In Paradise Lost Milton’s prelapsarian Eve is given a dream by Satan, in which she walks alone “to the tree of interdicted knowledge.”1 As she tells Adam upon waking, her dream-self finds the tree more fair than her waking self had previously, and its fruit “so quick’nd appetite, that I, methought, could not but taste.”2 This worries Eve, but Adam reassures her, arguing that while he dislikes her evil dream, she harbors no evil of her own:
Yet evil whence? in thee can harbour none,
Created pure . . .
Evil into the mind of God or Man
May come and go, so unapprov’d, and leave
No spot or blame behind: Which gives me hope
That what in sleep thou didst abhorr to dream,
Waking thou never wilt consent to do.3
Adam’s theory is that dreams are produced not by the will but by “fancy,” which shapes imagination. In persons who are asleep, imaginings are free from the influence of reason, and thus cannot receive consent, but only be viewed.4 There is nothing that might be considered sin in Eve’s unsettling dream of a fall, because “the approval of the will alone makes a mind evil [but] the presence of evil as an object of thought does not.”5
St. Augustine sympathizes in his Confessions with Milton’s view that persons should not be held responsible for what they do in their dreams. Even in that relatively early work, however, Augustine is more uncertain about this view than Milton, and over time Augustine nuances his treatment of the possibility of sinning in one’s dreams. This essay focuses on his mature views, in comparison with his earlier thoughts about dreams, and in doing so provides a novel angle from which we can view Augustine’s mature thoughts about the human body, sin, and responsibility as well.
Our discussion will be easier to follow if I begin by clarifying what I mean by two key terms. First, to inquire whether we can be responsible for the evils in our dreams is to ask whether the content of our dreams is attributable to us in such a manner that it is fair to consider us accountable for them, and therefore to blame us for them. For the sake of clarity, I refer to this sense of the term “responsibility” as “deep responsibility.” The term “responsible” is also often used to mean being a virtuous person, or one who fulfills one’s duties or roles, but that sense of the term is not the one with which we will be primarily concerned in this essay.6
Second, it may be helpful to indicate what I take a dream to be. I will follow the suggestion of recent dream researchers that dreams are “an experience of complex and organized perceptual imagery” that undergoes “some temporal process or change”; a definition that excludes “the isolated visual images, fragmented auditory recall, and thoughts that are also part of the more general category of sleep mentation.”7
Augustine has a number of reasons to believe that it is possible to sin in dreams, but it takes him some time to fully acknowledge and develop the sometimes radical implications of his own thought. He is forced to do so, however, in his controversies with the Pelagians, especially in response to his most capable opponent, Julian of Eclanum. As a result, Augustine’s latest works tend to be more internally consistent, at least on the topics of sin, grace, and responsibility, than those written in the middle of his career.8 We will focus on those works later in this paper, but let us first turn our attention to his initial treatment of what Owen Flanagan calls “Augustine’s problem” about dreams.9
In the Confessions, Augustine is divided over whether he can sin in his dreams. He explicitly concludes that he cannot, because “the very difference between sleep and waking is obvious enough to convince us that we did not really do the disgraceful thing [we dreamed], even though we are sorry that it was in some sense done in us” (Conf 10.30.41).10 Yet, as Gareth Matthews astutely notes, Augustine’s conscience is obviously not clear as he reflects on the content of his dreams. Moreover, Augustine’s conclusion that he is not responsible for sexual or other improper content in his dreams is unwarranted because “he has no way to understand the difference between waking and sleeping that will justify his putatively clear conscience.”11
Augustine seems to find it difficult to believe that he can sin in his dreams because he finds something like the following argument persuasive: “I am not directly responsible for what I dream because I do not have properly robust ownership of what happens in my dreams.” This argument allows for the possibility that one might be indirectly responsible for one’s dreams and dream-actions. If one could voluntarily take steps to influence the content of one’s dreams—perhaps by trying to practice the ability to have “lucid” dreams, in which dreamers are not only aware that they are dreaming but apparently often able to voluntarily control what they do in their dreams—one might be able to influence one’s dreams in such a manner as to take ownership of them, thereby making oneself responsible for their content.12 Most of us, however, lack such control over our dreams. Moreover, we tend to doubt—as Augustine seems to have doubted—that we can gain such a high degree of control over our dreams. Thus, we wonder if we “have” our dreams in the manner necessary to be accountable for them.
When Augustine raises this question, he quickly rejects one reason for thinking that we do not own our dreams: a strong dissociation of one’s self from one’s dream-self. One owns one’s evil dreams in at least the weak sense that we are the subjects who dream, and act in our dreams; a person’s identity remains constant in sleep. Yet Augustine also believes that “the moment of passing from wakefulness to sleep or back again certainly marks a great change in me” (Conf 10.30.41). Thus, while we are ourselves even in dreaming, in sleep we are ourselves with a difference.
Augustine’s initial attempt to explain the difference between his sleeping and waking self is much like Milton’s: perhaps during sleep his reason, which enables him to resist sensual suggestions made by his imagination while he is awake, is “shut down along with my eye-lids” (ibid.). If so, only Augustine’s lower faculties are involved in dreaming, and he can rest easy in the knowledge that the better part of himself is not tainted by the evil in his dreams.
