutatio) that a corporeal form engenders when it replicates itself in matter (as when fire makes something hot) is quite different from the alteration whereby the form communicates itself to the senses. Thomas characterizes alteration that produces sensation, whether such alteration occurs in the intervening medium or in the sense organ itself, as intentional or spiritual; and he characterizes the alteration by which material forms replicate themselves in matter as material or natural. 59 A sensible form may affect the sense organ materially as well, but that is accidental to sensory De Verit., q. 10, a. 11. This line of argument becomes obscure and perhaps circular if, with some modern Thomistic scholars, one construes the resemblance between interior species and external form in a purely intentional manner, reducing the resemblance to a correspondence of the sort that we expect in DNA or in computer information. (Nothing in the redhead's genes need actually be like redness.) 58 IV Sent., d. 49, q. 2, a. 1. 59 IV Sent., d. 44, q. 2, a. 1, qcla. 3; ibid., ad 2. STh I, q. 78, a. 3; 1-11, q. 22, a. 3; II De anima, c. 14 (Leonine edition, vol. 45.1). 56 57 522 SPIRITUAL COGNITION IN TIIOMAS AQUINAS cognition and may even obstruct it (as when a bright light dazzles the eye).60 Thomas uses several arguments to show that no mental likeness, from whatever source, is adequate for quidditative knowledge of God. The gist of all of them is that a mental likeness is a created form, and that no created form is sufficiently like God. In his commentary on book 4 of the Sentences, Thomas distinguishes between the modus essendi and ratio speciei of the representative species. A species in the knower need not (and usually does not) have the same mode of being as the external form has, but it must have the same ratio speciei. Since any created thing is like God neither in species nor in genus but only by analogy, 61 no species in a created intellect can serve as the means by which one knows the divine essence. 62 Thomas uses essentially the same argument in his gloss on 1 Corinthians 13: 12 and in the Compendium Theologiae.63 In his commentary on Boethius's De Trinitate, Thomas observes simply that "any likeness impressed on the human intellect would not suffice to make God's essence known, since that infinitely exceeds every created form. For this reason, God is not accessible to the intellect through any created form, as Augustine says."64 Thomas has in mind a familiar auctoritas, apparently not authentic, in which 1-11, q. 22, a. 2, ad 3. See also IV Sent., d. 14, q. 1, a. 1, qcla. 6, ad 2. On species, genus, and analogy as progressively remote modes of unity, see Aristotle, Metaph. 5.1016b31-1017a3. See also De partibus animalium 645b27-28 and 645b3-8. Aristotle's analogia is a relational resemblance, such that A is to B as C is to D, and not analogy in the peculiarly Scholastic sense (a modus loquendi between equivocity and univocity). 62 IV Sent., d. 49, q. 2, a. 1. Here Thomas outlines a scale of defectiveness: (1) things in the same species with different intensities, such as two white things with different degrees of whiteness; (2) things in the same genus but in different species; and (3) things in different genera that are analogically akin, such as a likeness of a man and a likeness of his whiteness insofar as both are beings. 63 In I Cor., c. 13, lect. 4. Comp. Theo/. I, c. 105. 64 De Trin., q. 1, a. 2: "Similitudo etiam quecumque impressa ab ipso in intellectum humanum non sufficeret ad hoc quod faceret eius essentiam cognosci, cum in infinitum excedat quamlibet formam creatam, ratione cuius intellectui per formas creatas peruius non potest esse Deus, ut Augustinus elicit." 60 STh 61 PHILIP L. REYNOLDS 523 Augustine says that God "escapes every form of our understanding. " 65 In the Summa Theologiae, Thomas presents three arguments to prove that one cannot know the divine essence through any mental likenesso66 The first is an argument from authority. Dionysius says that one cannot know superior things "through likenesses of the inferior order of things" (per similitudines inferioris ordinis rerum)o67 For example, Thomas explains, "the essence of an incorporeal thing cannot be known through the species of a body!' The auctoritas is ambiguous, for one can construe the genitive in two ways (as in "pictures of Picasso"). Notwithstanding Thomas's example (which is reminiscent of his critique of Avempace), 68 he takes the dictum to refer not only to likenesses resembling and representing inferior things, but to any likeness that in itself belongs to the inferior ordero Thus he proceeds with the following argument a fortiori: "How much less, therefore, can the essence of God be seen through any created species whatsoevero" The other two arguments likewise eliminate any mental likeness as an adequate meanso Since God's essence is the same as his being, no created form can match or represent his essenceo Furthermore, the divine essence is boundless, containing in itself "super-eminently" and at once every perfection that a created intellect is able to signify or to understand separately, such as wisdom, power, and being. No created species can represent such an essence because every created form is limited and determinateo Thomas uses similar arguments when commenting on John 1: 18 69 0 65 Cf. ibid., obj. 2: "Set sicut o{ figuratively to refer to members of the Christian community, the NRSV employs a variety of paraphrases: "brothers and sisters" (Gal 1: 11; 3: 15; 5:13; 6:18; PhH 1:14; 3:1, 17; 4:1), "friends" (Gal 4:12, 28, 31; 5:11; 6:1; Phil 4:21), "beloved" 1:12; 3:13; 4:8), or even "members of God's family" (Gal 1:2; d. Rom 8:29). 17 Some proponents of indusive language claim that, far from distorting 15 By contrast, even a very competent reader could be confused by the translation "children" in the NRSV and the NAB Revised NL Note that the translators of the latter revision find it necessary to include a marginal note (on 3:26) explaining that the Greek term literally means "sons" and that this is in contrast to the young child under a disciplinarian. It is also interesting to note mat both the King James Version (1611) and the Rheims NT (1582) use "children" to translate uioi (or Latin filii) in 3:26. But they are constrained by the force of Paul's argument to use "son(s)" throughout the rest of the passage. 16 W. Bauer, F. W. Danker, W. F. Arndt, and F. W. Gingrich, A Greek-English Lexicon of the New Testament and Other Early Christian Literature (3'd ed.; Chicago and London: University of Chicago Press, 2000), 18. 17 Such variety is itself curious. Did the translators fear that using "brothers and sisters" for every figurative use of ali£Aoi as "an especially delicate problem," chose to retain the traditional rendering "brothers," noting that "[t]here has never been any doubt that this designation includes all members of the Christian community, both male and female" ("Preface to the Revised Edition," in Saint Joseph Edition of the New American Bible [New York: Catholic Book Publishing, 1986], 7). INCLUSIVE LANGUAGE 587 the meaning of the original, such renderings are actually "more accurate" than the traditional "brothers" or "brethren. " 18 But whatever one thinks of this argument, it does not apply to the case of Greek uloc; and English "son." While lexemes in different languages rarely have identical semantic ranges, these two actually come pretty close. In the nonbiblical Koine Greek of Paul's day, uloc; had a range of senses similar to English "son. " 19 Thus, the singular uloc; referred to a male offspring, not to a daughter (0uych11p). And to refer to a mixed group of male and female offspring, one normally used a gender-neutral term such as TEKVa ("children") or rrmo(a ("[young] children"), not the plural ulo( ("sons"). 20 Or one could specify that both sons (ulol) and daughters (0uyaTfpEc;)were involved. 21 cf. Carson, The Inclusive-Language Debate, 130-31. It is amusing to note, however, that in a literal context where there is almost no doubt that both males and females are meant, and where it is a question of those who will betray Christ's disciples to death (Luke 21:16), the NRSV renders aoe;\(j>ol simply "brothers"! Apparently accuracy is less a concern in such cases. 19 The claims made in this paragraph are based on my examination of hundreds of nonliterary Koine papyri (mostly from the period 200 B.C. to A.D. 200) including searches run via the Perseus Digital Library (www.perseus.tufts.edu). For a convenient overview of the use of uiCx;, see James Hope Moulton and George Milligan, The Vocabulary of the Greek New Testament: Illustrated from the Papyri and Other Non-Literary Sources (Grand Rapids, Mich.: Eerdmans, 1930; repr. 1972), 649. 20 The one clear exception of which I am aware is in a "deed of disownment" from the Byzantine period, approximately 500 years after the time of Paul (P. Cairo Maspero 67353, as found in A. S. Hunt and C. C. Edgar, Select Papyri I, Loeb Classical Library [Cambridge, Mass.: Harvard University Press, 1932], 262-67). The author refers to his "parricidal children" (rnic; umpo;\q)oic; µou uioic;) and then gives their four names, two male and two female. 21 Thus, for example, in P. Oxyrhynchus 1464 (A.D. 25 0), we read: "I have sacrificed and made libation and tasted the offering along with Taos my wife and Arnmonius and Arnmonianus my sons [uioic;] and Theda my daughter [0uymp(]" (A. S. Hunt and C. C. Edgar, Select Papyri II, Loeb Classical Library [Cambridge, Mass.: Harvard University Press, 1934], 352-53). Other clear examples accessible through the Perseus Digital Library include (with my translation of the pertinent passages): P. Michigan 5.322a (A.D. 46), line 38 ("my aforementioned sons and my four daughters"); BGU 5.1210 (circa A.D. 150), line 67 (the property of the deceased is to be given "to their sons and daughters and [other] heirs"); and P. Oxyrhynchus 3.533 (second or third century A.D.), lines 26-27 ("greet Statia my daughter and Heraclides and Apion my sons"). Even in biblically influenced discourse and in a context where the terms are used to refer to spiritual offspring, uioi Kai 0uym(pec; ("sons and 18 Strauss, Distorting Scripture, 15; 588 GREGORY VALL Generally speaking, this usage is reflected even in the translation-Greek of the Septuagint (LXX), provided it is a question of the proper or literal sense of terms. 20 Thus when Hebrew banim means "children" of both sexes, 21 it is typically rendered TEKVa (e.g., Gen 3:16; 30:1; 31:16; 32:12); but when it refers specifically to males, ulo( is used (e.g., Gen 5 :4; 6: 10; 9:19; 10:25; 11:11; 19:12; 29:34; 34:25; 35:29; 37:35; 46:15). In the NT, similarly, the plural ulo( may refer to two or more male offspring (e.g., Matt 20:20-21; Luke 15:11; Acts 7:29; 19:14; Gal 4:22; Heb 11:21) but never to a mixed group of (literal) sons and daughters. 22 In the latter case one finds Tiat&ia (e.g., Matt 11:16; 14:21; Luke 11:7) or more commonly TEKva (e.g., Matt 7:11; 27:25; Mark 10:29; Luke 14:26 [cf. Matt 10:37]; Luke 23:28; Acts 21:5; 1 Cor 7:14; 2 Cor 12:14; Col 3:20-21; Titus 1:6). The point of all this is that the Greek lexeme ul6c; is every bit as much a "masculine-oriented" term as is English "son." According to its literal or proper sense it refers to males. This literal usage forms the foundation upon which any figurative use will be built. Or, to change the image, the proper sense is the sounding board off which the various figurative senses of ul6c; will daughters") is a natural enough expression (Epistle of Barnabas 1:1). 