Chapter IV -- 17
Therefore, since the will is an active principle not determined to one thing but having itself indifferently [disposed] to many things, God moves it thus, because it is not determined from necessity to one thing; but its movement remains contingent and not necessary, except in those things to which it naturally moves.
St Thomas wishes to preserve human freedom in view of his doctrine that God is the only external cause which can move the will. He does this by means of a doctrine which he has taken from St Dionysios the Areopagite, that God acts on his creatures analogically with how he has created them. Here, since the human will was created contingent and not necessary, except in those things to which it naturally moves, notably the good and the Beatific Vision, God moves the human will in the same way, preserving the contingency and non-necessity of its movement, except in its desire for the good and for the Beatific Vision. St Thomas seems to mean that God moves our will by attracting it towards the good and towards its final end, beatitude.
St Thomas then turns to the concept of ‘enjoyment’.
St Thomas establishes that enjoyment is an act of appetitive power. St Thomas defines enjoyment, quoting St Augustine as follows: ‘To enjoy is to adhere with Eros (amore) to something on account of itself.’ He continues that enjoyment is of the end, and since the end is a matter of the will, which is an appetitive power, enjoyment is an act of the appetitive power.
St Thomas establishes that enjoyment has as its sole object the final end. But in St Thomas, the final end of man is beatitude, the Beatific Vision. We understand St Thomas to say in this article that the only full and proper and complete enjoyment that a man can have is of the Beatific Vision. However, St Thomas goes on to say, in the case of some other end, to which certain other things are referred and which has in itself a certain delectation, the end can in a certain way be said to be an object of enjoyment, but not properly and not according to the complete nature of an object of enjoyment.
St Thomas establishes that, given its definition as adherence with Eros (amor) to something on account of itself, enjoyment can occur even before possession of the object. This is clear: enjoyment is adherence with Eros to the object desired without the necessary implication of its actually being possessed.
There is a certain ambiguity in St Thomas’ treatment of enjoyment. On the one hand, enjoyment has its common-sense meaning. On the other hand, it is adherence to something with Eros for the sake of the thing itself. This second sense of enjoyment is quite similar to the psychoanalytic notion of ‘cathexis’, of the psychological bonding of the person to his love object.
However, the Thomist notion that enjoyment has as its proper object only the final end, beatitude, suggests that we cannot, according to the common-sense meaning of the term, enjoy anything, or, in the broader sense of the term, adhere with Eros to any end for its own sake, except the Beatific Vision, other than in an imperfect way.
But surely the bride and groom have each adhered with Eros to the other for the sake of the other even before the two have come together. And surely the football fan adheres with enjoyment to his favourite team and is grieved when it loses the Cup. And surely the political partisan adheres with enjoyment to the party and agonizes when the struggle is on and the issue in doubt. And surely the rock fan adheres with enjoyment to his favourite star and is elated when his idol puts out a new song. Yes, says St Thomas, but these enjoyments are improperly spoken of as enjoyment. But while it may be true that man’s final repose and good is to rest in the lap of the Beloved, is not St Thomas’ notion of enjoyment too restricted?
Gilson discusses St Augustine’s distinction between ‘enjoyment’ and ‘use’. This is important for us, because St Thomas discusses ‘use’ below, clearly depending on St Augustine for the concepts both of enjoyment and of use. As Gilson’s presents the matter, the Augustinian distinction between enjoyment and use seems close to M. Buber’s (1878–1965) notion of the distinction between the I-Thou and the I-It relationships: To enjoy something is to fix the will on a thing with love for the sake of that thing only. To use something is to make use of it as a means to obtain another end. Therefore one enjoys what he considers an end; he uses what he does not consider except as a means. According to Gilson, St Augustine goes on to say that, ultimately, we ought to enjoy only the final end, beatitude, or, indeed God, making all things means to that end. However, St Augustine has put all this on the psychological and moral plane, not on the plane of a metaphysical analysis of reality and of the will. St Thomas Aquinas has reinterpreted the Augustinian doctrine within the context of his Aristotelian metaphysics, changing the doctrine from a psychological and moral one to a metaphysical doctrine of the nature of the will. Moreover, his metaphysical analysis is carried out in the context of a strict rationalism. How well does St Augustine’s doctrine survive such a treatment, however? It is as if today we were to express the Augustinian doctrine by means of symbolic logic. The result would freeze the suppleness of St Augustine’s analysis.