Yet Augustine goes on to note that he is sometimes able to resist consenting to pleasurable sensations and images even in his dreams, which leads him to conclude that his reason is not “asleep” during sleep. The implication is that his reason is involved in Augustine’s dreams, and that his whole mind is corrupted by evil in a manner that sometimes manifests itself in his dreams. Augustine hints that he sees this implication when he goes on to pray that God “heap gift after gift upon me, that my soul may shake itself free from the sticky morass of concupiscence” so that his soul will not be divided against itself any more (Conf 10.30.42).13 The division of which Augustine speaks likely lies within his reason, as well as between reason and abilities such as imagination; even if he has only the latter in mind, that implies the fallenness of reason, since Augustine believes that it should not be so weak that it cannot guide and direct other aspects of his mind.
Yet Augustine does not pursue this insight. Although he indicates that his reason is, or at the least could be active, to some degree, while he is asleep, he seems to believe, nevertheless, that he lacks a kind of active control while he is dreaming, a control that is required for him to be blameworthy for sexual images in his dreams, as well as for the bodily erections and emissions that sometimes seem to be provoked by those images. Augustine does not explain why he believes there is such a great difference between waking thoughts and actions and those in his dreams that he is not responsible for the latter—but perhaps we can make some inferences from his conclusion that the difference between waking and sleeping is such that what happens in our dreams is not really done by us, but rather is “in some sense done in us” (Conf 10.30.41). Augustine may mean to imply, as Flanagan suggests, that dreams are “happenings not actions.”14 Things are done in our dreams—they happen to us—but what simply happens to us is not done by us, in that we do not own what happens in dreams in the robust manner required for responsibility. In effect, Augustine appears to be saying that although it is clearly he who dreams, he is passive before his dreams, not active in them. Thus, while his dreams are indeed his, they are his in a manner akin to the way in which an accidentally upset stomach is his—that is, they are not robustly his own.
So Augustine’s belief that he does not really do the evil that happens in his dreams seems to be tied to a sense that he lacks robust ownership of his dream-evil because he lacks an appropriate kind of activity in and control over it. Yet Augustine’s argument falters here, for he not only fails to say why we should think that we do not do the things done in our dreams, but contradicts the reasons to believe the contrary that he himself offers. After he concludes that dreams are done in but not by us, Augustine repeats the suggestion that the fault lies with “bestial imagination” which “drives the flesh to the point of polluting itself” (presumably, he is thinking of “nocturnal emissions”; Conf 10.30.42). Yet he should not return to simply faulting imagination, because he has already indicated that his reason is also implicated in his evil dreams.
In Augustine’s day, as in our own, many believed that one is only deeply responsible for what one is able to choose or reject. If, as seems plausible, we are unable to make such choices in our dreams, those who accept this condition should deny that we are responsible for what happens in our dreams. Yet Augustine has two reasons for rejecting this reason to think that we do not properly own our dreams. First, as we have seen, Augustine indicates that on some occasions he has refused consent to the seductions his dreams offer. It appears, therefore, that some of his dream-actions pass this test.
Second, and more importantly, Augustine’s arguments with Pelagius were started by his repeated insistence, in the Confessions, that we must confess as sin even evils from which we cannot, by our own power, refrain (but which can be resisted with the help of divine grace). Augustine therefore rejects the notion that the robust ownership required for responsibility depends on the ability to do or not do a thing on one’s own; to the contrary, his view is that we can be deeply responsible even for things, like Augustine’s own conversion, that we cannot choose on our own. Augustine seems to feel that dreams fall into this category, given his prayer in Confessions 10.30.42 that God reshape his soul so that it will no longer be titillated by sexual images in his dreams. Thus, even if Augustine cannot, by himself, control whether he consents to evil in his dreams, he should not take that as a reason to believe he cannot sin in his dreams.
In summary, Augustine’s discussion of his “dream problem” in the Confessions is internally divided: Augustine tries to reject the idea that he can sin in his dreams by arguing that his dreams simply present themselves before his mind’s eye, much like seeing happens to a person who is wide-eyed and awake. His own distinctive views, however, imply otherwise: because Augustine believes that his full mind is implicated in his dreaming, and because he believes that he can be responsible even for what he cannot, without divine help, refrain from doing, Augustine should credit his uneasy conscience with more wisdom than he wants to admit it possesses. On the account he offers in the Confessions, Augustine should believe that he owns his dreams in the robust manner required for him to sin in them.
Dreams, and dream-activity that includes what we call “nocturnal emissions,” remain a matter of interest and concern to Augustine for the rest of life. He rarely discusses these topics at length, yet he returns to them on a number of occasions, developing and revising the views expressed in the Confessions. He does so not mainly because of prurient interest, or because he blames the body for sin but for three other reasons.15 First, Augustine’s own flock appears to have wondered what to make of their dreams, especially when those dreams seemed to involve apparitions of deceased relatives or embarrassing nocturnal erections and emissions. The latter may have been a problem Augustine dealt with in the confessional, while the former was a matter that Augustine personally investigated, even to the point of interviewing people about the visions they claimed to have had (cf. Cura Mort 15; CJul 6.10; DeTrin 11.2.7).
Second, while the question of whether one can sin in one’s dreams is a very practical pastoral problem for Augustine, he also finds dreams of theoretical interest because they provide him with useful thought-experiments. The question of whether we can be responsible for what we do in, or as a result of, our dreams is a fascinating test case for Augustine, one sometimes invoked in his arguments with the Pelagians. Augustine also discusses the nature and meaning of dreams with other interlocutors, most prominently a young Donatist convert, Vincentius Victor, who argued that the content of Perpetua’s visions prove the corporeality of the soul, a view Augustine rejected.