20 When uio( refers beyond the first generation of offspring and means "descendents," it does include females. But here it is a question of Semitic idioms such as "the sons of Israel" (= "the Israelites" [e.g., Judg 1: 1]). Other more figurative idioms involving uio( and inclusive of both genders are frequent in the LXX (e.g., "sons of unrighteousness" = "unrighteous persons" [2 Sam 3:34]) and in the NT (e.g., "sons of light" = "persons who live in the light" [1 Thess 5:5]). But such expressions do not touch upon our argument concerning the translation of Gal 3:23-4:7. 21 Because the Hebrew words for "son" (ben) and "daughter" (bat) are cognates (from Proto-Semitic binu* and bintu* respectively), the grammatically masculine plural form banim serves as the "unmarked" term with respect to sexual gender and can thus mean either "sons" or "children." When a Hebrew author wishes to specify that females are involved, the "marked" form banot ("daughters") is employed, as in the common phrase banim u-banot ("sons and daughters"). In this respect, Hebrew is like Latin (cf. filius, filia, filii, filiae) and unlike Greek or English or German (where the words for "son" and "daughter" are not cognates). 22 In John 4:12 the reference is probably to Jacob's twelve male children (thus even the NRSV renders uio( "sons" here); it is less likely that the author wishes us to think also of Dinah (cf. NAB: "children"). INCLUSIVE LANGUAGE 589 be heard. Thus the usage by which Paul refers to human beings of both sexes as "sons of God" (ulol 0rnu) represents an extension of the "masculine-oriented" term u\6c; to include females. Of course, Paul did not originate this usage, and so we must briefly consider its somewhat complex background. IV. THE INCLUSIVEUSE OF "SONS" In the Old Testament, Yahweh refers to Israel corporately as his "son" (Hebrew ben = Greek u16c; [Exod 4:22; Jer 31 :20; Hos 11:1]), 25 or in the plural as his "sons" (banim). In the latter case, because banim is unmarked for sexual gender (see note 23 above) and because the various contexts do not indicate otherwise, it is quite natural to assume that Israelite females are included. And in at least one passage (Isa 43:6), this is made explicit by Yahweh's reference to "my sons . . . and my daughters" (banay . . . ubenotay), which the LXX, naturally, renders Touc; ulouc; µou ... Kai Tac; OuyaT[pac; µou. But in the remaining instances of this usage, the LXX translators must choose between rendering banim as ulo( (Deut 14:1; Isa 1:2; Jer 3:14; Hos 2:1) or as TEKva (Deut 32:5; Isa 63:8). The former has the advantage of preserving the theologically important correlation to Israel's corporate identity as Yahweh's "son" (ben = uloc;), but the latter more readily suggests that females are included. Thus, while uloi 0rnu and TEKva 0rnu may both refer to the Israelites as chosen for a filial relationship with Yahweh, these phrases are not completely interchangeable. Nor is it simply the case that u1o{ represents a wooden rendering of banim and TEKVa a more idiomatic one. These assertions are confirmed by a brief examination of the Book of Wisdom, which was composed in idiomatic Greek just before the turn of the era (probably in Alexandria) and which seems to have played a formative role in 25 LXX Hos 11: 1 refers to the "children" (Tfrva) of Israel, not to God's "son." This may reflect a Hebrew Vorlage that differed slightly from the Masoretic Text. 590 GREGORY VALL Paul's thinking. 26 The Book of Wisdom contains about as many references to Israel's sonship as do all other OT books combined and represents in this regard (as in so many other ways) something of a synthesis of OT theology and a bridge to the NT. In light of these observations, it is striking to note how the author of Wisdom favors ulo( (Wis 9:7; 12:19, 21; 16:10, 26; 18:4} over TEKVa (only 16:21). That we should not ascribe this tendency to sexism is dear from the fact that in the first of these passages the author explicitly refers to Israel as God's "sons and daughters" (ulwv ... Kal 0uyaT€pwv [9:7]}, echoing Isaiah 43:6 and anticipating 2 Corinthians 6:18. Why, then, does the author of Wisdom favor ulo( over TEKva? Clearly he wants his readers to perceive the connection between his motif of seven references to "the sons/children of God" (plural) and the two references to "the son [ul6<;] of God" (singular) found near the beginning and end of the book (2:18; 18: 13 ). The first part of the book (chaps. 1-5) is dominated by the figure of the righteous sufferer who is both "servant of the Lord" (2:13) and "son of God" (2:18). Because he "boasts that God is his father" (2: 16), this man is mocked, tortured, and put to death by those who do not understand that man was made in God's "image" precisely in order to enjoy the incorruptible life of divine filiation (2:23). But God grants imperishable life to the righteous sufferer and raises him up to the glorious status of "the sons of God" or "holy ones" (5:5 [probably a reference to the angels]). Writing just a decade or two prior to the Incarnation, and drawing upon the riches of Torah, prophecy, and wisdom, this inspired author seems to have thus glimpsed the manner by which Israel would realize its true identity and vocation as "servant of the Lord" and "son of God": not through a return to the glorious wealth and political power of Solomon but in the person of a 26 Wisdom is probably the only OT book to be composed entirely in Greek (even 2 Maccabees contains a lengthy passage [1:1-2:18] that seems to be translated from a Semitic original), and it is almost certainly the last book of the OT to be written (circa 20 B.C.). Its influence on Paul's Epistle to the Romans is generally recognized, but one finds echoes also in Galatians. For example, Paul's use of the verb ("send forth") in Gal 4:4-6 draws upon Wis 9:10 ("Send her [Wisdom] forth from the holy heavens"). INCLUSIVELANGUAGE 591 poor, humble, and righteous sufferer imbued with divine wisdom. 27 In the central portion of the book (chaps. 6-9) the author adopts, with no little irony, the persona of Solomon (sans the greed, lust, oppressiveness, and idolatry of the historical figure) and humbly prays to receive the wisdom necessary to be a just king over God's "sons and daughters" (9:7). Finally, through his midrash (interpretive retelling) of Israel's early history (chaps. 10-19) the author constantly reminds his contemporary Jews (who, like the Hebrews of old, live in Egypt) that as God's "sons" they are called to be holy and righteous in the midst of a pagan world and that they will be sustained in persecution and nourished by revelation. The motif of sonship concludes when the Egyptians, chastised by plagues, are compelled to acknowledge Israel to be "God's son" (0rnu u16v; 18:13). This collective use of uloc; brings us full circle, recalling not only Wisdom 2: 18 but Exodus 4:22, the OT's first reference to God's fatherhood and Israel's sonship. The Book of Wisdom's theology of sonship thus sets the stage for the NT, where, of course, we find further developments. Under the title "Son of God" (6 uloc; Tou 0cou), Jesus is revealed to be the definitive embodiment and representative of Israel, the Messiah, the righteous sufferer, and the eternal Son of the Father. Israel's prerogative of sonship is fully realized in him and is the heritage of his followers, who are called "sons" (ulo() or "children" (TEKva) of God. The former usage is found in the Synoptic Gospels (Matt 5:9, 45; Luke 6:35; 20:36), the latter in the Johannine Literature Uohn 1:12; 11:52; 1 John 3:1-2, 10; 5 :2). In the Pauline epistles we find an even split, each term used (with this sense) six times. But the distribution is uneven, with Tfrva used more broadly (Rom 8:16-17, 21; 9:8; Eph 5:1; Phil 2:15). In fact, aside from OT quotations (Rom 9:26; 2 Cor 6:18), ulo( is confined to our passage in Galatians (3 :26; 4:6) and the parallel passage in Romans (8:14, 19). But it is striking that 27 "Now the further off a thing is, the less distinctly it is seen; and so those who were near Christ's advent had a more distinct knowledge of the good things to be hoped for" (STh II-II, q. 1, a. 7, ad. l; trans. Anton C. Pegis, Basic Writings of Saint Thomas Aquinas, vol. 2 [New York: Random House, 1945], 1066). 592 GREGORY VALL whereas in Romans 8 Paul alternates between ulo( and TEKva, in Galatians he uses only ulo(, apparently avoiding Tfrva. The reasons for this will be discussed below. We may now summarize this part of the argument. The Greek lexeme uloc;, according to its literal or proper sense, refers to males (both in the singular and the plural). Like its English counterpart "son," it is a "masculine-oriented" term. The biblical expression by which Israelites or Christians are referred to as God's ulo( thus represents an extension of the term to include females (precisely the sort of usage that the NRSV translators have set out to eliminate). The author of the Book of Wisdom strongly favors ulo( (which he clearly intends in a gender-inclusive sense [cf. 9:7]) over TEKva, apparently in order to underscore the correlations among: (1) Israel's corporate covenantal identity as uloc; 0rnu, (2) the plurality of righteous Israelites as ulol 0rnu, and (3) the individual righteous sufferer as uloc; 0rnu. Paul, for his part, refers to Christians both as ulol 0rnu and Tfrva 0rnu but in Galatians 3:23-4:7 has restricted himself to the former expression. In other words, like the author of Wisdom, Paul has chosen to use a "masculine-oriented" term "inclusively." As will become clear below, our capacity to understand his meaning and to receive revelation through his words depends in large part on our appreciation of this fact. V.THESTATUSOFSONS Next, we must attend to the distinction between inclusive language that is "anthropological" (referring to human beings) and that which is "theological" or "christological" (referring to God or Christ). Moderate inclusive-language translations such as language and have for the most part retained traditional masculine language when referring to God or Christ. 28 Thus, in our passage the NRSV renders uloc; with "child(ren)" when the reference is to mere human beings but retains the traditional "Son" when the 28 See note 10 above. INCLUSIVE LANGUAGE 593 reference is to Christ. By contrast, Burton H. Throckmorton's more radical translation employs inclusive language that is anthropological, theological, and christological, as exemplified by his rendering of Gal 4:6. "And because you are children, God has sent the Spirit of God's Child into our hearts, crying, 'God! Mother and Father!'" 