10 December 2009

Sensation, Understanding, and Epistemology in the Summa Theologica’s Treatise on Man

Please note that what follows is still subject to editing.

The Summa Theologica’s Treatise on Man[1] devotes considerable attention to explaining how it is human beings are capable of understanding, and this is necessarily intertwined with Thomas Aquinas’ theory of perception. A result of such discussion is an understanding of the Treatise’s epistemology. The current discourse shall begin by describing the theory of man presented in the Treatise. Having done this, discussion will turn to Aristotle, affectionately known as the Philosopher. Specifically, an epistemological link shall be shown to exist between Aristotle’s Posterior Analytics and Aquinas’ Treatise through the views on human understanding presented in each, a link that shall prove to bear considerable utility in understanding the latter text. After delineating the Aristotelian epistemology, the Thomistic theories of perception and understanding shall enlighten the Treatise’s epistemology, and all shall be concluded with a comparison of the two.

In order to understand what the Treatise on Man is about, it is fitting to establish what a man, or otherwise a human being, is. There is always the Aristotelian position that man is a “rational animal,” which is referenced later on, but the Treatise begins simply by stating that man “is composed of a spiritual and corporeal substance”[2] These, of course, are respectively the human soul and the human body. It sounds as if Aquinas is about to adopt a form of substance dualism in the way that Descartes eventually would, but the student of Aristotle knows that such is not the view of the Philosopher. To show that this is not the case, a distinction is drawn in how a thing is able to subsist. Thomas explains:

This particular thing may be taken in two senses. Firstly, for anything subsistent; secondly, for that which subsists, and is complete in a specific nature. The former sense excludes the inherence of an accident or of a material form; the latter excludes also the imperfection of the part, so that a hand can be called this particular thing, in the first sense, as being something subsistent; but not in the second, for in this sense, what is composed of body and soul is said to be this particular thing.[3]

It is the composite of body and soul that is the substance, human being. Neither the body nor the soul is a human being, as an atomist or a Platonist would respectively argue. That the composite is the human being Aquinas asserts repeatedly.[4] Still, Aquinas claims, “the nature of the human intellect is not only incorporeal, but it is also a substance.”[5] The intellectual soul is substance. According to the above distinction, the composite human being subsists in the second way, while the body and intellectual soul subsist in the first. The body and intellectual soul are not “complete in a specific nature,” but rather are parts of a composite that is, each being akin to a severed hand in terms of subsistence. The hands cannot subsist unless the substance of which they had been a part first existed, just as neither body nor soul may subsist unless the composite first subsist.

This becomes even clearer when it is put in to Aristotelian, hylomorphic terms of matter and form. Form actualizes matter, which is potentiality, and the soul is the form of the body. Aquinas says, “since the soul is united to the body as its form, it must necessarily be in the whole body, and in each part thereof.”[6] Without a form, matter is undifferentiated, and without matter, a form is not actualized, as a form must be actualized in matter. In this way, neither a soul nor a body can come to be on its own, but only as a composite, as the soul is actualized through the body.

This is true of the nutritive, sensitive, and intellectual souls that Aquinas has borrowed from Aristotle. These are such that whatever is sensible is also nutritive, and whatever is intellectual is also both sensible and nutritive, but the combination of these parts is but one soul. The mere nutritive soul belongs generally to plants, while the sensitive soul belongs to all animals, and the intellectual soul belongs to human beings: “the difference which constitutes man is rational, which is applied to man on account of his intellectual principle. Therefore the intellectual principle is the form of man.”[7] Additionally, “in man the sensitive soul, the intellectual soul, and the nutritive soul are numerically one soul.”[8] Man’s is differentiated by his intellectual principle, and it is also nutritive and sensitive.