Third, Augustine often returns to a variety of questions about sexual desire, whether in dreams or elsewhere, because he believes that topic proves with special clarity a larger point about our post-fall existence. This point will be more clear if we first summarize his mature moral psychology, according to which sin resides in both the postlapsarian “flesh” and soul (GPO 1.12.13). As this terminology is often misread, it is essential to keep in mind that Augustine is attempting to appropriate a scriptural terminology, which refers to the whole human being, often negatively, by the term “flesh” (AnOr 1.18.31). Augustine believes that the flesh cannot desire without the soul; indeed, pains and pleasures are really an experience of the soul (CD 21.3; GnLit 10.12.20). Thus, when he speaks of the desires of the flesh, he is speaking of the soul desiring in a carnal (i.e., evil) manner (PerIust 8.19). Correspondingly, he writes of “the desires of the soul which are called desires of the flesh, because the soul has carnal desires, when it has such desires that the spirit, which is its better and higher part, has to resist” (CJul 5.5.28).
In accord with his psychology, Augustine believes (unlike Julian) that human sexual desire is not merely animal or biological; it reaches and expresses our deepest inner being.16 Yet in our fallen state, this desire comes and goes illogically, and without integrating peacefully with the other powers of our minds and desires of our hearts; it seems not only to resist and distort reason but also move the body in a manner that bypasses the activity and desires with which we would like to identify ourselves (CJul 4.14.71). Thus, it violates the order that ought to exist within human persons. In fact, Augustine believes sexual desire exemplifies the disordered lives of sinners with especially painful clarity. Yet though it is a clear and poignant example, Augustine does not consider sexual desire unique, but merely illustrative. The effects of the fall are such that our flesh and spirit are in conflict; spirit does not rule over flesh as it ought—this is the point Augustine is trying to highlight in his discussions of sexual desire, and one of the issues he attempts to explore in raising his problem about dreams.
Paul Ramsey has expressed a common misconception regarding the implications of Augustine’s idea that the “spirit” should rule the “flesh,” criticizing him for deploring postlapsarian sexuality because it lacks the “personal presence” of those involved in it. He thinks Augustine holds this belief because Augustine cannot imagine any other form of personal presence in the body than a conscious and voluntary presence.17 Ramsey argues, to the contrary, that not every “personal appearance” need be or is a “command performance.” Indeed, he suggests that if sexuality is as deeply tied to human identity as he and Augustine believe, persons can be disclosed all the more in a sexuality that is not at our beck and call.18
Ramsey is right to note that Augustine’s discussion of prelapsarian male sexuality in City of God (14.23) is tied to an excursus on the amazing things even some fallen persons can command their bodies to do (CD 14.24).19 Yet he misinterprets Augustine’s mature view and his motivations for holding it. This is partially because Augustine modifies his position on prelapsarian sexuality only after he writes the portion of City of God just cited. Augustine continues to prefer the view that Adam and Eve had conscious, passionless control over all their sexual activity—because he thinks it would result in fewer unsatiated desires, and less temptation (CJul 3.14.28)—but he grants that Julian could be right in speculating that they had a sexual desire that was not opposed to the good (DNC 2.35.59–60; CJul 4.5.35, 14.69, 5.15.60). The deeper issue, however, is that Augustine is able to make this concession because he is not committed to the idea that having proper ownership of one’s desires (or actions) requires their being consciously controlled; his main concern is to assert that they should not be a source of division, perversely at odds with the good desires that Augustine calls the concupiscence of the spirit (cf. DNC 2.10.23, 2.12.25; CJul 4.13.62, 5.15.63). Thus, even when it comes to prelapsarian sexuality, Augustine is not stuck on a “command performance” view. Nor is he opposed to Ramsey’s suggestion that persons can be disclosed in their involuntary behaviors—in fact, as the discussion below indicates, he embraces that suggestion wholeheartedly.
The insight that one’s robust self can be disclosed especially clearly in behaviors concerning which one does not rationally deliberate, or make voluntary choices, is central to Augustine’s anti-Pelagian philosophical soteriology in general, and his treatment of the problem of dreams, in particular. Evil dreams can, Augustine suspects, be one example of “personal presence” without “command performance”—but to understand why he thinks so, we must first understand, at least in broad outline, Augustine’s mature theory of deep responsibility. Considering the basis on which Augustine defends our being deeply responsible for waking thoughts and actions will make it easier to understand and evaluate, by comparison, his discussions of dreams. Only a brief overview of this complex topic is possible here—a summary that not only ignores dissenting views but inevitably makes Augustine’s views seem more coherent than they are in his own words—but it should not be hard to see the ways in which his mature response to his dream problem connects to his debates with Pelagius and Julian.
First, we should remind ourselves that, for Augustine, what counts in assessing a person is that person’s motivations, beliefs, and other states of mind. For him, the central questions about sin are “what does one love?” and “towards what is one oriented?” He makes this explicit when he tells Julian that if one desires carnally, even without consent of the mind or action of the bodily members, such concupiscence of the flesh is evil in itself (OpImp 5.50). Moreover, evil beliefs and desires deserve blame even if they are not put into action (CG 16.4; CJul 5.7.28), and God judges us for what we would prefer to do, if we could get away with it (SL 8.13; CD 1.18). Thus, we are judged in our waking lives not primarily for our physical actions but for our states of mind, our mental activities. I will call this view Augustine’s “mentalism.”