In artificial intelligence, enjoyment would seem to have to do with the degree of adherence that the robot had to a goal, how easy it would be to deflect it to another goal according to circumstances which developed after the adoption of the first goal. These levels of commitment to a goal would have to be provided for in the design of the will subsystem. They would be different from the different degrees of emotional intensity generated by the desire and temper subsystems. However, the level of commitment by the will subsystem to a goal would necessarily, by way of messages to them, affect the way the desire and temper subsystems responded to the robot’s perceived context, just as, reciprocally, messages from the desire and temper subsystems could affect the will and intellect (evidently, the robot could become ‘conflicted’).
St Thomas next turns to ‘intention’. Intention is distinguished from the simple act of the will and from the goal of the will in that it is a conscious program or project—in ordinary language, intention—to do some specific set of actions which comprise both the end and the means to the end. For example, I want to become a monk. To be accepted into a monastery, I must first pay off my school debts. I formulate an intention to go to the Middle East to work at a high salary in the oil industry to pay off my school debts: I have applied for a job at such and such a company; they have accepted me; I am waiting for my visa. This is my intention: to go to the Middle East quickly to make a lot of money to pay off my school debts to become a monk. This is my end: to become a monk.
St Thomas establishes that intention is an act of the will and not of the intellect.
St Thomas establishes that our intention embraces both the end and the means to the end.
St Thomas next turns to ‘choice (electio)’. Choice is the actual selection of the means to the end. One way quickly to make a lot of money to pay off my school debts to become a monk is to rob a bank. However, by ‘deliberation’, I conclude that it would be better to go to the Middle East. I choose going to the Middle East to work in the oil industry over robbing a bank as the means to become a monk.
St Thomas establishes that choice is an act of the will and not of the reason. While by deliberation with my intellect I may conclude that it is better to go to the Middle East than to rob a bank, choosing to go to the Middle East is an act of the will.
St Thomas establishes that choice has as its object the means to the end, not the end itself. Quoting Aristotle, he says: ‘The will is concerned with the end; choice, however, with those things which are towards the end.’ In other words, my choice is not over whether to become a monk or not—unless the problem be seen in a much broader context than how I am now considering it, for example how I might obtain the Beatific Vision—but over the means by which I might become a monk.
In artificial intelligence, the implications are obvious. Once the end is adopted, then after the possible means to the end are found and evaluated, a means to the end is adopted—chosen—by the will subsystem.
We will discuss certain aspects of St Thomas’ doctrine of choice when we discuss ‘consent’, below.
St Thomas turns to ‘deliberation (consilio)’. This is the act of the reason to evaluate the possible means to the end.
St Thomas establishes that deliberation is a research.
In artificial intelligence, this is important: the designer of the robot must introduce heuristic programs into the intellect subsystem to search for possible means to the end. Here, given that computer systems are always closed, discrete systems, the designer must provide for ‘vocabularies’ or databases of means that he will preprogram the robot to be able to execute. The intellect subsystem must search its databases or repertoires of means to find an action, or even a sequence of actions, which will, according to evaluation criteria that the designer will have programmed, best ‘fit’ the end that the will subsystem will have adopted.
St Thomas establishes that deliberation has as its object only those things that are means to the end. Evaluating a possible goal is a different sort of thing from deliberating over the possible means to attain that goal.
St Thomas establishes that deliberation proceeds in an analytical fashion.
In artificial intelligence, the designer would proceed in the way we have just described.
St Thomas establishes that deliberation comes to an end somewhere.
In artificial intelligence, this is important. The search and evaluation algorithms in the intellect subsystem over the means must terminate, and within a reasonable time. Who wants to buy a robot that can’t make up its mind?
St Thomas discusses the next stage of the human act, ‘consent’.