The nutritive and sensitive souls are actualized entirely by the corporeal body, as Aquinas says, “one cannot sense without a body.”[9] The intellectual soul, however, possesses the powers of intellection and volition, and these are not realized corporeally but immaterially. In this way the intellectual soul persists after the corruption of the body as a substantial form, for “the substantial form makes a thing to exist absolutely.”[10] This is why human beings persist after death, but brute animals do not; the brute animals do not have any substantial form. Even the nutritive and sensitive principles persist in human beings, not actually, but virtually, that is in virtue of the persistence of the intellect: “the composite being destroyed, [the powers of the sensitive and nutritive parts] do not remain actually; but they remain virtually in the soul, as in their principle or root.”[11] Because the intellectual soul persists, all its parts persist.

It is important that Aquinas’ highly nuanced theory of mind be generally understood, since without it, it will not be possible to appreciate his theories of perception and intellection, which shall be treated below. First, however, some comments must be made on Aristotle. As stated above, the epistemology of the Treatise is best understood in light of Aristotle’s Posterior Analytics, specifically the notable final chapter on “how [the principles] become familiar and what is the state that becomes familiar with them.”[12] First principles are necessary in the epistemology of APo because it is from these that demonstration is possible, where demonstration is defined as “a scientific deduction; and by scientific I mean one in virtue of which, by having it, we understand something,” which Aristotle goes on to explain must “depend on things which are true and primitive and immediate and more familiar than and prior to and explanatory of the conclusion.”[13] A demonstration is therefore akin to a sound deduction, whose premises are true and whose inferences are valid. The first principles serve as the true premises that fulfill all those conditions listed above, so all valid inferences drawn from them are necessarily sound. The principles may be accurately described as the foundations of the APo epistemology, and so the aforementioned purpose of II.19 is of central importance.

Returning now to how the first principles are known, Aristotle first eliminates the possibility of their being known innately, but rather they come to be known by some capacity, which is identified as perception. In order to explain this, Aristotle draws an analogy to the formation of a battle line following a rout. One soldier makes a stand, and others join him in support, and eventually “a position of strength is reached.”[14] After that, he explains:

when one of the undifferentiated things makes a stand, there is a primitive universal in the mind (for though one perceives a particular, perception is of the universal—e.g. of man but not of Callias the man); again a stand is made in these, until what has no parts and is universal stands—e.g. such and such an animal stands, until animal does, and so in this a stand is made in the same way. Thus it is clear that it is necessary for us to become familiar with the primitives by induction; for perception too instills the universal in this way.[15]

By perception of particulars or individuals, a universal begins to be formed until eventually, one is irrefutably understood. It likely sounds strange to think that the understanding of a universal should suddenly manifest following a certain set of perceptions; for example, Christopher Shields remarks, “there is something mysterious, even occult-sounding about Aristotle’s doctrine of nous [understanding] in the last chapter of the Posterior Analytics.”[16] Be that as it may, this is very near to Aquinas’ view, and in what follows a parallel will be drawn, showing that sensation and intellection in the Treatise allow for abstraction to a universal in exactly the way Aristotle describes above.

In the most general terms, sensation is for the purpose of receiving a phantasm, a “material image”[17] and “the likeness of an individual thing,[18] which can be abstracted to a universal in the intellect. Aquinas says, “sense is a passive power, and is naturally immuted by the exterior sensible,”[19] meaning that perception of the world is not an act of the senses, but is an act on the senses. He goes on to draw a distinction between natural and spiritual immutation. The former is exemplified by the immutation of heat into the thing heated, while the latter is exemplified by the immutation of the form of color into the pupil, whereby the pupil does not become colored. All the senses involve spiritual immutation, and all but sight involve natural immutation, also. If the sensory powers are actualized entirely in mater, how can it be that sensation involves a spiritual change? This has been the subject of some debate, which shall be related with all brevity.