Second, Augustine believes that we are responsible for what we will, because we cannot deny that we own the things we will. More precisely, Augustine ties praise and blame to ownership, and he appears to believe that if something is part of or originates in what he calls the voluntas, that is a sufficient condition for responsibility. Unlike one who is coerced by threats, the sinner who acts from his or her own voluntas is not under compulsion by some external power; rather, the power that moves him is his or her own disordered desires. So Augustine writes:
That person is, however, very much mistaken who, while consenting to the concupiscence of the flesh and definitely deciding to do what it desires, still supposes it is right to say, It is not I who do it. After all, a person consents, even if one hates the fact. For these two can coexist in one person: both the hating it because one knows it is evil and the doing it because one decided to do it. (DNC 1.28.31; cf. Retr 1.23.1-2)
Augustine goes on to say that even if those who consent to sin are displeased with their behavior, they must nevertheless see themselves in their consent.
In order to understand the implications of these claims, it may be helpful to turn to Augustine’s early Commentary on the Sermon on the Mount, written a few years prior to his Confessions, where Augustine argues that “there are three things which go to complete sin: the suggestion of, the taking pleasure in, and the consenting to” (SermInMont, 1.12.34).20 He expands on the meaning of these terms in the following manner:
Suggestion takes place either by means of memory, or by means of the bodily senses, when we see, or hear, or smell, or taste, or touch anything. And if it give us pleasure to enjoy this, this pleasure, if illicit, must be restrained. Just as when we are fasting, and on seeing food the appetite of the palate is stirred up, this does not happen without pleasure; but we do not consent to this liking, and we repress it by the right of reason, which has the supremacy. But if consent shall take place, the sin will be complete, known to God in our heart, although it may not become known to men by deed. (ibid.)
This tripartite rubric continues to be important to Augustine over time, and he often invokes it both explicitly and implicitly.21
As Augustine’s views mature, he comes to believe that what comes to mind—what is suggested to us—is something over which we lack direct control, even while we are awake (DDP 8.20). Moreover, what we find pleasurable or delightful is no more voluntarily controlled by us than what is on our minds. As we become adults, and our minds develop, we find that we take delight in some things, and not others. One might hope that we can control these delights by the power of choice, but Augustine increasingly emphasizes that our choices actually depend on what motivates us, and that our motivations are provided by the very delights we have just noted the difficulty of controlling. Thus, it is often impossible to voluntarily control what we consent to. The fact that we often cannot do otherwise than consent to what seems pleasurable to us is one reason why Augustine concludes we need the assistance provided by divine grace.
Augustine develops a technical terminology for talking about this situation by distinguishing what we might consider the will (properly so-called) from an act of the will, that of choice or consent. The former Augustine names the voluntas, a term that implies for him what we call wishes, desires, and attachments—though we must keep in mind that Augustine thinks of these goods as deeply rational, in the sense that desiring a thing implies believing that it is good, and while desires can be foolish, they are never without their reasons. The latter Augustine speaks of as the liberum arbitrium, the ability of the voluntas to endorse and identify with that which the voluntas seeks.22 The fact that Augustine thinks of liberum arbitrium (free choice) as being “of the voluntas (will)” is significant, because it implies that that to which a person consents is tied to, and depends upon, what a person desires.
In Augustine’s mature thought, “liberum arbitrium” need not imply a power for alternatives. Many of Augustine’s modern interpreters have either denied or overlooked this, fearing perhaps that accepting it makes Augustine a “compatibilist” who believes responsibility and necessity are compatible. I believe that the late Augustine is fairly obviously a compatibilist, and that that is a good thing, but all I can offer here in defense of those claims is the following: Augustine gives a fair indication of how he conceives of liberum arbitrium when he writes that “If we look for the free choice of a human being that is inborn and absolutely unable to be lost, it is that by which all will to be happy, even those who do not will those things which lead to happiness” (OpImp 6.11).23 In this passage, Augustine makes it clear that he considers the paradigm example of liberum arbitrium the natural affirmation of the desire for happiness, a consent we find unavoidable. This is not a choice that involves alternative possibilities, because Augustine believes that we cannot desire to be unhappy; we seek happiness by natural necessity (OpImp 4.92–93).
In summary, what we might as well call Augustine’s “compatibilism” is his belief that what is required for deep responsibility is a kind of ownership, one that does not depend on an ability to choose between alternative possibilities, but is signified by whether a person has a thing willingly—whether a person desires and consents to that thing with his or her own voluntas and liberum arbitrium. Thus, merely bodily states or actions are not owned in the manner necessary for deep responsibility, and being physically beautiful or having leprosy are (under ordinary circumstances) not things for which persons should be considered responsible. Expressions of one’s loves, beliefs, and commitments, however, can very well be owned in the manner necessary for responsibility, even when they are not voluntarily chosen or expressed.24 Augustine draws on St. Paul to defend his views:
Who ever said . . . that anyone else commits anyone’s sin? For even he who says “It is no longer I who do it” immediately adds “I know that the good does not dwell in me.” He shows that whatever that it is it is his own, because the flesh belongs to the one who is composed of flesh and spirit. (OpImp 2.15)
Augustine draws on the ideas summarized above as he thinks about his dream problem, not only because the theory of responsibility just sketched provides the context for his thinking about sin in dreams, but because Augustine believes that what happens in dreams is often much like what happens when a person who is awake responds to sensory experiences.