St Thomas establishes that consent is an act of the appetitive power. In establishing his position, St Thomas quotes St John of Damascus. Here is what St Thomas says in the contrary assertion:
But it is the contrary, as Damascene says in Book II, 22, [of the Exact Exposition of the Orthodox Faith], that ‘If one judges and does not love (diligit), it is not opinion (sententia),’ that is, consent (consensus). But to love (diligere) pertains to the appetitive virtue. Hence, also consent.
It is important to see what St John says in the Greek original:
…Then he judges the better, and it is called judgement. Then he is disposed and loves the thing judged from the counsel, and it is called opinion (gnome). For if he judges and is not disposed towards the thing judged, that is to say, does not love it, it is not called opinion (gnome).…
We have italicized the sentence that St Thomas is quoting from the Latin translation of St John.
It seems to us that St Thomas is forcing St John of Damascus.
St Thomas’ use of St John of Damascus must be understood in the context of St Thomas’ definition of consent: the assent of the will to the good in each of the possible means to the end that the intellect discovers in its act of deliberation over the means to the end. What St John seems to have in mind, however, is the interior adherence to our judgement that we have when we adopt an opinion. While adopting an opinion is not far from assenting to the good in a set of possible means, it is not the same thing.
We wonder whether the problem is not this: in the passage of the Exposition that St Thomas obviously has before him, is not St John discussing and defining a loose and overlapping set of concepts, whereas St Thomas is taking him to describe what he himself wants to describe, a rigorously formal model of human action such as we might find today in symbolic logic? Moreover, as will be seen from a consideration of the full passage of St John, the ‘opinion (gnome)’ that St Thomas has assimilated to ‘consent (consensus)’ seems more to apply to the judgement that the end is good, not the means to the end. St Thomas seems to have been misled by the fact that St John refers to an examination of means before referring to the formation of the ‘opinion (gnome)’. However, the full passage indicates that when St John discusses the formation of the ‘opinion (gnome)’, he has returned to a discussion of the evaluation of the end.
The full passage of St John is this:
It is necessary to know that there is naturally sown in the soul an appetitive power (dunamis) desirous (orektike) of that being which is according to nature, which power embraces all those properties which are in [that being’s] nature essentially. This power is called the will (thelesis). For the essence of being and living and moving, when it aspires, desires, in accordance with the mind (nous) and the perception (aisthesis), its own natural and full being. For that reason they also define the natural will thus: The will is a rational and living appetite depending only on natural things. Therefore the will is that natural and rational appetite, or simple power; for the appetite of the irrational animals, not being rational, is not called will.
Purpose (boulesis), then, is a certain natural will (thelesis), that is to say, natural and rational appetite (orexis), for some thing. For there lies in the soul of man a power rationally to desire (oregesthai). When, then, that rational appetite moves (kineitai) towards a certain thing, it is called purpose (boulesis). For purpose (boulesis) is a rational appetite and aspiration for a certain thing.
Purpose (boulesis) is named both concerning those things that are in our power and concerning those things that are not in our power, that is, both concerning those things that are possible and concerning those things that are impossible. For we often purpose to commit fornication or to be chaste or to sleep or one of those sorts of things, and these things are those which are in our power and possible. We also purpose to reign as king; this is not one of the things that are in our power. Perhaps we also purpose never to die; this is one of the things that are impossible.
There is, then, the purpose (boulesis) of the end or of those things which are towards the end. On the one hand, then, that which is willed is the end, as to reign as king, as to be in good health; on the other hand, there is the matter for deliberation towards the end, that is to say, the means through which we must be in good health or reign as king. Then, after the purpose (boulesis), search and thought. And after that, if the matter is one of those things which is in our power, there occurs counsel (boule), that is to say deliberation (bouleusis). Counsel (boule), then, is a searching appetite that occurs concerning those things which are in our power to be done. For he deliberates if he must pursue the thing or not. Then he judges the better, and it is called judgement. Then he is disposed and loves the thing judged from the counsel, and it is called opinion (gnome). For if he judges and is not disposed towards the thing judged, that is to say, does not love it, it is not called opinion (gnome). Then after the disposition there occurs intention (proairesis), which is to say, choice (epiloge); for intention (proairesis) is to choose and select this in preference to that in regard to two things which lie before us. Then he rushes towards the act, and it is called impulse. Afterwards he uses, and it is called use. Then after the use he ceases from the appetite.