The so-called “received interpretation” of D.W. Hamlyn, as Sheldon Cohen called it, put forth that the spiritual immutation was a non-physical change in the mind, a view that Cohen went on to challenge, arguing that in fact sensible forms are always received physically.[20] This position was challenged by John Haldane, who argued that both the Hamlyn and Cohen accounts were sufficiently meritorious as to imply contradiction on Aquinas’ part.[21] After this, Paul Hoffman claimed to settle the issue by arguing that Haldane had misunderstood Cohen as meaning that the physical reception of the forms meant it was a material reception, when in fact Cohen argued that it was the physical reception of immaterial forms. Hoffman proposed his own solution to the puzzle by arguing for the halfway state, meaning that there are different degrees between materiality and immateriality, and in this way spiritual immutation is possible in a corporeal organ.[22] Robert Pasnau challenges this view, himself arguing that Aquinas’ views on sensation can be construed in a way that is acceptable to the modern materialist, espousing a “semimaterialist” interpretation of Aquinas, which allows for “cognition to occur in wholly physical entities.”[23] This, however, has its opponents, such as Gabriele De Anna,[24] but at this point a sufficient summary of the main viewpoints on this subject has been given.

In light of the above views, it is worth noting that Aquinas elsewhere speaks of the spiritualization of the corporeal: “After the resurrection man will be like an angel, spiritualized in soul and body,”[25] and, “The empyrean heaven is the highest of corporeal places,” a fitting abode for angels.[26] What these texts indicate is, in fact, a gradation between corporeality and incorporeality, since corporeal creation is ruled from the empyrean heaven, and the increased spiritualization of soul and body at resurrection elevates man toward his final state in the empyrean heaven. These are but two texts, but they indicate that Hoffman’s discussion of gradation is likely on the right track. More persuasive, perhaps, is the sheer unity of body and soul necessitated by Aquinas’ Aristotelianism. Sensation, recall, is done by the composite of body and soul. It is realized in an entirely corporeal way, but only in virtue of the animal existing as a composite. Because an animal can be realized in an entirely physical way but still be ensouled by an immaterial form, a semimaterialist account also seems plausible; however, this reminder on no account makes the Hoffman view less plausible. Unfortunately, a detailed look at this particular debate is beyond the current scope.

Whatever the best interpretation of Aquinas may be, the fact that phantasms are received in the sense organs from sensible objects is established, and this is what really matters from an epistemological perspective. Thus the senses are said to receive phantasms, and “corporeal phantasms...are in corporeal organs.”[27] Through the phantasms of individual forms, it is eventually becomes possible to have understanding. According to Aquinas, the intellect is dependent upon phantasms in order to achieve understanding, and so he avers, “we must needs say that our intellect understands material things by abstracting from the phantasms.”[28] Intellection, however, involves two parts. There exist both an active intellect and a passive intellect in the soul. The passive intellect is in potentiality with respect to intelligible species, while the active intellect “is the cause of the universal, by abstracting it from matter,”[29] or in other words, it is the cause of the intelligible species, which it abstracts from phantasms. The intelligible species must be abstracted because sensible forms are not intelligible.[30] Additionally, “the object of our intellect in its present state is the quiddity of a material thing, which it abstracts from the phantasms,”[31] but a phantasm is a form existing immaterially by some combination of natural and spiritual immutation in the sense organ, which means that it is a differentiated and individualized, numerically distinct.[32] For such a thing to be understood intellectually, the active intellect must discern the quiddity from the individual, and that is abstraction to the intelligible species that is received into the passive intellect.

Abstraction itself requires that “the things which belong to the species of a material thing…can be thought of apart from the individualizing principles which do not belong to the notion of the species.”[33] Abstraction may be done from the particular to the universal and likewise from the phantasm to the intelligible species, and it is done by the active intellect. Once the active intellect has established a universal or intelligible species in the passive intellect, one is said to understand; however, it is also possible to misunderstand. In that event, the abstraction of a quiddity from a material thing is misunderstood as something else, for example abstracting the quiddity of a human being from what is in fact a sculpture. This can happen by some deficiency in the sense organs, such as farsightedness, or perhaps by inadequate experience of things of the intelligible species abstracted. If the latter is the case, then it makes sense that an individual who has only ever experienced one or two phantasms of something should not have a perfect grasp of its quiddity and cannot be said to understand perfectly, perhaps attributing certain accidents to the quiddity that necessarily do not belong.