On Augustine’s view, the suggestions offered in dreams usually occur when a sleeper’s memory presents images to the “mind’s eye.” This is not always what happens in dreams, however, and that fact allows Augustine to distinguish between kinds of nocturnal narrative experiences. He differentiates, in particular, between dreams and visions. The latter differ from dreams in that they are presented to the mind’s eye not by the mind, via memory, but by the operation of God, in a manner much like the way objects are presented to waking vision (AnOr 4.18.26, 21.34; Cura Mort 15; DeTrin 11.2.7; Ep 159, 162). Such experiences do not violate human proper function, though Augustine argues that they are more rare than many of his flock seem to have thought. Augustine also holds that, waking or sleeping, persons can have delusions, visions improperly produced in the mind by diseases, drugs, or even demons (CG 22.22).25
In any of these cases, the subject of the dream-experience is not someone or something other than oneself. Augustine also rejects the idea that only part of the soul is active in one’s dream-self—as his psychology, summarized above, implies.26 These views are most clearly delineated in On the Nature and Origin of the Soul, where Augustine treats a number of questions about the nature and reality of dream-pains and pleasures. Among other things, Augustine argues that these feelings are real, though the experiences had in dreams are not always what they may initially seem to be. Thus, he writes that “even in dreams, when we endure harsh and troublesome experiences, it is certainly we ourselves who do so” (AnOr 4.17.25, cf. 4.18.27). One really does have pleasures and pains, and even consent to them (even if one does not choose to), in dreams. This means that the late Augustine admits what the Augustine of the Confessions did not: one really does act, mentally, in one’s dreams.
In his commentary on Paradise Lost, C. S. Lewis appears to agree with Milton that persons are not responsible for what they dream, or what they do in their dreams; his succinct defense of Milton is the comment that “common sense tells us that we no more become bad by thinking of badness than we become triangular by thinking of triangles.”27 Augustine appears to find Lewis’s suggestion partially right: he indicates that “there is no harm in . . . imaginative fancies, just as there is no harm in experiencing sensory things and retaining them in the memory, provided you do not desire them covetously” (DeTrin 11.8, cf. GnLit 12.14.30-31). So Augustine agrees with Lewis that simply seeing things in one’s mind’s eye is not necessarily blameworthy or harmful; this includes even the bad things that Augustine himself at times imagined, without bad motives, in his writing.28
However, Augustine also distinguishes between thinking a person blameworthy and thinking that person responsible. Thus, while he agrees that there are times when it is not blameworthy to think of bad things—in a fallen world, one might have a good reason to think about an evil thing—he is committed to believing that we are often responsible for our thought-lives, waking and sleeping. In sleep, Augustine thinks humans are often personally present in their dreams, and that one may learn something about oneself from the content of one’s dreams because of what we might call the psychological reality of dream-experiences.
In light of the distinctions Augustine makes between kinds of dreams, it is important to note that he does not think we mentally act in all of our dream-experiences: only those dreams that are not visions or delusions are properly said to be produced by one’s own mind, because “it makes all the difference whether on the one hand the senses of the body are lulled in sleep, or shaken from their inner moorings by madness, or otherwise alienated in divination or prophecy” (DeTrin 11.2.67). In a vision, one’s mind is at least partially taken over by a person other than oneself; one might be able to control one’s dream-response to the content given in a vision, but much depends on what sort of vision one has. One’s actions in a delusion cannot be trusted to be one’s own, because the mind is too incapacitated for that to be the case. What we have seen so far, then, is that Augustine is confident that our own minds are properly the authors of our dream-actions and experiences only when we are dreaming, as opposed to experiencing a vision or delusion. Many visions, and all delusions, can plausibly be said to fit the claim Augustine made in the Confessions about all dreams: they happen to us, but are not done by us—not in the sense required for them to be attributed to us, or for us to be praised or blamed for them.
Augustine’s distinction between robustly owned dreams authored by oneself and visions and delusions makes it possible for him to agree with Milton’s Adam in thinking that Eve is not responsible or blameworthy for her prelapsarian vision. His reasoning, however, differs from Milton’s, because Augustine does not believe that Eve’s dream is typical. Rather, it is a special case, in that the visions in her mind are forced on her by Satan. According to Augustine’s moral psychology, many dreams are not forced on us, but rather produced by our own minds, functioning under their own impulse. We are personally active in such dreams, and the question of responsibility for them remains, along with the possibility of sinning in them.29
In dealing with this question, Augustine does not always make explicit use of his mentalism and compatibilism, but we have already seen the impact of these views in his suggestion that the kinds of dreams in which a person might sin are those produced by that person’s mind when it is not under external influences including drugs, disease, or another person. We have also seen that while one’s experiences and actions in such dreams differ from one’s waking sensory experiences and actions, Augustine does not indicate that they differ in any respect that would make it impossible to sin in the former. While we lack the ability to choose what we dream, and only rarely can choose what we do in our dreams, Augustine believes that a similar story can be told about most of our waking mental lives. Though we do not typically choose to find them attractive or annoying, the people and things around us often provoke thoughts and desires that enter our heads unbidden, and we may or may not be able to suppress further reactions related to such perceptions. Given this analysis, the fact that Augustine holds persons responsible for much of their waking mental lives presents a prima facie reason for him to hold human persons responsible for the content of their dreams.