29 While such a rendering leaps across the hermeneutical gap and transforms Paul's richly evocative biblical diction into the banal jargon of political correctness, 30 it does, almost in spite of itself, preserve one important dimension of Paul's argument that has nearly vanished in the NRSV. By maintaining a certain consistency in his rendering of ul6<; ("Child" when the reference is to Christ, "children" when it is to Christians), Throckmorton enables us to glimpse the correlation Paul wishes us to see between Christ's eternal Sonship and our filial adoption. The latter is a participation in the former. This point is, moreover, hardly peripheral to Paul's argument. His tight Trinitarian formula ("God has sent forth the Spirit of his Son") indicates that the Holy Spirit is precisely the Spirit of eternal filiation, whose action in our hearts and minds is required if we are to enter into Jesus' relationship to the Father. This is further underscored by Paul's coupling of the Aramaic abba of Jesus' prayer with Greek 6 naTtjp (cf. Mark 14:36). The reader of the NRSV could easily overlook all of this, since the first due is obscured. Apparently Jesus is "Son," but we are mere "children. " 31 In a passage in which Paul draws a sharp contrast between the "child" (vtjmo<;)who is still in his minority and the 29 Burton H. Throckmorton, Jr., The Gospels and the Letters of Paul: An InclusiveLanguage Edition (Cleveland: Pilgrim Press, 1992), 212. Throckmorton even employs what might be termed "ancestral inclusive language." For example, he has Paul say that Christians are "offspring of Abraham and Sarah, heirs according to promise" (Gal 3:29; emphasis added). 30 To Throckmorton's credit, he has at least preserved a modicum of intimacy in his rendering of af3fla 6 ucrrtjp. The same cannot be said for the translators of the NTILB, who render the phrase, "O, My dear Parent!" (The New Testament of the Inclusive Language Bible [Notre Dame, Ind.: Cross Cultural Publications, 1994], 215). 31 According to Strauss, "To argue that this connection is 'obscured' in the NRSV is probably nitpicking" (Distorting Scripture, 162). Given the importance of the truths of revelation that are at stake, I am willing to run that risk. 594 GREGORY VALL "son" (uloi:;) who has come of age, the text as translated in the NRSV seems to qualify our relationship with God a way that runs counter to Paul's intention. There is, of course, a huge difference between Christ's Sonship and ours, as Paul indicates by using the word ulo0i::cr(a ("adoptive sonship") in reference to the latter. One might even be tempted to defend the NRSV as safeguarding this distinction, especially in light of the fact that elsewhere in the NT the Johannine Literature consistently refers to Christ as uloc;; but studiously avoids this term when speaking of Christians, using TEKVa ("children") instead. But that is exactly the point. The various formulations of Paul, John, and the other sacred authors play complementary roles within t:he NT's overall witness t:o the mystery of Christ. A translation that fails to aHow each of these authors to speak in his own voice risks upsetting this delicate canonical balance and to that degree does not serve revelationo Next we should note that the NRSV committee's decision to use the gender-inclusive "child(ren)" to translate the anthropological occurrences of uhk; in Galatians 3:23-4:7 has caused a RSV (and most other EngHsh transripple effect. For in the lations) "chHd(ren)" already been employed in this passage to render the word vtjmoc; (4: 1, 3)0 Since Paul's argument, as just noted, hinges on a sharp distinction between vtjmoc; ("babe, child" [here: the heir during his minority]) and uloc; ("son" [here: the heir having come of age there was no question of using the same English word to translate both Greek terms. Such a procedure would have made nonsense of the passage. 32 Instead, the NRSV translators use the word "minor(s)" to render vtjmoc;. In and of itself, this is an excellent choice and arguably an improvement over the RSV, since "minor" pinpoints the precise sense of vtjmoi:; that Paul intends in this context But only serves to highlight a further problem with using "child(ren)'' for uloc;. For English "child" by no means connotes n, 32 This does not stop Throdanorton or the NTILB translators, however. In blatant disregard for Paul's train of thought, they employ "child(ren)" to render both and The resulting translation is not only inaccurate but unintelligible. INCLUSIVELANGUAGE 595 that one so called has come of age, is no longer in need of guardians, and is ready to take possession of the inheritance. If anything, the word is suggestive of just the opposite. Although the NRSV translators attempt to establish a semantic opposition between "minor" and "child," the two words sound more like synonyms. Thus, despite considerable ingenuity, the NRSV translators are not able both to eliminate "masculine-oriented language" and to give the reader the best possible opportunity to follow Paul's dense and subtle line of argumentation. As Bruce Metzger confesses in his preface to the NRSV, "more than once the Committee found that [its] several mandates stood in tension and even in conflict." 33 In the case of Galatians 3:23-4:7 the tension between inclusive language and accuracy of translation has been resolved in favor of the former, Making explicit reference to Galatians 4:7, Carson attempts to forestall this sort of criticism and to defend the procedure of rendering ulOi:; with "child(ren)." He argues that since Greek Tfavov ("child") refers to an heir elsewhere in the NT (Mark 12:19; Luke 1:7; Acts 7:5), there is no reason why English "child(ren)" cannot refer to the heir(s) in the Galatians passage, even if it happens to be translating uloi:; (rather than Tfavov). His examination of Romans 8:14-21 seems to clinch the mattero In this passage, which contains a very dose parallel to Galatians 4:47, Paul freely alternates between uloc; and Tf'Kvov, so that "it is difficult in the flow of this context to detect [any] significant semantic distinction between the two terms. " 34 Carson thus implies that while Paul in fact uses u'l6c; in Galatians, he could just as easily have used TflOVTWVKaTa TOU d vm OKOAOU0T]atv TO a(nov cmwaouv 0mEpuan t.f:yolT' av" (Aristotle, Categories and Propositions, trans. Hippocrates G. Apostle [Grinnell, Iowa: The Peripatetic Press, 1980]). 28 See John J. Cleary, Aristotle on the Many Senses of Priority (South Bend, Ind.: St. Augustine's Press, 2002), 29. 618 AUGUSTINE, ARISTOTLE, AND THE CONFESSIONS In book 12 of the Confessions Augustine also lists four kinds of priority: eternity (aeternitas), time (tempus), choice (electio), and origin (origo) (12.29.40). He describes them in the following way: [Something is said to be prior by] eternity, as God [is prior all things; by time, choice, as the fruit [is prior to] the flower; as the flower [is prior to] the fruit; by origin, as the sound [is prior to] the song. 29 these chapters Augustine has been canvassing a number of different orthodox interpretations of Genesis 1:1 and 1:2, an endeavor which yidds no definitive condusions about the meaning of the text but which at least demonstrates the legitimacy of interpretative pluraHty. 30 The priorities are introduced in an attempt to clarify one such valid interpretation, the opinion that the verse "In the beginning God made" means "First God made" (12.28.38). In order to avoid a possible contradiction with the rest of the Genesis passage, the subscriber to this view would need to understand "first" in a non-temporal sense-hence the need to explore different notions of priority. Despite differences in nomenclature, a comparison of Aristotle's and Augustine's lists reveals striking similarities. Augustine, for example, keeps Aristotle's initial topography of four priorities 31 but readily accepts Aristode's unstated demotion of the fourth, the so-called priority by nature, to a priority of human whim. Thus in his own list Augustine will refer to the "priority of choice" (electio) and give as an example the fact that most men and women will choose the delicious and nutritious fruit over the pretty but useless flower. There is also a strong affinity, if not identity, between Augustine's "priority of origin" and Aristotle's "priority of order." Not only is Augustine's example of sound and song comparable to Aristotle's example of 29 "aeternitate, sicut deus omnia; tempore, sicut flos fructum; electione, sicut fructus florem; origine, sicut sonus cantum" (12.29.40). 30 Augustine here does not seem to have any particular set of "opponents" or even interlocutors in mind, as none of the various positions he mentions were commonly associated with any recognizable Christian or even Manichean authorities (cf. O'Donnell, Augustine Confessions, 3:316-17, 328, 329). 31 Augustine most likely does not mention the fifth priority because its subject matter is not relevant to his present concerns. MICHAEL P. FOLEY 619 letter and syllable, but his renaming of it in terms of ongm conforms to Aristotle's treatment of this priority as a logical ordering of first principles and propositions, the former of which constitutes the starting point for the latter. 32 Even the priorities most distinct from each other, Aristotle's "priority of existence" and Augustine's "priority of eternity," betray an intriguing compatibility. For Aristotle, the priority of existence means that A's existence follows from B's, but B's does not follow from A's. Were one to substitute "God" and "all things" (the two referents Augustine uses to exemplify the priority of eternity) for A and B, one would arrive at the following statement: God is prior to all things as God's existence follows from the existence of all things, but the existence of all things does not follow from God's existence. Such a compatibility by no means suggests that the two priorities are the same; on the contrary, as A Solignac, S.J., notes in his explication of 12.29.40, the priority of eternity is the only one of Augustine's priorities that requires a uniquely Christian view of God "immutably creating mutable beings. " 33 But if this is the case, then Augustine's priority of eternity would mark a development of Aristotle's priority of existence in light of divine revelation. Thanks to the proficiency wrought by his ongoing conversions, Augustine has indeed learned to make good use of the Categories. IV. POSSIBLE DIFFICULTIES Before proceeding any further, however, I should point out that there are two possible difficulties with my contention that Confessions 12.29AO includes a silent but salient use of chapter 12 of Aristotle's Categories. First, given Augustine's liberal use of Neoplatonic sources, it may be more likely that he took the 32 Augustine's development of this priority in terms of matter and form ("sonus eius materies eius est ... formatur ... ut cantus sit [12.29.40]) most likely accounts for his changing the name. 33 "La priorite selon l'eternite comporte en effet que !'on saississe le paradoxe d'un Dieu creant immuablement des etres muables; c'est une priorite de transcendance" (CEuvresde Saint Augustin, vol. 14 [Paris: Desclee de Brouwer, 1962], 612 n. 26). 620 AUGUSTINE, ARISTOTLE, AND THE CONFESSIONS foundation for his theories on priority from some work other than the Categories. This would explain, for instance, his alteration of the Aristotelian order of the categories listed in book 4 from "substance, quantity, and quality" to "substance, quality, and quantity," the latter order being the one transmitted by one or more veins of the Neoplatonic tradition. 