Although there are obvious connections and similarities between St John and St Thomas here, St John is not so rationalistic and rigorously formally logical as St Thomas. Moreover, the doctrine of consent that St Thomas is advancing can only with a certain forcing be derived from this passage of St John.
Let us make some remarks on this admittedly somewhat obscure passage of St John. First, St John does, evidently following Aristotle, define the will, with St Thomas, as a rational appetite. However, in St John, the will is an appetite ‘desirous (orektike) of that being which is according to nature’, and it ‘embraces all those properties which are in [that being’s] nature essentially’: we take this to refer to the (well-) being of the soul itself: ‘For the essence of being and living and moving, when it aspires, desires, in accordance with the mind (nous) and the perception (aisthesis), its own natural and full being.’ We see nothing here of St Thomas’ doctrine that the will tends necessarily to the Beatific Vision, and to secondary goods only insofar as they bear a metaphysical relation of goodness to the Beatific Vision. While such a doctrine might on the basis of Aristotelian metaphysics be inferable from what St John says, St John has here made no such inference. The doctrine of St Thomas is his own Aristotelian adaptation of the related doctrine of St Augustine. St John has a more flexible and open doctrine of the will than St Thomas: while the soul certainly finds ‘its own natural and full being’ in perfection in the gnosis of God himself, St John’s doctrine is more supple as a doctrine of human psychology than St Thomas’ doctrine.
We next see St John’s definition of purpose (boulesis): it is ‘a rational appetite and aspiration for a certain thing’: we take this to be the end or the means towards the end: ‘There is, then, the purpose (boulesis) of the end or of those things which are towards the end.’ St John continues with a passage that seems to mix discussion of evaluation of the end with discussion of evaluation of the means to the end. For after purpose (boulesis), he names ‘search and thought’, seemingly passing from the end to the means to the end. He then passes to counsel (boule). St Thomas has adopted counsel (boule) as deliberation (consilio), but taken as search and thought only over the means to the end, at least insofar as a formal stage of his theory of action is concerned: that appears to be how St Thomas has taken ‘concerning those things which are in our power to be done’—as being concerned only with the possible means to the end. However, when St John then says, ‘For he deliberates if he must pursue the thing or not,’ is he not then discussing deliberation over the suitability of the end, something that St Thomas himself recognizes? If that is so, as we think, then ‘the better’ that the man judges ‘concerning those things which are in our power to be done’ is not the best means to the end, but which end the man should pursue, or even whether he should pursue an end at all, although St John is not being so rigorously formally precise that deliberation over the means can be excluded. However, if our interpretation be correct, then St John goes on to say that the man is positively disposed towards the end judged best from the counsel and loves that end, and that this is called ‘opinion’. This ‘opinion’ is not a matter of consent to the means, taken as adherence of the will to the good in each and every possible means to the end, as St Thomas would have it, but of the formation of and adherence to an actual opinion that that end is what the man actually ought to do. St John then proceeds to intention, which is a choice of one thing over another. While this certainly is consistent with choosing one means to an end over another, we wonder if St John is not still discussing the choice of one end over another.
If we take St John to be talking about the end when he discusses the formation of the opinion (gnome), is not the positive disposition and love of the man for ‘the thing judged from the counsel’ to be ‘the better’, the enjoyment of the end that St Thomas himself refers to, that psychological bonding to the love object that we spoke of that occurs even before the end is attained? In that case, there would be no support for St Thomas’ use of the passage in his definition of consent as the assent of the will to the good in each of the possible means to an end that the intellect discovers in its act of deliberation.
Let us continue with St Thomas: St Thomas establishes that consent is only towards those things that are means to the end, not towards the end itself.