Augustine seems to hold persons responsible for their dreams in two ways. First, when it comes to the actions of one’s dream-self, that to which one’s dream-self consents can be attributed to a person as a real disclosure of who that person is. While dream-actions do not take place in the physical world—Augustine makes it clear that if a person is baptized in a dream, that is not at all the same as actually being baptized—if one chooses to be baptized in a dream, that can be a psychologically real action of one’s heart (and one should indeed be baptized! Cura Mort 15). In fact, Augustine believes that Scripture itself teaches in the story of King Solomon’s vision that we can be responsible for our dream-actions (1 Kings 3:5–15). When God offers to give Solomon whatever he most desires, and he answers that he desires wisdom, this expression of Solomon’s desires was not necessarily a conscious choice, but it was nevertheless an action of his mind in which he was personally present. Augustine claims, accordingly, that “some . . . merits shine out even in dreams” and that Solomon’s wish “found favor with the Lord, who was not slow to give him a good reward for a good desire” (GnLit 12.15.31). Augustine’s suggestion is that we can be responsible for our dream-actions because we can be personally present in them in a way that makes those actions self-expressive. And if persons can be praised and even rewarded for their self-expressive dream-actions, it follows that persons can be responsible for their dream-actions, and that persons can also be blamed (and perhaps even punished?) for their dream-actions, as well.
Second, Augustine extends this logic to cover not just one’s dream-actions but the content of one’s dreams themselves. Whether or not one’s dream-self consents to various actions in one’s dreams, it appears to mean something to Augustine that one’s dreams have the specific narrative content and images that they have. Insofar as one’s own mind is the author of one’s dreams, the narratives one constructs say something about who one is; that is why it makes some sense for Augustine to feel guilty about the content of his evil dreams, and pleased insofar as the evil content in his dreams diminishes. We have evil dreams, Augustine believes, because our wills are corrupt: “if in paradise, where there was no evil of concupiscence, there was the cycle of waking and falling asleep, their dreams while they were asleep were as peaceful as their life while they were awake” (CJul 5.10.42). Misplaced loves express themselves in both the waking and sleeping lives of sinners. Thus, the content of our dreams reflect our character, and it does so in the active sense that one’s beliefs and desires influence one’s dreams, and are disclosed in one’s dreams. Dreaming is, of course, quite different from authoring a novel, or deliberately making up a story, but one can nevertheless be personally present in the content of one’s dreams, when dreaming expresses one’s hopes, fears, or loves.30
A few passages in Augustine’s later works have led some scholars to believe that his mature view of dreams is more like his view in the Confessions than I am suggesting.31 However, attention to these texts shows that Augustine’s position is more subtle than most of his readers have recognized. In Augustine’s Literal Commentary on Genesis Augustine argues (1) that
When the mental pictures that occur in his thoughts while he is speaking [in a sermon] are also imprinted on his vision while he is dreaming in such a way that he cannot distinguish between them and the real coupling of bodies, the flesh is at once stirred into movement, and the result is what usually follows upon this movement; and this happens without sin . . . (GnLit 12.15.31)
In City of God, in the midst of an argument that those who are about to be raped should not commit suicide to avoid being implicated in sin, Augustine indicates (2) that
that lustful disobedience which still dwells in our dying members sometimes moves itself as if by its own law, apart from the law of our wills: when we are asleep, for instance. In this case also, however, there is still no guilt in the body of one who does not consent. (CG 1.25)
In an argument with Julian, however, Augustine suggests (3) that concupiscence can steal consent even from the chaste, by deceiving the sleeping senses (CJul 4.2.10, cf. 3.20.38). He later adds (4) that “when sleep holds our members in its grasp, it does not make them disobey the will, because it takes from the will its power of command by distracting the soul with visions in dreams . . .” (CJul 5.10.42).
Passages (1) and (4) have been taken to mean that dreams cannot be an occasion of sin because they lack the voluntary choice many believe to be required for responsibility. Yet we have already seen Augustine challenge such ideas about responsibility; he believes we can consent without choice. Moreover, in (3), written not too long before the other passages, Augustine clearly indicates that there can be a kind of consent in nocturnal emissions. Perhaps Augustine is simply being inconsistent, but I propose that we can make sense of these passages once we note that they all focus on the question of what to make of nocturnal emissions, which are not themselves dreams or actions in dreams, but bodily movements that occur in response to dream-images. Thus, the question in which Augustine is interested in these passages is not whether we can sin in our dreams, but whether nocturnal emissions are sinful.
Augustine’s answer to the latter question is that because the will is not properly involved in nocturnal emissions, we are not responsible for them. As he indicates in (1) and (2), our members—our bodies—can be moved by instinct, rather than via the will. This happens when we are awake, too, of course: we have already seen Augustine note the hunger pangs and salivation those who are fasting must endure when they encounter the sight and smell of food (SermInMont, 1.12.34). Such unwilled “consent” can also happen, Augustine is concluding, in sleep: the consent that is stolen while we are sleeping is not the consent of the will, which cannot command the body in sleep, but merely the consent of the body, reacting to a stimulus. Augustine notes that the will loses command of one’s limbs while one is sleeping, and just as a person’s limbs may nevertheless move in sleep, without regard for what one desires, they may also respond sexually, but without sin, because one’s personal presence is bypassed insofar as the voluntas and liberum arbitrium are.
Thus, Augustine considers emissions a sign of the fallenness of the body much like other disorders of the body that do not engage the will, including death. Augustine does not clarify how it is possible for the will to be bypassed in the action of a sleeping person’s bodily members. Nor do these passages focus on assessing the sexual images that provoke nocturnal emissions; Augustine’s attention is on asserting the point that the latter, as bodily actions that bypass the voluntas, should not be considered sinful.