By way of example if you see an individual Jabberwock, then the sensible form of that individual Jabberwock will be received into your sense organs as a phantasm. Abstraction from that phantasm will then be made by the active intellect, and an intelligible species, a universal Jabberwock, will then be received into the passive intellect. Still, having seen only one thing with this quiddity, it is not plausible to think that the quiddity will be perfectly understood; however, repeated experiences of various Jabberwocks and their respective phantasms do lead to a perfect understanding of the quiddity of the Jabberwock.

This should be quite familiar to the student of Aristotle, as is clear form a bit of replacement and interpolation, as indicated by boldness.

A single phantasm makes a stand, there is a primitive universal in the mind (for though one perceives a particular, intellection is of the universal—e.g. of man (an intelligible species) but not of Callias the man); again a stand is made in these phantasms, until what has no parts and is universal stands (is abstracted)—e.g. such and such an animal stands, until animal does, and so in this a stand (an abstraction) is made in the same way. Thus it is clear that it is necessary for us to become familiar with the primitives by induction; for intellection too instills the universal in this way.

Aquinas’ epistemology in the Treatise on Man is aptly described as a more robust version of the epistemology of the Posterior Analytics. Aquinas makes a plethora of comments to that effect, of which one specific example will suffice, but reference to other relevant texts is provided: “intellectual knowledge in some degree arises from sensible knowledge: and, because sense has singular and individual things for its object, and intellect has the universal for its object, it follows that our knowledge of the former comes before our knowledge of the latter.”[34] In both cases, sensation of the particular gives rise to the understanding of the universal, and eventually to infallible understanding of the universal.

It has been argued that Aquinas’ epistemology in the Summa Theologica’s Treatise on Man can be interpreted as an expansion of the APo II.19 epistemology of Aristotle, which can easily be construed as being empiricist and foundationalist, as the foundations of knowledge for both Aquinas and Aristotle are known empirically, and as both explicitly argue, not innately. Of course, it should come as no surprise that the theories of a devoted Aristotelian and Aristotle himself agree in this way.



[1] It might rightly be asked why only one text should be analyzed in discussing Aquinas’ views. The purpose is historical. Because Aquinas’ work was written over time, analysis of one text at a time does not attempt to find a single theory sensation, understanding, or epistemology to be representative of Aquinas; rather, interpretation by text allows for analysis within a single argumentative framework, and if each text is interpreted by itself, they may be compared to see if a comprehensive view exists. So, in that vein, here is a section of the Summa Theologica.

[2] St. Thomas Aquinas. Summa Theologica. Fathers of the English Dominican Province, tr. Ave Maria Press, Notre Dame, IN: 1948. Bk. I, Q75

[3] Ibid. Q75a2ad1. This particular thing might be taken to mean primary substance, as in Cat. 5, 2a13-18

[4] Ibid. Q75aa3-5

[5] Ibid. Q75a2 sed contra, See also Q79a1an1

[6] Ibid. Q76a8

[7] Ibid. Q76a1

[8] Ibid. Q76a3, See also Q77a4, Q78a1

[9] Ibid. Q76a1, See also Q75a3ad3, Q77a5, Q85a1, Q101a2

[10] Ibid. Q77a6, See also Q75a7

[11] Ibid. Q77a8, See also Q76a5ad2, Q89a7

[12] APo II.19, 99b17-18

[13] Ibid. I.2, 71a18-22

[14] Ibid. II.19, 100a10-14. The analogy is especially forceful when one thinks of the Greek phalanx.

[15] Ibid. 100a15-100b5.