In St Thomas, what is the difference between enjoyment and consent? For in St Thomas’ doctrine, consent appears to be a mild version of enjoyment. In St Thomas, enjoyment relates to the end, and especially to the final end, beatitude, whereas consent relates to the good in all possible means to attain a certain contingent end. Gilson points out that we make an act of the will, called consent, that assents to the good in each of the possible means that we encounter in our intellect’s deliberation over the means to the end, making it necessary for us to make another act of the will, choice, before we can proceed to action. This agrees with certain of St Thomas’ remarks to which Gilson refers.
Surely, however, given that St John of Damascus has defined opinion (gnome) as arising from the judgement of ‘the better’, St Thomas’ definition of consent is a forced interpretation. It seems to us that St Thomas is here trying to fit the loose and overlapping set of concepts of St John into a rigorously formal model of human action that has no untidiness. But that leads him to rationalistic excess. For we consent to the best means, not to the good in every possible means that we discover by deliberation: who makes an overt act of the will consenting to the good in every possible means that he discovers by an act of deliberation of his intellect, an overt act of the will different from the actual choice of the means that he actually will pursue?
However, in artificial intelligence, the designer of a robot should proceed in the way St Thomas has outlined: he should use heuristic programs in the intellect subsystem to find all the possible means to the end; he should ‘score’ each means on the basis of preprogrammed criteria—this would be an assessment by the intellect subsystem of the good in each possible means—; and he should have the intellect subsystem send to the will subsystem for its consent a message which lists all the possible means with their ‘scores’. Then it is a matter of the robot’s choosing the best means to the end.
St Thomas seems to say that it is not the intellect which determines which means will be adopted at the stage of choice, but the will. For, St Thomas says, once the intellect has deliberated over all the means and the will has consented to the good in each of the possible means, then the will chooses the means. But there is no provision in St Thomas’ theory for the will to deliberate over anything. This seems to imply that the will makes an ‘existential leap’ in choosing one of the possible means, but that seems uncharacteristic of St Thomas: surely the reason commands the will which means to adopt by providing the will with its choice. St Thomas does, however, introduce the notion of residual influences of the will on the intellect and of the intellect on the will, evidently in order to address this issue.
It seems to us that the weakness of St Thomas’ model of human action is in its treatment of the freedom of man and that this weakness is manifest in the above considerations.
In artificial intelligence, the designer should have the will subsystem necessarily choose the means with the highest ‘score’: it would not make sense for the will subsystem to override the ranking of the means provided by the intellect subsystem: that would introduce other criteria into the ranking.
St Thomas turns to the final stage of the human act, ‘use (usus)’. In discussing enjoyment, above, we have already seen that St Augustine defines ‘use’ as a sort of I-It relationship in contradistinction to the I-Thou relationship of ‘enjoyment’. St Thomas’ ‘use’ is just what we would understand today by ‘use’: the use we make of something, especially of the bodily members, to bring about the goal through the means we have chosen.
St Thomas establishes that use is an act of the will. It is the will that makes use of the bodily members, not the intellect.
In artificial intelligence, we have already seen this: the will is the executive subsystem. It is the will subsystem that will put the other subsystems—including the intellect subsystem—and the physical members of the robot’s body into action in order to accomplish the means to the end.
St Thomas next discusses the last aspect of the human action, ‘command’. There are two aspects of command: the actual act of command and the act commanded. The act of command is the issuance of the order; the act commanded is what the order says should be done. Here we understand that ‘command’ is taken by St Thomas to refer to how the powers of the soul and the members of the body are put into motion in order for the means to the end to be accomplished. Although his remarks can certainly be read in a broader light—for example, as bearing on how the intellect comes to present the will with its end—the placement of the discussion of ‘command’ in his exposition and the general structure of his theory of action seem to restrict the notion of command to the actual accomplishment of means.