At the same time, there are signs that Augustine is divided in his thinking about some of the dream-images produced by his memory. As we see in (1), he sometimes speaks of sexual images in dreams as merely left over in the memory from a day’s work that included writing or speaking about sexuality. This suggests that Augustine sometimes thinks of dream-images as a kind of “mental junk,” dissociated fragments of his waking thoughts, now meandering meaninglessly through his brain.
On balance, however, he cannot accept that all dream-images are debris in which he is not personally present, because he believes that the minds of those who were fully pure would not produce disturbing images in their sleep—a prelapsarian Eve could not have had evil dreams of her own. Sinners, by contrast, find disordered loves and beliefs within themselves even when they have made up their minds to seek what is best. Thus, it is not surprising that the secret depths of one’s heart can be disclosed in one’s dream-life—when, for instance, a person takes pleasure in narratives that may be the product of desires on which that person fears to act (cf. Ser 161.8).
Augustine’s mature response to his dream problem is incomplete in certain ways. Perhaps his most important oversight is that he says nothing about the degree of responsibility persons have for their dream-desires, pleasures, and actions—surely an evil desire indulged only in one’s dreams is less bad than one indulged in one’s waking life? He also leaves us wondering about the status of other actions persons might perform in their sleep—if you are angry at your spouse, have a dream about hitting him or her, and wake to find that you have actually thrown a punch across the bed, how should we think about that?32 Finally, Augustine leaves us with questions about a number of boundary issues: he does not make it easy to know precisely what a person is responsible for in his or her dreams, or when; he does not indicate exactly how to differentiate between dreams, visions, and delusions; and he fails to clarify the extent to which we should think of dream-images as “mental junk” as opposed to narratives in which we can be personally present. Yet his answers to the questions he first posed in the Confessions are a fascinating example of and window into his mature views about sin and responsibility. They are also provocative, in that they challenge us to engage Augustine’s dream problem for ourselves.
As we consider the merits of Augustine’s views, it is important to recognize that recent dream-research appears to support key elements of Augustine’s view of dreams. Freud, of course, thought of dreams as subconscious wishes in need of interpretation, because their true meaning is hidden behind bizarre mental images. Dream-research suggests that while Freud was right in thinking that dreams are often much more than a merely meaningless by-product of brain activity, he was wrong in thinking that dreams are an unusual form of mentation that tends towards the bizarre.33 As William Domhoff has argued, the popular imagination has exaggerated the degree to which a person’s dreams are disconnected from that person’s everyday life, or differ from waking mentation.34 In part, the problem is that popular ideas about dreams rely heavily on self-reporting that tends to be skewed to emphasize the unusual and disturbing.
More rigorous studies in which large numbers of subjects have been awakened and interviewed about their dreams suggest that dreaming is a complex cognitive achievement, one that children—whose dreams tend to be more static than dynamic until they are over ten years of age—grow into only gradually, as their ability to engage and represent the world around them grows. In adults, Domhoff concludes that “dreams dramatize conceptions and concerns, and . . . are generally consistent over time and continuous with waking thoughts.”35 Some dreams are indeed inexplicable, but such dreams tend to occur when the conceptual abilities of the dreamer are under strain, such as in times of illness or during transitions from waking to sleeping. In general, “comparison of dream content with waking life suggests that dreams express our conceptions of the people and activities that concern us in waking life.”36 Thus, while some dreams may be mental junk, and even though dreams are typically not a command performance, we are often personally present in the content of our dreams. Indeed, whether a person’s dreams disclose that person’s personality has little to do with whether a person can voluntarily, consciously direct his or her dreams. And that makes Augustine’s notion that we can involuntarily disclose who we are in our dreams seem quite plausible.
The main barriers to agreeing with Augustine’s solution to his dream problem are, therefore, more theoretical than empirical. It may be that those who accept Augustine’s “mentalism” are likely to find his compatibilism attractive as well. The two are connected by the conviction that certain states of the heart and mind are intrinsically blameworthy and sinful. Indeed, Augustine believes that the individual sins we knowingly choose to enact are less significant than the states of our hearts, because the former are less new and independent sins than external signs of the continuance of the more fundamental sins that already exist inside us. It is congruent with this understanding of the nature of sin to believe that our dreams, like our other, more overt actions, can be a barometer of our hearts. Given these convictions, Augustine dreams of a day when our secret mental lives will, by the grace of God, be more responsible than they are in this life.
1. John Milton, Paradise Lost, ed. Barbara K. Lewalski (New York: Blackwell, 2007), 5.51–2. This and the following passages were brought to my attention by the discussion in Brian Horne, “Human Sin and Human Freedom: A Reading of Milton’s Areopagitica,” in God and Freedom: Essays in Historical and Systematic Theology, ed. Colin E. Gunton (Edinburgh: T & T Clark, 1995), 21.
2. Paradise Lost, 5.84–6.
3. Paradise Lost, 5.99–100, 117–121.
4. Paradise Lost, 4.800f; 5.101-116.
5. C. S. Lewis, A Preface to Paradise Lost (New York: Oxford University Press, 1960), 84.
6. It is worth noting that while the latter use of the term “responsible” seems to invoke moral concerns, the former need not, because we can be deeply responsible for non-moral aesthetic creations or other actions that are not necessarily a matter of moral concern. Thus, whether what one does in one’s dreams is immoral or not, one could be deeply responsible for them.