[16] Shields, Christopher. Aristotle. Routledge, New York: 2007. pg 132

[17] ST I Q84a1 sed contra

[18] Ibid. Q84a7ad2

[19] Ibid. Q78a3

[20] Cohen, Sheldon. “Thomas Aquinas on the Immaterial Reception of Sensible Forms.” The Philosophical Review, Vol. 91, No. 2 (Apr., 1982), pp. 193-209

[21] Haldane, John. “Aquinas on Sense Perception.” The Philosophical Review, Vol. 92, No. 2 (Apr., 1983) pp.233-239

[22] Hoffman, Paul. “St. Thomas Aquinas on the Halfway State of Sensible Being.” The Philosophical Review. Vol. 99, No. 1 (Jan., 1990), pp. 73-92

[23] Pasnau, Robert. Theories of Cognition in the Later Middle Ages. Cambridge University Press, Cambridge: 1997. Pg. 36

[24] De Anna, Gabriele. “Sensible Forms and Semimaterialism.” The Review of Metaphysics, Vol. 54, No. 1(Sep., 2000), pp. 43-63.

[25] ST I Q98a2an1

[26] Ibid. Q102a2ad1

[27] Ibid. Q89a1

[28] Ibid. Q85a1, See also Q76a2, Q76a5ad4, Q86a2an2, Q89a1,2,4-7

[29] Ibid. Q79a5ad2, See also Q79a3ad3, Q79a4, 9

[30] Ibid. Q84a4ad2

[31] Ibid. Q85a8, See also Q85a6

[32] Ibid. Q76a2ad3, Q85a7ad3

[33] Ibid. Q85a1ad1

[34] Ibid. Q85a3, See also Q75a5, Q78a4, Q79a3ad3, a5ad3, a8-9, Q84a1 sed contra, a7, Q85a1ad1, a2, 6, Q86a1, Q89a2, Q89a6

12 November 2009

Respectability and the Promise of Superdupervenience: Comments on Horgan's "From Supervenience to Superdupervenience"

Superdupervenience as we know it was introduced by Terence Horgan (1993), though in a footnote he attributes the word itself to Bill Lycan (1986). As Horgan defines it, superdupervenience is “ontological supervenience that is robustly explainable in a materialistically acceptable way” (566). This definition is notable for its addition of the requirement that supervenience be “robustly explainable,” which Horgan says is necessary if ontological supervenience is to have materialistic “respectability.” This is never precisely defined, but the idea seems clear enough. A reasonable account of respectability in Horgan’s sense would seem simply to be the difference between supervenience and superdupervenience, but that is just explainablity. Respectability and explainability cannot be the same, however, since respectability is conferred by a “robust explanation;” it is not the explanation itself. Horgan’s favored example for a non-respectable materialism is G.E. Moore’s moral non-naturalism, and he cites J.L. Mackie’s criticism thereof in order to explain the contention, that is the non-respectability. In short, the trouble is that a non-respectable relation, such as the “supervenience” required by something like Moore’s moral non-naturalism, is not anything “in the world.” A respectable relation, then, should be explainable in the world, and here we have what Horgan calls superdupervenience.

The utility of superdupervenience hinges on the notion of materialistic respectability. If respectability is not useful, then superdupervenience is not useful. It is explainability that confers respectability, and it is explainability that differentiates supervenience and superdupervenience. Without respectability, explainability is not useful, and if explainability is not useful, there is no use for superdupervenience. We will be left just withsupervenience,[1] for better or for worse. Let the following clarify what has just been said:

P1 A relation is materialistically respectable if and only if it is robustly explainable.

P2 Supervenience is not in itself robustly explainable.

P3 Superdupervenience is by definition robustly explainable supervenience.

C1 Therefore, superdupervenience is materialistically respectable supervenience

C2 Therefore, materialistic respectability is not useful if and only if robust explainability is not useful.

C3 Therefore, if materialistic respectability is not useful, then superdupervenience is not useful.

C4 Therefore, if materialistic respectability is not useful, then supervenience may still be useful.

C4 leaves us with the possible utility of supervenience, still saddled with the problems superdupervenience was intended to solve, but only if either respectability or explainability turn out to be useless or unimportant.[2] The big problem with mere supervenience is that without explanation it is unacceptably sui generis to be “materialistically respectable.” Superdupervenience must therefore be considered within materialism, which shall now be done. Thus, it is for the sake of supervenience in a materialistic metaphysics (since this is not necessarily problematic for the non-materialist) that superdupervenience is invoked, as is clear from the very opening of Horgan’s paper.