St Thomas establishes that command is an act of the reason, an act of the will being presupposed. St Thomas writes these important but difficult lines on the subject:
To command is an act of the reason, there being supposed, however, an act of the will. Towards the proof of this it is necessary to consider that since acts of the will and of the reason can act on each other reciprocally, as, that is to say, the reason reasons concerning what is to be willed, and the will wills to reason, it occurs that an act of the will comes from an act of the reason and conversely. And since the virtue (virtus) of the prior act remains in the act which follows, it sometimes happens that there is a certain act of the will according to which there remains in it by virtue something of an act of the reason, as has been said concerning ‘use’ and ‘choice’; and, conversely, a certain act of the reason, according to which there remains in it by virtue something of an act of the will. However, to command is essentially an act of the reason; for, commanding, it orders, by intimating or declaring, the [power] that it orders, to do something: however, it is [the nature] of the reason thus to command [a power] after the fashion of this intimation. But the reason can intimate or declare something in two ways: In one way absolutely, and here the intimation is expressed by a verb in the indicative mood, as if one were to say to another: ‘This is to be done by you.’ However, at another time, the reason intimates something to someone by moving him to it; and this sort of intimation is expressed by a verb in the imperative mood, for example when it is said to someone: ‘Do this!’ However, as has been said, among the powers of the soul, the first mover to the exercise of an act is the will. Therefore, since the second mover does not move except in virtue of the first mover, it follows that that itself which the reason moves by commanding should be in it out of the virtue of the will. Whence it remains that to command is an act of the reason, there being presupposed an act of the will, in virtue of which the reason moves [a power] to the exercise of the act by means of command.
The gist of this passage is that it is the intellect which tells the will what to do. However, the intellect must first have been put into motion by the will, since it is the will which must move the other powers of the soul to act. St Thomas seems to be running the risk of an infinite regress here, since it is the intellect that presents to the will its object. Where does the cycle start?
In artificial intelligence, the implications are clear.
St Thomas’ doctrine of the reciprocal interactions of the intellect and the will creates serious conceptual problems for his model that are evident in this article. For he wants to say that to command is the sort of thing that applies to the reason, that it is an intellectual sort of thing. However, as he puts it, ‘[A]mong the powers of the soul the first mover to the exercise of an act is the will.’ To resolve this contradiction, he must introduce a doctrine of residual influences: When the reason commands the will, because command is the sort of thing that the reason and not the will does, it does so because the will has first moved the reason to act and that act of the will has left a residue in the reason. However, when the will ‘chooses’ or ‘uses’, because choosing or using is the sort of thing that the will and not the reason does, it does so because the reason has deliberated over the means to the end, and that act of the reason has left a residue in the will. This last appears to be how St Thomas handles the objection that the act of the will in choosing the means appears to be an ‘existential leap’ since in his system there is no provision for the will to deliberate: he says that in the act of the will’s choosing there remains a residue of the reason’s deliberation over the means. Surely this logical complexity demonstrates the weakness of St Thomas’ model: in his effort to produce a logically rigorous model of human action he has got himself into serious logical difficulties which require the supposition of residual influences of one power of the soul on the other. While in the context of his Aristotelian metaphysics, his solutions may be precise and well-taken, surely because of this logical complexity his model fails as a model of human psychology! For while on the philosophical plane we may agree with his theoretical solutions, do they have anything to do with the lived experience of ordinary men?
In artificial intelligence, St Thomas’ remarks in this article are well-taken. The designer should consider what St Thomas is saying.
St Thomas next establishes that to command precedes use. This is simply the common-sense notion that we command the use of a bodily member before we put it to use. We obviously do not command the use of our bodily members after they have been put to use. The model here is the soldier who obeys the command of the general: he does not obey before the general commands but after.
St Thomas establishes that the command and the act commanded are one and the same thing. That is, when I command my bodily members, and this is taken to be a human act, the command—to raise my arm, say—and the act commanded—the raising of my arm—are one and the same thing. That is, there is no gap between the command and the act commanded: taken as a human act; they are one and the same thing: I raise my arm. However, according to St Thomas, this unitary human act can be analysed into parts and I can logically distinguish between the command of reason—that my arm be raised—and the actual act commanded—the raising of my arm.