34 It is certainly true that Augustine could have culled the different meanings of "first" from such sources, although concrete evidence is wanting. Porphyry's Commentary on the Categories and Isagoge, both of which are dedicated to an analysis of Aristotle's Categories, make no mention of the priorities, and neither does Plotinus' s Enneads; for that matter, none of the great investigators of Augustine's sources-namely, Pierre Courcelle, Harold Hagendahl, John J. O'Meara, Robert O'Connell, and James D. O'Donnell-have ever drawn a connection between 12.29.40 and a Platonic text. Yet even if a Neoplatonic derivation is possible, it is still not unreasonable to turn first to the Categories as a likely source for Augustine's views on priority, as this is the only relevant book that we know Augustine read. (Which Neoplatonic works he read, on the other hand, remains a hotly contested point. 35 ) Nor would our hypothesis preclude a Neoplatonic influence. For if Augustine's list of the categories in 4.16.28 betrays a Neoplatonic hermeneutic (which to some scholars, incidentally, is a big "if"), then this merely establishes that Augustine's memory or appropriation of Aristotle was affected to some degree by his interaction with Neoplatonism. The same dynamic could be operative in his knowledge of the priorities without in any way denying that Aristotle is his primary, albeit quasi-mediated, interlocutor. The second difficulty concerns the consensus among modern scholars that the last five chapters of the Categories (which include Cf. O'Donnell, Augustine Confessions, 2:265. Cf. Pier Franco Beatrice, "Quosdam Platonicorum Libras: The Platonic Readings of Augustine in Milan," Vigiliae Christianae 43 (1989): 248-81; Frederick Van Fleteren, "Plato, Platonism," in Augustine through the Ages, 651-54; R. H. Nash, "Some Philosophic Sources of Augustine's Illumination Theory," Augustinian Studies 2 (1971): 47-66; O'Donnell, Augustine Confessions, 2:416-17. 34 35 MICHAEL P. FOLEY 621 the discussion on the priorities) are spurious. 36 This may indeed be the case, but what is essential for our purposes is not whether Aristotle wrote 11b15-15b32 of the Categories (dubbed the postpraedicamenta), but whether Augustine thought that Aristotle wrote them. Based on the evidence we have of late-fourth-century views on the Aristotelian corpus, it seems relatively safe to conclude that Augustine did indeed consider the postpraedicamenta genuine. It is generally acknowledged that Andronicus, a Peripatetic editor from the first century B.C., knew of the last five chapters 37 and that subsequent commentators on the Categories-for example, Dexippus (early 4th century A.D.), Arnmonius (5th century), Philoponus (mid-5th century), and Simplicius (early 6th century)-either wrote on these sections or alluded to them. 38 Porphyry also knew of the postpraedicamenta and considered them authentic. 39 This being the case, when the Ten Categories came into Augustine's hands, the work almost certainly included 11b15-15b3 2, all of it dutifully ascribed to Aristotle. V. CONCLUSION Augustine's use of Aristotle in the Confessions is not only consistent with his qualified appreciation of the philosopher in the rest of his writings, but it also suggests an engagement with Aristotelian philosophy that extends more deeply than has generally been recognized. Through a dexterous return to the Categories, Augustine was able to overcome the pitfalls of his adolescent arrogance by taking Aristotle's teachings on time and priority and weaving them into his own mature reflections on these topics in books 11 and 12 of the Confessions. This, of course, does not render Augustine an "Aristotelian" any more than Thomas Aquinas's frequent appeal to the writings of pseudo36 Cf. L. Minio-Paluello, Aristotelis Categoriae et Liber de Interpretatione (Oxford, 1949) v, vi. Cf. Simplicius, Commentarius in Aristotelis Categoriae, 379.8. 93.9-12; 167.21-168.3; 379.8ff, respectively, of each author's Commentary on the Categories. 39 Cf. Simplicius, Comm. in Aris. Categ., 379.13. 37 38 Cf. 17.7-9; 622 AUGUSTINE, ARISTOTLE, AND THE CONFESSIONS Dionysius him a "Neoplatonist." But it does attest to a certain resourcefulness and independence of mind on Augustine's part, to say nothing of a bold willingness to fuse together diverse schools of thought. For that matter, Augustine's "synthesis" of different philosophical traditions may even suggest a greater compatibility between Aristotle and Plato than our own stereotypes currently allow. than disregarding Augustine's conclusions about the recondite harmony of classical philosophy's twin giants as the aping of a Ciceronian or Plotinian tenet (both of these authors being themselves often unjustly dismissed as "edectic"), 40 Augustine's position on the Academy and the Lyceum may be one that merits our serious consideration" 40 The very term "eclectic," in fact, does not fully deserve its disparaging connotations. Cf. Pierlugi Donini, "The History of the Concept of Eclecticism," in The Question of "Eclecticism": Studies in Later Greek Philosophy, ed. John M. Dillon and A. A. Long (Berkeley: University of California Press, 1988), 15-33. The Thomist 67 (2003): 623-43 A LONG DISCUSSION REGARDING STEVEN A. LONG'S INTERPRETATION OF THE MORAL SPECIES STEVEN JENSEN University of Mary Bismarck, North Dakota L IKE STEVEN LONG 1 I am disturbed by the view maintaining that crushing a baby's skull is not necessarily an act of killing the baby; yet I am unconvinced by his counter analysis of Aquinas. The difficulty concerns the nature-or the moral species-of a doctor's act when she performs a craniotomy in order to save the life of the mother. Is the doctor's action a resizing of the head or a killing of the baby? Aquinas, of course, does not discuss craniotomies, but Long thinks he may find an answer to the difficulty in Aquinas's discussion of self-defense. After all, the problem is more general than craniotomies, encompassing the nature of all moral actions. What is, for example, the moral nature of the act of killing in self-defense? Is it a killing in order to save one's life or is it a repelling of the attack, with the side-effect of the assailant's death? In what follows, I do not dispute many of Long's criticisms of the view he opposes, but I do dispute his own interpretation of Aquinas. He has attempted to construct a theory from a few difficult passages, while he has failed to address the many passages that might call his view into question. I propose, therefore, to do three things. First, I will lay out Long's view; second, I will show that his view cannot be reconciled with Aquinas's; and finally, I 1 See Steven A. Long, "A Brief Disquisition regarding the Nature of the Object of the Moral Act according to St. Thomas Aquinas," The Thomist 67 (2003): 45-71. 623 624 STEVEN JENSEN will tentatively suggest the direction to go for a better interpretation. 2 I. LONG'S VIEW The dispute focuses on the text of STh 11-11, q. 64, a. 7, in which Aquinas discusses self-defense. Aquinas claims that the defender can intend to preserve her own life, but she cannot intend to kill the assailant. The relevant passages are worth quoting: Nothing hinders one action from having two effects, one of which is within intention and the other of which is outside intention. But moral actions receive their species according to that which is intended, and not from that which is outside intention, since this is per accidens,as is plain from what has been said. From an act of self-defense, then, two effects can follow: the preservation of one's own life and the killing of the assailant. Acts of this sort, if what is intended is the preservation of one's own life, do not have the formality of being unlawful, since it is natural to anything to preserve its own existence insofar as it can .... It is unlawful for a man to intend to kill a man in order to defend himself, except for those who have public authority, who while intending to kill a man for self-defense, refer this act to the public good, as is plain for soldiers fighting the enemy, or for a minister of the judge fighting against robbers, although even these sin if they are moved by private desires. 3 2 Many, of course, have written extensively on the nature of moral actions in Aquinas, including Servais Pinckaers, "La role de la fin dans !'action morale selon Saint Thomas," in Le renouveau de la morale (Paris-Tournai, 1964); Stephen L. Brock, Action and Conduct: Thomas Aquinas and the Theory of Action (Edinburgh: T & T Clark, 1998); Eric D'Arcy, Human Acts: An Essay in their Moral Evaluation (Oxford: Clarenden Press. 1963); and (at least in a Thomistic vein) G. E. M. Anscombe, Intention (ld ed.; Ithaca, N.Y.: Cornell University Press. 1963). 3 STh II-II, q. 64, a. 7: "Nihil prohibet unius actus esse duos effectus, quorum alter solum sit in intentione, alius vero sit praeter intentionem. Morales autem actus recipiunt speciem secundum id quod intenditur, non autem ab eo quod est praeter intentionem, cum sit per accidens, ut ex supradictis patet. Ex actu igitur alicuius seipsum defendentis duplex effectus sequi potest, unus quidem conservatio propriae vitae; alius autem occisio invadentis. Actus igitur huiusmodi ex hoc quod intenditur conservatio propriae vitae, non habet rationem illiciti, cum hoc sit cuilibet naturale quod se conservet in esse quantum potest. ... Illicitum est quod homo intendat occidere hominem ut seipsum defendat, nisi ei qui habet publicam auctoritatem, qui, intendens hominem occidere ad sui defensionem, refert hoc ad publicum bonum, ut patet in milite pugnante contra hostes, et in ministro iudicis pugnante contra LONG ON THE MORAL SPECIES 625 The uncontroversial first claim, that it is legitimate to intend to preserve one's own life, poses no difficulty, but the second claim, that the defender cannot intend to kill the assailant, has generated much controversy, largely over the meaning of the word 'intend'. Some say that intention includes the means aimed at, and therefore in self-defense one can in no way legitimately aim to kill, either as an end or as a means, but Long claims that in this text 'intention' refers only to intention of the end, which is Aquinas's primary meaning of intention. On Long's reading, then, Aquinas is saying that one cannot intend to kill as an end, but it might be permissible to kill as a means to the goal of saving one's life. In favor of his interpretation Long cites many passages in which Aquinas says that intention concerns the end while choice is of the means. "Just as intention concerns the end, so choice concerns those that are ordered to the end. " 4 The act of self-defense, then, may have two effects, the preservation of one's life and the death of the assailant. While one can intend (as an end) to preserve one's life, it is impermissible to intend to kill (as an end) in order to save oneself. It is permissible, according to Long's interpretation, to choose to kill as a means. The proscription applies only to intention, not to choice, and therefore only to the end and not to the means. Long's idea seems to be something like the following: killing is legitimate if one chooses it as a necessary but unwelcome means to achieve the good goal of preserving one's life, but one cannot begin to want the assailant's death as something desirable in itself, apart from its utility in saving. As long as one desires the killing simply as a means, it does not give moral species to the action, for it is not intended, and so the action falls under the species of selfpreservation. But when one begins to desire the killing as an end, as something desirable apart from its utility, then it is intended and the action falls under the unlawful species of killing. latrones. Quamvis et isti etiarn peccent si privata libidine moveantur." 