If this is true, then, for materialism, mere supervenience might be considered useless or unimportant, and the implications of this for C4 are clear. If the additional premise “Supervenience is useful if and only if superdupervenience is useful” is taken into account, then C4 reads, “Therefore, if materialistic respectability is not useful, then supervenience is not useful.” At this point in the argument, superdupervenience is the only thing giving supervenience any utility at all, since superdupervenience requires supervenience by definition, but perhaps it is not clear that “Useful(supervenience) iff Useful(superdupervenience),” for materialism, of course. The “only if” clause is obviously true, since superdupervenience contains supervenience in its definition. The “if” clause comes from the very objection that superdupervenience seeks to resolve: supervenience is unacceptably sui generis and obscure and so forth, and hence not materialistically respectable. Of course, the biconditional holds only if superdupervenience is the only possible solution to the problem, but this just means that robust explanation is the only possible solution, which can only be true if robust explanation is the one and only sufficient condition for materialistic respectability. By saying that robust explainability confers materialistic respectability (in the relevant sense, anyway), it is admitted that robust explanation is the one and only sufficient condition thereof, and so robust explanation is assumed to be the only solution to the problem of supervenience’s materialistic non-respectability. Since robust explanation is the only solution, then superdupervenience is the only solution, so, “If Useful(supervenience) then Useful(superdupervenience).” Thus, “Useful(supervenience) iff Useful(superdupervenience).” Hence C4 shall read as I have claimed, since the necessary premise follows from Horgan’s definitions, assuming I have rendered them properly.

All this is to point to one conclusion. If one is to be a materialist and materialistic respectability is a useful concept, then superdupervenience is the only hope for supervenience, which is useless on its own. The problem is that “respectability” is not as clear a concept as one should like, and so it might be that some definitional finagling might be able to deftly dodge such a strong conclusion. It might, for example, be less closely tied to robust explainability than I have rendered it here, but I have tried to render Horgan’s meaning accurately. How that might be done is not obvious to me. One might also deny that supervenience as sui generis is respectful, after all, but in doing that, respectability would still probably need some definitional modification. One could also deny that “respectability” is meaningful, useful, important, etc. and thereby dodge the problem altogether, but if the denier is a materialist, Horgan shall likely quote Mackie and explain just how queer such a thing in a materialist ontology would be.

Superdupervenience is supposed to answer the problem presented by the concept of respectability, and if superdupervenience is possible, it appears to do just that. If respectability is not a chimera, then superdupervenience is a great hope for any materialist seeking to avoid reductionism, especially with respect to mind. Even if it is a chimera, then mere supervenience might still be useful. Because of superdupervenience, reductionism does not necessarily result for the materialist who accepts the concept of material respectability. Thus, Horgan asks a trio of questions which, if answerable, allow for a superdupervenient account, and these are the standpoint question, the target question, and the resource question, which respectively boil down to what sort of facts are needed, what specific facts are they, and do adequate explanatory resources exist for a superdupervenient account? Putting superdupervenience to work shall thus determine its usability, because if Horgan is right, that determines the usability of supervenience itself, and regardless of Horgan’s accuracy it is certainly true that the utility of a superdupervenience, which includes supervenience, certainly means that supervenience is useful, just not by itself, only alongside robust explanation.

Work Cited

Horgan, Terence. “From Supervenience to Superdupervenience: Meeting the Demands of a Material World.” Mind, New Series, vol. 102, issue 408 (Oct. 1993), pg. 555-586.



[1] For clarification, this discussion is about ontological supervenience, following Horgan.

[2] One might alternately phrase the argument in terms of importance rather than usefulness. I consider them to be approximately equal for my purposes. The argument may also be framed in terms of possibility and in terms of robust explainability in C3 and C4, which is allowed because of the biconditional in P1. This leads to the conclusion that if supervenience is not robustly explainable, then superdupervenience is not possible. The same claim can be made with respectability, of course. Such is largely analytic in superdupervenience so defined, though.