St Thomas next establishes that an act of the will can be commanded. St Thomas says this:
As has been said, command is nothing else than an act of the reason with a certain motion ordering (ordinantis) [a power] in regard to something that is to be done. However, it is manifest that the reason can order (ordinare) concerning the act of the will; for just as it can judge that it would be good to will something, thus it can order, by commanding, that the man will: from which it follows that an act of the will can be commanded.
In artificial intelligence, this raises the interesting problem of those cases in which the intellect subsystem commands the will subsystem to will something, evidently on the basis of a prior command of the will subsystem to the intellect subsystem to evaluate some matter.
St Thomas establishes that acts of the reason can be commanded. He says this:
Since the reason reflects on itself, as ordering concerning the acts of the other powers, thus can it even order concerning its own act: whence even its own act can be commanded.
St Thomas has some provisoes concerning whether the reason can command the comprehension or assent of the reason itself.
The significance of this passage, which is important, is that in St Thomas’ doctrine, it is the will which ‘makes things happen’, but it is the intellect which commands the will ‘what it should make happen’. Hence, in St Thomas’ doctrine, the reason can give a command to the will to put the reason itself into operation: I conclude by reason that a good means to progress spiritually would be to think about my salvation and my reason commands my will to put my reason into operation to think about my salvation. As a matter of logical analysis for the designer of a robot, this is, honestly, very valuable advice. But as a matter of human psychology, has not St Thomas been carried away by his rationalistic and intellectualistic effort at formal logical precision?
In artificial intelligence, the implication is that the intellect subsystem can command the will subsystem to direct itself, the intellect subsystem, to conduct certain evaluations or calculations. The designer will have to be careful to avoid an endless regression or ‘loop’ in such interactions between the two subsystems.
St Thomas next establishes that the act of the sensitive appetite can be commanded, but not insofar, first, as it rests on the disposition of a bodily organ nor insofar, second, as the movement suddenly tends independently of the reason towards the apprehension of the imagination or sense.
In the first case, the key to St Thomas’ argument is that the powers of the soul other than the reason and the will depend on the body to a greater or lesser extent. To the extent that a movement of the sensitive appetite depends on a movement of the (sensitive) soul, it can be commanded; to the extent, however, that it depends on a movement of the body, it cannot.
In the second case, St Thomas does not explain why a sudden movement of sensitive appetite towards an apprehension of the imagination or sense is outside the control of our reason. He seems to accept the matter as a common fact of life, and observes that if the reason were able to control such a movement, then it would be able to prevent it, something that he considers impossible.
This second case is very important for us, for these sudden movements of the sensitive appetite towards imagination or sense are the basis of the ascetical model of temptation and sin of Evagrius Pontikos, St Mark the Ascetic and St Hesychios that we discussed earlier. For the ascetical program that leads to dispassion (apatheia)—a concept surely foreign to St Thomas—is centred precisely on these sudden movements of the imagination and sense. These are the initial images that we spoke of that occur in the mind’s eye of the ascetic.
Moreover, the first case, the movements of the sensitive appetite which depend on a disposition of a bodily organ, is in the ascetical theory of Evagrius Pontikos the object of a bodily ascesis for the diminution of the bodily passions.
In artificial intelligence, the implication is that within certain limits, the intellect subsystem must be able to send instructions to the will to ‘reset’ the desire and temper subsystems—to calm them down, for example, or even to make them a little more excited so that they take things a little more seriously.
St Thomas next establishes that the acts of the vegetative soul are not subject to the command of reason. His reasoning is that the acts of the vegetative soul are not apprehended (i.e. not consciously perceived), and, since the reason operates by means of the power of perception, it is impossible that it should be able to control them.
However, in the ascetical program of Evagrius, these involuntary movements of the vegetative soul, which always have a bodily substrate, are the object of a bodily ascesis, although not entirely.
In artificial intelligence, the implication is that it would not do for the intellect subsystem to tell the vegetative subsystem not to say to the desire subsystem that the battery is low: when the battery is low, it’s low.