4 STh I-II, q. 13, a. 4: "Sicut intentio est finis, ita electio est eorum quae sunt ad finem."' Also see STh I-II, q. 12, a. 4, ad 3. In STh I-II, q. 8, intro., Aquinas lists intention as an action that concerns the end, and in STh I-II, q. 13, intro.; ITh I-II, q. 13, a. 3; and ScG III, c. 6, he states that intention does not concern the means. 626 STEVEN JENSEN This interpretation seems to imply the unacceptable conclusion that the means never give moral species to an action, and that many wrong actions can be justified on the basis of good intentions. Robin Hood, for example, does not intend to steal, for he desires it only as a means to give to the poor; his action, as almsgiving. Similar therefore, is not specified as theft, arguments can be made for just about any action, from adultery to murder. Long's position, it seems, reduces to a morality of intention, an Abdardianism in which only a person's goals determine the good or evil of her actions, the actions themselves never playing a role in morality. Long is ready with a reply to this difficulty. He says that the species to human actions, for means chosen do in fact give its own the action chosen as a means has its own object natural order to some end, from which it receives its species. There are two sources of the species of human actions, namely, the object of the action and the end intended by the agent. The action of theft has its own object and gives species to the action, no matter the intentions of Robin Hoodo At this point, it seems that Long is getting himself into deeper water, for he is trying to interpret a text that states that what is outside intention does not give species. Long claims that the means is outside intention, but then he goes on to daim that the means does in fact give species. In other words, he says that what is outside intention gives species, when the very text he is trying to understand explicitly states the opposite. Long's attempt to extricate himself from this perplexity is rather convoluted. Nevertheless, he reaches a coherent position, however much he must stretch the text of STh II-II, q. 64, a. 7 to do so. Without ever referring us to any text or passage of Aquinas, Long relies upon a teaching that undoubtedly arises from STh I-H, q. 18, aa. 6 and 7. As Long puts it, "When the object [of an action] is naturally ordered to the end, then the moral species derived from the end is the defining species. " 5 Or as Aquinas says, 5 Long, "Object of the Moral Act," 58. LONG ON THE MORAL SPECIES 627 When the object is not per se ordered to the end, then the specific difference from the object is not per se determinative of the species from the end, and vice versa. It follows that one of these species is not under the other, but the moral act is under two disparate species, for example, we say that he who steals in order to commit adultery, commits two evils in one act. On the other hand, if the object is per se ordered to the end, one of the differences is per se determinative of the other, and one of the species will be contained under the other. 6 Aquinas goes on to say that the species from the object is contained under the species from the end. We have, then, Long's two sources of moral species, the object and the end; we have the idea of the object being naturally (or per se) ordered to the end; and we have the idea of the species from the end being the defining species. Unfortunately, it is still far from evident how this teaching is supposed to get Long out of his dilemma. Add one further teaching-that killing is sometimes naturally ordered to self-preservation-and we are still left in the fog. Long derives this last teaching, plausibly enough, from the fourth reply of STh 11-11, q. 64, a. 7. The objection says that no one is allowed to commit adultery or to fornicate in order that he may save his own life. Since killing is worse than either of these sins, then neither can someone kill to save his own life. Aquinas responds that, "The act of fornication or of adultery is not ordered of necessity to preserving one's own life, as is the act from which killing sometimes follows. "7 We have yet to see how these ideas are supposed to reconcile the claims that what is outside intention does not give species, yet the means, which is outside intention, does give species. Let us "Sic igitur quando obiectum non est per se ordinatum ad finem, differentia specifica quae est ex obiecto, non est per se determinativa eius quae est ex fine, nee e converso. Uncle una istarum specierum non est sub alia, sed tune actus moralis est sub duabus speciebus quasi disparatis. Unde dicimus quod ille qui furatur ut moechetur, committit duas malitias in uno actu. Si vero obiectum per se ordinetur ad finem, una dictarum differentiarum est per se determinativa alterius. Uncle una istarum specierum continebitur sub altera." Aquinas does not use Long's wording of "naturally ordered to an end"; rather, he speaks of something being "per se ordered to an end." 7 STh II-II, q. 64, a. 7, ad 4: "Dicendum quod actus fornicationis vel adulterii non ordinatur ad conservationem propriae vitae ex necessitate, sicut actus ex quo quandoque sequitur homicidium." 6 STh I-II, q. 18, a. 7: 628 STEVEN JENSEN begin by considering the ideal case in which the act of killing is naturally ordered to the end of self-preservation, and in which the defender chooses to only as a means, without intending to kiH as an end (which is precisely the case that Long thinks is justified by STh II-H, qo 64, ao 7)o The means gives us the species of killing, the end intended gives us the species of self-preservation, but since the object of killing is naturally ordered to the end, it follows that its species is not entirely disparate; rather, it falls under the species of self-preservationo The species from the object serves merely to narrow the defining species of self-defense, which is derived from the end, into something like 'lethal selfdefense' o8 What matters is that it remains, in species, an act of self-defense, which Aquinas explicitly states is legitimate. Now suppose that the defender wants to kill not simply as a means but as an end. No longer can his act of killing fall under the species of self-defenseo Rather, killing is intended and so becomes an independent species of its own, which Aquinas states is unlawfuL Finally, consider the act of fornication as used to preserve one's own life (eog., one is threatened with death if one does not fornicate)o Suppose that the person chooses to fornicate only as a means, and no way intends it as an end. Can we then say that fornication is legitimate because it falls under the broader species of self-defense, so that it becomes 'fornicating self-defense'? No, we cannot, for fornication is not naturally ordered to preserving one's life, so it, like the instance of stealing in order to commit adultery, must give rise to an entirely separate specieso Similarly, if some act of killing were not properly proportioned to selfpreservation, it would not simply narrow the defining species of self-defense, but would generate its own species, which again would be unlawful. We have, then, three cases. The act of kiHing is naturally ordered to self-defense and it is not intended as an end; then we have a single species of lethal self-defenseo (2) The act of killing is naturaHy ordered to self-defense, it is intended as a good in 8 Long, "Object of the Moral Act," 64. LONG ON THE MORAL SPECIES 629 itself; then we have killing as an independent species of its own, not falling under self-defense. (3) The act of killing (or fornicating) is not naturally ordered to self-defense, so that even if it were chosen merely as a means it would nevertheless give rise to an independent species of its own. Because self-defense is an acceptable act, the first case, which falls under the species of selfdefense, is morally good. Since a private individual cannot kill (as an action with its own independent species), it follows that the second two cases are evil. Long, by the way, fits the craniotomy case, in which the doctor crushes the head of the baby (and so kills the baby) in the third case; killing the baby is not naturally ordered to saving the mother, he says, for the baby is in no way engaged in an action of endangering the mother. 9 The position so far laid out is internally consistent, even if it leaves gaping holes, such as the problem of determining when an action is naturally ordered to an end, but it still has not got Long out of the woods. It is a fine account of self-defense, but it is no account of Aquinas's statement that what is outside intention, being per accidens, does not give species. Why would Aquinas make such a statement so dearly contrary to his teaching that the means do give species? First of all, by praeter intentionem Aquinas means what is outside the general intention. 10 Suppose I intend to take a trip to Chicago. Such a general intention is as yet indeterminate as to whether I will take a plane, a train, or an automobile. In other words, the means to get to Chicago is outside the scope of the general intention to go to Chicago. Similarly, the means of defense is outside the general intention to defend myself. Now since, in the ideal case (1) above, the species of the action ultimately falls under the end of self-defense, it is, says Long, quite appropriate to say that the object (or means) does not give species; 11 it merely determines the species, rather than gives an independent species of its own. For the first case, then, Aquinas's statement is true: what is praeter intentionem (the means) does Ibid., 68. Ibid., 60. 