St Thomas establishes that the acts of certain bodily members such as the hands can be commanded, whereas the acts of certain other bodily members cannot. He says this:
The members of the body are certain organs of the powers of the soul: whence, in that manner by which the powers of the soul are obedient to the reason, in that same manner the members of the body are also. Therefore, since the powers of the sensitive soul are subject to the command of reason, not, however, the natural [vegetative] powers, on account of this, all the movements of the members which are moved by the sensitive power of the soul are subject to the power of the reason; however, the movements of the members which follow the natural [vegetative] powers are not subject to the command of reason.
What is St Thomas saying? All the bodily members are organs of one or another power of the soul. Since each member of the body is the organ of a certain power of the soul, it is subject to the conscious control of the reason precisely to the extent that the corresponding power of the soul is subject to the conscious control of the reason. The two types of powers that St Thomas considers are the vegetative and the sensitive. As we have already seen, the vegetative powers of the soul—this is the realm of biochemical reactions, endocrine reactions, reactions of the sympathetic and parasympathetic nervous systems, in a word, of involuntary bodily responses—are not under the conscious control of the reason. Hence, St Thomas says, the bodily members controlled by the vegetative powers of the soul are not subject to conscious control. This is clear: there are many bodily reactions that are involuntary.
However, the sensitive powers of the soul are subject to conscious control by the reason. Therefore the bodily members controlled by the sensitive powers of the soul are subject to conscious control by the reason. However, it is well to recall that an overwhelming by a passion, which overwhelming has a bodily substrate, can incapacitate the reason.
It is well to note in passing that St Thomas here ignores those cases where a member of the body is partly under the control of the vegetative soul and partly under the control of the sensitive soul. Some members of the body are immune from conscious control under some aspects and subject to conscious control under other aspects. This, however, does not present any serious difficulty for St Thomas’ analysis, being a mere detail.
In artificial intelligence, the implication is that it would not do for the intellect subsystem to meddle with the support circuitry of the vegetative subsystem. That should be handled directly from within the vegetative subsystem, leaving the intellect subsystem to its work of evaluating ends, evaluating means, and logical and arithmetic computation.
In the ascetical program of Evagrius Pontikos which underlies the Philokalia, the involuntary movements of the bodily members due to the vegetative soul are the object of bodily ascesis.
 ST Ia IIae, 10, 4.
 ST Ia IIae, 11.
 ST Ia IIae, 11, 1.
 De Doctr. Christ. 1 and De Trinit. 10.
 ST Ia IIae, 11, 3.
 ST Ia IIae, 11, 4.
 Gilson Aug p. 217–8.
 ST Ia IIae, 12.
 ST Ia IIae, 12, 1.
 ST Ia IIae, 12, 4.
 ST Ia IIae, 13.
 See below.
 ST Ia IIae, 13, 1.
 ST Ia IIae, 13, 3.
 Nicomachean Ethics, 3.
 ST Ia IIae, 14.
 ST Ia IIae, 14, 1.
 ST Ia IIae, 14, 2.
 However, see our remarks under St Ia IIae, 10, 2, above, concerning deliberation over the end.
 ST Ia IIae, 14, 5.
 Under ST Ia IIae, 14, 1.
 ST Ia IIae, 14, 6.
 ST Ia IIae, 15.
 ST Ia IIae, 15, 1.
 In ST Ia IIae, 6, 2.
 In ST Ia IIae, 11.
 ST Ia IIae, 15, 3.
 Gilson E p. 309.
 In ST Ia IIae, 15, 3, ad 3.
 ST Ia IIae, 15, 3, ad 3.
 ST Ia IIae, 17, 1.
 ST Ia IIae, 16.
 ST Ia IIae, 16, 1.
 ST Ia IIae, 17.
 ST Ia IIae, 17, 1.
 ST Ia IIae, 9, 1.
 ST Ia IIae, 17, 3.
 ST Ia IIae, 17, 4.
 ST Ia IIae, 17, 1.
 ST Ia IIae, 17, 5.
 ST Ia IIae, 17, 6.
 ST Ia IIae, 17, 7.
 In discussing ST Ia IIae, 10, 3, above.
 ST Ia IIae, 17, 8.
 ST Ia IIae, 17, 9.