11 Ibid. 9 '0 630 STEVEN JENSEN not give species. Aquinas's statement does not apply to the second case, of course, since killing is intended as an end. But the third case, in which the means is outside intention yet gives its own independent species, remains a difficulty for Long. The third case simply does not fit under Aquinas's general principle that what is outside intention does not give species. Why, on Long's reading, does Aquinas ignore this third case when he gives the general statement about species? Quite naturally, thinks Long, because by knowing that the action is lethal, we do not yet know whether it fits the third case; we must further determine whether the action is naturally ordered to the end, which is why Aquinas goes on to say that even with a good intention, the act of self-defense can be unlawful, if it is not proportioned to the end. 12 All of this may seem quite natural to Steven Long, but I will not, even at this point, hide my skepticism. Aquinas seems to be giving a general principle, that what falls outside intention does not give species, but in his actual theory, the principle turns out not to be general at all. It is, in fact, only a principle for the ideal case, the situation in which the means chosen is naturally ordered to the end intended. Even then the principle holds only in an attenuated way, since the chosen means, which is outside intention, further determines the species (e.g., making it defense by killing or lethal defense). Rather than tell us that his principle is not general, Aquinas gives it a quasi-general application because, says Long, we don't know just from the intention whether we have a case in which it applies. All the more reason, it seems to me, for Aquinas to clarify the scope of this principle when he lays it out at the beginning of STh 11-11, q. 64, a. 7. With that note, let us turn to our second task, criticizing Long's account. II. PROBLEMS WITH LONG'S ACCOUNT I will cover the following four difficulties with Long's account: (1) According to Aquinas, intention does indeed concern the 12 Ibid. LONG ON THE MORAL SPECIES 631 means as well as the end; (2) Long ignores the fact that the proximate end is most important in determining the species; (3) Long's use of praeter intentionem does not correspond with Aquinas's; and (4) even the public official cannot do what Long says he can, namely, intend to kill as an end. As I criticize Long's view, I will invariably make points that tell in favor of the view opposed by Long, what he calls the Cajetanian interpretation, but my arguments should not be taken as a defense of the Cajetanian interpretation, a view with which I have my differences, at least as it is usually applied to selfdefense in current discussions. 13 A) Intention of the Means According to Cajetan's reading of STh II-II, q. 64, a. 7, Aquinas prohibits not only intending killing as an end but also intending killing as a means; the death must be a result, a side-effect, of the action that is intended, which is simply an action of repelling the attack. Long, on the other hand, says that the means chosen does not fall under intention, so that one may legitimately choose to kill in order to save one's life. As we have seen, on this point Long actually provides textual evidence corroborating his claim, leaving no doubt that for Aquinas intention concerns the end while choice concerns the means. Unfortunately, Long ignores crucial passages that throw his interpretation into doubt. 14 In particular he ignores STh I-II, q. 12, aa. 2 and 3, tvvo articles that directly address the question of whether intention could ever concern the means, or what is ordered to the end (ad finem). The first of these articles asks whether intention concerns only the ultimate end. In response, Aquinas says, 13 I do not doubt Cajetan's interpretation of STh II-II, q. 64, a. 7 so much as I doubt its application to most instances of self-defense. 14 For a fuller account of intention in Aquinas see Joseph Boyle, "Praeter intentionem in Aquinas," The Thomist 42 (1978): 649-65. See also John Finnis, "Object and Intention in Moral Judgments according to Aquinas," The Thomist 55 (1991): 10-14. 632 STEVEN JENSEN As was said, intention refers to the end insofar as it is the term of the movement of the will. In a movement, however, a term may be taken in two ways, either as the ultimate term, which is rested in and which is the term of the whole movement, or as some mediate term, which is the beginning of one part of the movement and the end or term of another part. For example, in the movement which goes from A to C by way of B, C is the ultimate term, while B is a term but not ultimate. Intention may bear upon both of these sorts of terms, so that while it is always of the end, it need not always concern the ultimate end. 15 The next article gives us an application of this teaching. Intention is not only of the ultimate end, as was said, but also of the mediate end. Someone may, however, intend both the proximate end and the ultimate end at the same time. For example, at the same time someone may intend both to prepare medicine and to regain his health. 16 Clearly, as Cajetan recognized when he cited this example while commenting on STh II-II, q. 64, a. 7, preparing medicine is a means to the end of attaining health, yet it is intended as a mediate end, which may also be called ad finem {as in STh I-II, q. 12, a. 4 ), usually translated as 'means'. In other words, it is certainly true that intention concerns the end, both ultimate and mediate, and not the means. The problem is that the means is also an end, albeit a mediate end. Preparing medicine, when viewed in relation to the goal of health, is a means; at the same time it is an end of the agent or an end of other actions, such as moving one's hands. When Aquinas says, then, that intention concerns the end and choice concerns the means, he does not exclude the possibility that they both concern the same object, which may itself be ad finem; one may both choose to prepare the medicine 15 STh I-II, q. 12, a. 2: "Respondeo dicendum quod, sicut dictum est, intentio respicit finem secundum quod est terminus motus voluntatis. In motu autem potest accipi terminus dupliciter, uno modo, ipse terminus ultimus, in quo quiescitur, qui est terminus totius motus; alio modo, aliquod medium, quod est principium unius partis motus, et finis vel terminus alterius. Sicut in motu quo itur de a in c per b, c est terminus ultimus, b autem est terminus, sed non ultimus. Et utriusque potest esse intentio. Unde etsi semper sit finis, non tamen oportet quod semper sit ultimi finis." 16 STh I-II, q. 12, a. 3: "Est enim intentio non solum finis ultimi, ut dictum est, sed etiam finis medii. Simul autem intendit aliquis et finem proximum, et ultimum; sicut confectionem medicinae, et sanitatem." LONG ON THE MORAL SPECIES 633 and intend to prepare the medicine. Both these can concern the means; they differ in the formality under which they move toward that means. Intention moves toward the means as a mediate end, while choice moves toward it precisely as something ordered toward a more ultimate end. 17 Long claims that 'intention', when applied to the means, is being used analogously. Unfortunately, he cites no passages of Aquinas indicating such. Article 12 of the Prima Secundae, of which we have quoted articles 2 and 3 above, directly addresses the topic of intention. Therefore, if Aquinas were using intention in an extended sense of the term, we should expect him to tell us so. Yet he does no such thing. He has a single meaning of intention in mind that applies both to the ultimate end and to the mediate end, both to health and to the preparation of medicine. In contrast, when Aquinas asks whether the act of willing (voluntas) concerns the means, he says that properly speaking it does not, leaving open an improper or analogous sense in which it does.18 Or again, when he asks whether enjoyment ifruitio) concerns only the ultimate end, he says that a mediate end that has its own pleasure may be enjoyed in some way of speaking. 19 Aquinas uses no such terminology when he speaks of someone intending the mediate end of preparing medicine. Intention is intention, whether it concerns the ultimate end or the mediate end. If one kills in order to save oneself, then the ultimate end is the preservation of one's own life, but a more mediate end is the means of killing. Intention applies to both of these ends, even though the second one may also be called a means. To read STh II-II, q. 64, a. 7, then, in a Cajetanian fashion, as prohibiting intention of the means of killing is not, as Long suggests, using 'intention' in an extended sense, apart from Aquinas's usual meanmg. 17 See STh I-II, q. 12, a. 4, ad 3. 18 19 STh I-II, q. 8, a. 2. STh I-II, q. 11, a. 3. 634 STEVEN JENSEN B) The Proximate End Gives Species Long might very well reply that while intention can refer to the mediate end, and hence the means, nevertheless, the intention that determines the species of actions, which is the concern of STh IIII, q. 64, a. 7, is the further end rather than the mediate end, for, as we have seen, the end gives the defining or formal species while the means merely provides a further determination. So it would seem, if we read merely STh I-II, q. 18, aa. 6 and 7. Unfortunately, once again, Long ignores some crucial texts suggesting that the intention that specifies is precisely the intention of the mediate end or means. Aquinas repeatedly affirms that the species of human actions come from the proximate end rather than the remote end. For example, he says, Profit or glory is the remote end of the dissembler, as it is of the liar, but the species is not taken from this end, but from the proximate end, which is to show oneself other than one is.20 In order to avoid possible confusion, I have chosen a text in which it is clear that Aquinas is referring to the proximate end of the will, and not the end of the exterior action. Another such text is STh I-II, q. 1, a. 3, the very text in which Aquinas establishes that human actions are specified by the end. The corpus leaves no doubt that Aquinas is speaking of the end of the will, yet in the reply to the third objection, he says, One and the same action, insofar as it arises from the agent, is ordered to only one proximate end, from which the act has its species, but it may be ordered to many remote ends, of which one is the end of the other. 21 In another place he says that, "Those things that are ordered to an end may be diversified by the end in two ways. In one way 20 SI'h II-II, q. 111, a. 3, ad 3: "Dicendum quod lucrum vel gloria est finis remotus simulatoris, sicut et mendacis. Unde ex hoc fine speciem non sortitur, sed ex fine proximo, qui est ostendere se alium quam sit." 21 SI'h I-II, q. 1, a. 3, ad 3: "Dicendum quod idem actus numero, secundum quod semel egreditur ab agente, non ordinatur nisi ad unum finem proximum, a quo habet speciem, sed potest ordinari ad plures fines remotos, quorum unus est finis alterius." LONG ON THE MORAL SPECIES 635 because they are ordered to diverse ends, and this makes for a diversity of species, most of all if the end is proximate. " 22 Elsewhere, he identifies the proximate end with the object: "Moral actions do not have their species from the remote end but from the proximate end, which is the object. " 23 It is not clear whether this object is the same as the object that Long considers to be the specifying element of the means or exterior action, the materia circa quam that is mentioned in STh I-II, q. 18, aa. 6 and 7, but there is good reason to suppose it is. In STh I-II, q. 73, a. 3, ad 1, Aquinas says, "The object, even if it is the materia circa quam in which the act terminates, has the formality of an end, insofar as the intention of the agent is led into it. " 24 The materia circa quam, then, specifies insofar as it is an intended end of the agent. When one kills in order to save oneself, the object of the act of killing serves as an object only insofar as it is an end intended by the agent. How are these texts concerning the proximate end to be reconciled with STh I-II, q. 18, aa. 6 and 7, which say that the remote end intended 25 is formal with regard to the species of actions? Unfortunately, the answer to this question is far from clear, and I will not attempt it here. The only point that I need to make is that the proximate end intended, which is the same as the means, does indeed give species. It is simply false to say that the 22 STh I-II, q. 107, a. 1: "Ea autem quae ordinantur ad finem, secundum rationem finis dupliciter diversificari possunt. Uno modo, quia ordinantur ad diversos fines, et haec est diversitas speciei, maxime si sit finis proximus." Aquinas is here concerned with species of law, but he makes a general statement about things that are ordered to an end, which include actions. I quote the text because of the strong emphasis upon the proximate end. 23 De Malo, q. 2, a. 6, ad 9 "Ad nonum dicendum, quad actus moralis non habet speciem a fine remoto, sed a fine proximo, qui est obiecturn." Again, it is clear that Aquinas is speaking of the end intended, for the objection reads, "The end gives species in morals, for moral actions are judged good and evil from intention" ("Finis dat speciem in moralibus, quia ex intentione judicatur actus moralis bonus vel malus"). See also De Malo, q. 2, a. 4, ad 9; De Malo, q. 2, a. 7, ad 8; STh I-II, q. 60, a. 1, ad 3; STh II-II, q. 11, a. 1, ad 2; STh II-II, q. 66, a. 4, ad 2. 24 STh I-II, q. 73, a. 3, ad 1: "Dicendum quad obiectum, etsi sit materia circa quam terminatur actus, habet tamen rationem finis secundum quod intentio agentis fertur in ipsum, ut supra dictum est." 25 That he is speaking of the remote end in these articles is clear from the examples he gives. STEVEN JENSEN 636 intention of the means does not gives species. When Aquinas says in STh II-II, q. 64, a. 7, that actions take their species from what is intended, he could not possibly have meant to exclude intention of the means. C) "Praeter Intentionem" Does Not Concern General Intention Long claims that in STh II-II, q. 64, a. 7, when Aquinas says that what is praeter intentionem is per accidens and hence does not give species to an action, he is referring to a general intention, to which the means are outside intention, just as a general intention to go to Chicago need not include any determinate means of getting there. Long provides us with no other texts in which Aquinas uses praeter intentionem in this way, especially texts in which he uses it to delineate the species of action. In fact, Aquinas sometimes does use praeter intentionem in this way. In the De Veritate, when Aquinas is wondering whether God creates with one or many ideas, he says that if someone has a general intention to make a triangle, then it is outside his intention whether it be large or small, and more generally he says that if an agent has only a general intention, then the details (which would include the means of bringing it about) are outside his intention. 26 However, I cannot find anywhere that Aquinas uses this sense of praeter intentionem when he is referring to the specification of actions. A cursory examination of examples indicates that Aquinas does not use praeter intentionem to exclude the means from giving species. He uses it, rather, to exclude what Cajetan wanted to exclude, namely, a consequence or side-effect of an action. 27 For example, Aquinas says that active scandal is not a distinct species when the spiritual downfall of one's neighbor is outside intention, "as when someone in his inordinate deeds or words intends only to satisfy his own will and not to give someone the occasion of ruin. " 28 Clearly, Aquinas does not mean the occasion 26 De Veritate, q. 3, a. 2. 27 Again, Boyle ("Praeter Intentionem, ") provides a fuller account of praeter intentionem. 28 STh II-II, q. 43, a. 3: "Ut puta cum aliquis suo facto vel verbo inordinato non intendit alteri dare occasionem ruinae, sed solum suae satisfacere voluntati." LONG ON THE MORAL SPECIES 637 of downfall is chosen as a means, for it is not a means at all. Rather, the occasion for ruin follows from his own sinful behavior as a consequence. Aquinas says that a sin does not take its species from its punishment, because the punishment is praeter intentionem. 29 He cannot mean that the sinner merely chooses the punishment as a means. Plainly, the sinner foresees the punishment as an undesirable consequence of his actions. Although all virtuous activities reveal the truth of oneself, as an act of courage reveals that one has courage, only in the virtue of truthfulness is this per se intended; for the other virtues, this selfrevelation is praeter intentionem and so it does not give species. 30 Once again, it is clear that the act of revealing one's nature does not serve as a means of accomplishing the brave act; rather, someone does a brave act, and consequently he reveals his nature. 31 It is not impossible, of course, that in STh 11-11, q. 64, a. 7, Aquinas is using praeter intentionem to exclude the means from giving species, even though elsewhere he uses it to exclude the consequences of action, but there seems no good reason to suppose that he is, apart from Long's a priori assumption that Aquinas must think it is legitimate to choose to kill in order to defend oneself. Long's use of praeter intentionem also forces him into an awkward reading of STh 11-11, q. 64, a. 7, ad 4, which we have previously quoted as saying, "The act of fornication or of adultery is not ordered to preserving one's own life from necessity, as is the act from which killing sometimes follows. " 32 Long concludes that adultery and fornication are not naturally ordered to saving I-II, q. 72, a. 5. 11-11, q. 109, a. 2, ad 2 31 There are many texts in which Aquinas uses praeter intentionem to exclude what is done in ignorance from giving species, but clearly these texts concern neither STh 11-11, q. 64, a. 7, nor Long's interpretation of it. See, for example, STh 11-11, q. 39, a. 1; STh 11-11, q. 59, a. 2; STh 11-11, q. 110, a. 1; and STh 11-11, q. 150, a. 2. 32 STh 11-11, q. 64, a. 7, ad 4: "Dicendum quod actus fornicationis vel adulterii non ordinatur ad conservationem propriae vitae ex necessitate, sicut actus ex quo quandoque sequitur homicidium." 29 STh 30 STh STEVEN JENSEN 638 one's life, so they cannot under the species of the end, namely, other has a natural order to self-defense. KiHing, on so it does fall under the species of the end. The difficulty Long's interpretation involves "the act from which sometimes follows." Cajetan, of course, supposes this action has killing as a side-effect. Long cannot read the killing as a side-effect, for he thinks the kiUing is chosen as a means. In sense, then, does killing upon the act? According to Long, it follows upon the act of seH-defense because defense, by itself, does not require homicide, even if this defense requires homicide. 33 In other words, Long is relying upon his notion of the general intention to defend oneself, to which the particular means, such as killing, are praeter intentionem. The means, in effect, follow upon the end, just so long as a variety of means is sometimes employed for the end. This reading is awkward enough as it stands, but upon scrutiny it becomes almost incoherent. What is act of killing? Is it the act that is naturally ordered to self-preservation, or is it what sometimes follows upon such an act? Long's position demands that killing is the act is naturally ordered to selfif it were not, then the act of killing would not preservation, be contained under the species of defense. The fourth reply, follows upon the act ordered to however, says that killing is self-preservation. It cannot be both. Either kiUing is itself ordered to preservation, or it follows upon the act ordered to preservation. We might also ask what act is naturally ordered to preserving one's life. As we have seen, Long requires that killing itself be naturally ordered to preserving one's life, but that is not how he reads the fourth reply" Rather, he says that killing follows upon the very act of self-preservation. And then the act from which killing sometimes becomes a sort of tautology: follows, which is the act of preserving one's life, is naturally ordered to preserving one's hie. There is simply no satisfying way to read the fourth reply given Long's interpretation. 33 Long, "Object of d1e Moral 61. LONG ON THE MORAL SPECIES 639 D) The Public Official Does Not Intend to Kill as an End Long says that it is wrong to intend to kill because one somehow desires the killing as good in itselfo The private individual, he says, can desire killing only as a means; the public official alone can want killing as desirable in itself. Long is wrong on both points. The private individual cannot desire killing even as a means, and the public official does not desire killing as good in itselt Let us quote once again the pertinent text from STh II-II, q. 64, a. 7: It is unlawful for a man to intend to kill a man in order to defend himself, except for those who have public authority, who while intending to kill a man for selfdefense, refer this act to the public good, as is plain for soldiers fighting the enemy, or for a minister of the judge fighting against robbers, although even these sin if they are moved by private desires. 34 The opening statement could hardly be better worded to describe intending to kill as a means to save oneself. Aquinas is not talking about killing as desirable in itself; he is talking about killing in order to defend oneself. This sort of killing, the sort that is ordered as a means to one's defense, is permissible for the public official but not for the private individual. And the public official must "ref er this act to the public good," which does not seem like a description of wanting something as an end, a good in itself. Long uses this notion of ordering to the common good to suppose that public officials might "shoot to kill even in defense against a merely diversionary or weak delaying tactic by a criminal band-not because of the gravity of the threat to the officers themselves, but because should the band succeed in escaping this would pose a threat to society at large. " 35 Long gives no evidence 34 STh II-II, q. 64, a. 7: "Illicitum est quod homo intendat occidere hominem ut seipsum defendat, nisi ei qui habet publicam auctoritatem, qui, intendens hominem occidere ad sui defensionem, refert hoc ad publicum bonum, ut patet in milite pugnante contra hostes, et in ministro iudicis pugnante contra lattones. Quamvis et isti etiam peccent si privata libidine 35 Long, "Object of the Moral Act," 63. 640 STEVEN JENSEN to indicate that this is the sort of case Aquinas has in mind. While Aquinas is talking about intending to kill in order to defend oneself, Long's case seems more concerned with intending to kill in order to prevent escape, two quite different situations. No doubt public officials are sometimes justified in killing for this reason, but this is not the case Aquinas has in mind. According to Long, Aquinas allows the private individual to kill as a means (at least for self-defense); what he prohibits is killing as an end. We do not have to go far from STh 11-11, q. 64, a. 7, however, to see that Long must be wrong. Just one article earlier Aquinas asks whether it is ever legitimate to kill the innocent. He answers that one may kill sinners by ordering it to the common good, but that the innocent may never be killed. 36 Three articles earlier, Aquinas asks whether the private individual can kill sinners. He answers that he cannot, for only those who have care of the common good may order a killing to the common good. 37 Killing of the innocent is never allowed; killing of evildoers is allowed only to the public officials; it readily follows that private individuals are never allowed to kill.38 Indeed, Aquinas goes so far as to say, Doing something for the benefit of the common good that harms no one is lawful for any private person, but if the benefit involves harm to another person, then it should not be done, except on the basis of the judgement of him to whom it pertains to decide what may be taken from the parts for the safety of the whole. 39 11-11, q. 64, a. 6. STh 11-11, q. 64, a. 3. 38 Long ("Object of the Moral Act," 63) quotes in his favor STh 1-11, q. 100, a. 8, ad 3, which says that the Ten Commandments forbid unlawful killing but allow lawful killing. But Long makes an unwarranted jump from the fact that there are lawful killings to the claim that there are lawful killings for the private individual. The passage he quotes gives only two examples of lawful killings, namely, killing evil-doers and killing enemies in war. STh 11-11, q. 64, a. 3, ad 1, which addresses the same sort of objection as STh 1-11, q. 100, a. 8, ad 3 (people killing under the command of God), does not say that these killings are legitimate for the private individual, but that these killing are done by God. 39 STh 11-11, q. 64, a. 3, ad 3: "Ad tertium dicendum quod facere aliquid ad utilitatem communem quod nulli nocet, hoc est licitum cuilibet privatae personae. Sed si sit cum nocumento alterius, hoc non debet fieri nisi secundum iudicium eius ad quern pertinet existimare quid sit subtrahendum partibus pro salute totius." 36 STh 37 LONG ON THE MORAL SPECIES 641 Other texts, as well, confirm our second point, that the public official does not intend to kill as an end in Long's sense. The virtue of vengeance, for instance, aims at someone's evil, but it does not aim at this evil in itself, but rather only insofar as some other good is attained through it. In the case of vengeance we must consider the mind-set of the one seeking vindication. If his intention is led primarily into the evil of the one upon whom he takes vindication, and rests in it, then vindication is in all ways unlawful, for to take pleasure in another's evil pertains to hatred .... On the other hand, if the intention of the avenger is led principally into some good that is attained through the punishment of the sinner (for example, that the sinner may amend, or at least that he be restrained and not disturb others, and to maintain justice and to honor God), then the vindication may be lawful, supposing that other necessary circumstances are observed. 40 Even God, when he punishes, does not seek the evil of death itself, but only the justice associated with it. Death may be considered in two ways. First, it may be considered insofar as it is a certain evil of human nature; as such, it is not from God but is a defect following on human sin. Second, death may be considered insofar as it has a formality of goodness, namely, insofar as it is a just punishment; as such, it is from God. Thus, Augustine says that God is not the author of death, except insofar as it is a punishment. 41 The public official, says Long, will act until he succeeds in killing the condemned criminal, but the private individual will act to kill only so long as the assailant poses a threat. 42 He concludes 40 STh 11-11, q. 108, a. 1: "Est ergo in vindicatione considerandus vindicantis animus. Si enim eius intentio feratur principaliter in malum illius de quo vindictam sumit, et ibi quiescat, est omnino illicitum, quia delectari in malo alterius pertinet ad odium ...• Si vero intentio vindicantis feratur principaliter ad aliquod bonum, ad quod pervenitur per poenam peccantis, puta ad emendationem peccantis, vel saltem ad cohibitionem eius et quietem aliorum, et ad iustitiae conservationem et dei honorem, potest essevindicatio licita, aliis debitis circumstantiis servatis." 41 STh11-11,q. 164, a. 1, ad 5: "Mors dupliciter potest considerari. Uno modo, secundum quod est quoddam malum humanae naturae. Et sic non est ex Deo, sed est defectus quidam incidens ex culpa humana. Alio modo potest considerari secundum quod habet quandam rationem boni, prout scilicet est quaedam iusta poena. Et sic est a Deo. Unde Augustinus dicit, in Libro Retractat., quod Deus non est auctor morris, nisi inquantum est poena." See also STh I, q. 49, a. 2; STh 1-11, q. 19, a. 10, ad 2. 42 Long, "Object of the Moral Act," 65. 642 STEVEN JENSEN that the former is killing as an end, while the latter is not. He should rather conclude, it seems, that killing remains a means the private individual so long as there is an imminent threat, but that killing remains a means for the executioner independent of any immediate threat. Interestingly, Long brings lack of innocence into his account of self-defense. He says that the assailant must be performatively non-innocent, that is, he must be engaged in threatening activity. 43 Apparently, this performative guilt is a condition for kHling to be naturally ordered to self-defense, and so to under the species of self-defense. According to Long, the doctor is not allowed to order to save the mother because the baby is not a baby performativdy guilty. 44 None of this appears in STh II-II, q. 64, a. 7, in which Aquinas never refers to the guilt or innocence of the assailant. m. WHAT TO Do WITH STH U-H, Q. 64, A. 7 My analysis here has been largely negative, showing that Long's interpretation of STh H-H, q. 64, a. 7 is flawed; I have to supply a better interpretation in its place. What can I say positively by way of interpreting intention within this text? Unfortunately, not much. I think it is a mistake to begin with STh H-H, q. 64, a. 7, which is a very difficult text. Furthermore, I think that most readers approach the text the wrong presuppositions and intuitions. If one supposes, as Long does (as weH as do those he opposes), that Aquinas is defending contemporary intuitions concerning self-defense, then one has Aquinas gives little evidence to already got off on the wrong indicate exactly situations he thinks meet his conditions for not. I think the conditions are very self-defense and which stringent and few instances of self-defense meet them. 45 But Ibid., 67. Ibid., 68. 45 For two accounts that recognize more stringent restrictions on self-defense see G. E. M. Anscombe, "War and Murder," in R. Wasserstrorn, ed., War and Morality (Belmont, Calif.: Wadsworth 1970), 45; and Thomas A. Cavanaugh, "Aquinas'sAccountofDouble Effect," The 43 44 LONG ON THE MORAL SPECIES 643 whether I am correct or not, it is a mistake to begin with this text, and it is a mistake to begin with presuppositions about what it is saying. In my mind, Aquinas's action theory should be examined and understood apart from this text; when this theory is understood well, then we can turn to his article on self-defense. If we begin as Long does (and in this regard he is little different from most others), and suppose that STh II-II, q. 64, a. 7 must justify actions in which we kill as a means to preserve our lives, then we might very well distort the meaning of the text to meet this presupposition. What if, after all, Aquinas thought such actions are not justified? I think such actions are justified, but then I also think that STh II-II, q. 64, a. 7 does not justify them. 46 To make the text justify such acts of self-defense one must stretch and twist the texts too much, as Long himself has done. STh II-II, q. 64, a. 7 is dear on a few points about intention: (1) the species of human actions are taken from what is intended; (2) what is praeter intentionem does not give species; (3) some actions that have two effects can have one effect intended but the other praeter intentionem; (4) the public official can intend something the private individual cannot; and given the wording, I think it is hard to get around (5) the public official can intend to kill as a means, while the private individual cannot. These five points are indeed a meager start, but then again, if I am correct, they are not a start at all. They should be the conclusion of an examination of other aspects of Aquinas's thought. If we wish to find within Aquinas a solution to the problem of the c:raniotomy case, we should look elsewhere than STh II-II, q. 64, a. 7. While intention plays a prominent role in Aquinas's analysis of self-defense, I suspect that it may play only a minor role in an understanding the nature of craniotomy. Thomist 61 (1997): 107-21. 46 I hold the view that in a situation of emergency the private individual takes on the role of the public official, and hence he defends himself not as a private individual. This view is not found in Aquinas, but it may be derived from his principles. BOOK REVIEWS Thomas Aquinas' Trinitarian Theology: A Study in Theological Method. By TIMOTHY L. SMITH. Washington, D.C.: The Catholic University of America Press, 2003. Pp. 258. $59.95 (cloth). ISBN 0-8132-1097-6. The Trinitarian theology of Thomas Aquinas is difficult and its interpretation remains a subject of controversy. The majority of twentieth-century theologians have severely criticized it: some have seen it as an attempt to provide a rational demonstration of the Trinity, others have characterized it as speculative reflection detached from the economy of salvation. Going back to Theodore of Regnon, and continuing through Michael Schmaus and Karl Rabner, many theologians have seen in St. Thomas the representative par excellence of an essentialist "Latin tradition" as opposed to a personalist "Greek tradition." Even today, manuals of theology continue to reproduce cliches of this sort. For this reason, Timothy Smith's work, which purports to show the inaccuracy of such interpretations, is a welcome contribution. Smith does not concentrate on the doctrinal content of Aquinas's theology but rather on his method, for a proper understanding of this method is required in order to have a correct reading of the treatise on the Trinity. The trajectory of this study is not linear. First of all, Smith provides an exposition of the context and the structure of the questions on the Trinity in the Summa Theologiae (chap. 1), then the order observed in the study of essence and of the divine persons (chap. 2), next the coordination of essential and proper terms (chap. 3, with a discussion of Trinitarian appropriations), and then the historical context of theological language (chap. 4). Finally, he underlines St. Thomas's originality in what constitutes the heart of the matter: naming God (chap. 5). This project allows us to lay to rest the methodological criticisms often leveled at Aquinas. Smith's research has much to recommend it. He aptly demonstrates that history and soteriology occupy a central place in the structure of the Summa Theologiae (12-20). He convincingly demonstrates that, for St. Thomas, the doctrine of the Trinity is "the interpretive framework for understanding all other doctrines" (29). In comparing Augustine and Thomas, Smith shows that it is impossible to speak of a single "Latin tradition" in Trinitarian theology (68-70, 119, 231). He also clearly establishes that for St. Thomas, the persons are never conceived of as a derivation of the divine essence. This is a veritable leitmotif: it is impossible rationally to demonstrate the Trinitarian processions (70-79, 645 646 BOOK REVIEWS "'"'"""',...,,,