7. The first two quotations are from F. Snyder, “The Phenomenology of Dreaming,” in The Psychodynamic Implications of the Physiological Studies on Dreams, ed. L. Madow and L. Snow (Springfield, IL: Thomas, 1970), 129, quoted in G. W. Domhoff, “Realistic simulation and bizarreness in dream content: Past findings and suggestions for future research,” in The New Science of Dreaming: Content, Recall, and Personality Characteristics, vol. 2, ed. D. Barrett and P. McNamara (Westport, CT: Praeger Press, 2007); the final quotation is from ibid. The articles by Domhoff cited in this essay were accessed online at http://dreamresearch.net/Library/domhoff_2007b.html, http://dreamresearch.net/Library/domhoff_meyer_schredl_2006.html, and http://dreamresearch.net/Library/domhoff_2007a.html, and page numbers are not available.
8. See Phillip Cary, Inner Grace: Augustine in the Traditions of Plato and Paul (New York: Oxford University Press, 2008), 86–100.
9. Owen Flanagan, “Self-Expression in Sleep: Neuroscience and Dreams,” in Self-Expressions: Mind, Morals, and the Meaning of Life (New York: Oxford University Press, 1996), 33.
10. References to Augustine’s works are by book, if any, and chapter and paragraph, in Arabic numerals—not to page numbers. I refer to his texts by English titles in the body of my paper, but by Latin abbreviations in parenthetical citations. See the note at the end of this book for a list of titles and abbreviations, and for full citation information for Augustine’s works.
11. Gareth B. Matthews, “On Being Immoral in a Dream,” Philosophy 56 (1982), 53.
12. Cf. Stephen Laberge, Exploring the World of Lucid Dreaming (New York: Ballantine Books, 1991).
13. Augustine sometimes means to imply sexual desire when he speaks of concupiscence (this may be especially the case in the Confessions), but in general what he has in mind when he speaks of concupiscence is simply desire in general; see Mathijs Lamberigts, “A Critical Evaluation of Critiques of Augustine’s View of Sexuality,” in Augustine and His Critics: Essays in Honor of Gerald Bonner, ed. R. Dodaro and G. Lawless (New York: Routledge, 2000), 179–80 and Margaret Miles, Augustine On the Body (Missoula, MT: Scholars Press, 2000), 67–69. Concupiscence can be good or bad, though Augustine uses the term in a negative sense more often than a positive sense, particularly prior to his debates with the Pelagians; cf. Jesse Couenhoven, “Augustine’s Rejection of the Free Will Defence: An Overview of the late Augustine’s Theodicy,” Religious Studies 2007 (43): 279-298.
14. Flanagan, “Self-Expression in Sleep,” 33.
15. See Elizabeth Clark, “Vitiated Seeds and Holy Vessels: Augustine’s Manichean Past,” in Ascetic Piety and Women’s Faith (Lewiston, NY: Edwin Mellon Press, 1986); Elaine Pagels, Adam, Eve, and the Serpent (New York: Random House, 1988), ch. 6.
16. See Peter Brown, “Sexuality and Society in the Fifth Century A.D.: Augustine and Julian of Eclanum,” in Tria Corda: Scritti in Onore di Arnaldo Momigliano, ed. E. Gabba (Como: Edizioni New Press, 1983) 49–70, for discussion of how radical Augustine’s view would have seemed in his day.
17. Paul Ramsey, “Human Sexuality in the History of Redemption,” Journal of Religious Ethics 16, no. 1 (1988), 60.
18. Ramsey, “Human Sexuality,” 62.
19. My thanks to John Bowlin for reminding me of this point.
20. This passage was drawn to my attention by Matthews, “Immoral in a Dream.”
21. However, the manner in which he understands especially “consent” may change over time; cf. Eugene TeSelle, “Serpent, Eve, and Adam: Augustine and the Exegetical Tradition,” in Augustine: Presbyter Factus Sum, ed. Joseph T. Lienhardt, S. J., Earl C. Muller, S.J., and Roland J. Teske, S. J. (New York: Peter Lang, 1994); Couenhoven, “Augustine’s Rejection of the Free Will Defence.”
22. See N. W. Den Bok, “Freedom of the Will: A Systematic and Biographical Sounding of Augustine’s Thoughts on Human Willing,” Augustiniana 44 (1994): 237–70.
23. For further development of my views, see “Augustine’s Rejection of the Free Will Defence,” and “What Sin Is: A Differential Analysis,” Modern Theology, forthcoming in October 2009.
24. I am assuming that “voluntary” implies not just action that is not compelled but action done with knowledge of what one is doing.
25. See also Richard Sorabji, Emotion and Peace of Mind: From Stoic Agitation to Christian Temptation (New York: Oxford University Press, 2000), 414.
26. Cf. Matthews, “Immoral in a Dream,” 96.
27. Lewis, Preface, 84.
28. Cf. Gerard O’Daly, Augustine’s Philosophy of Mind (London: Gerald Duckworth & Co. Ltd., 1987), 117.
29. If we dislike our dreams but do not consider ourselves responsible for them, we might regret our dreams without considering ourselves guilty for them; John Rist considers this Augustine’s view in Augustine: Ancient Thought Baptized (New York: Cambridge University Press, 1994), 138, although he wonders why Augustine does not make more of the role of concupiscence in dreams in ibid. 86. As we will see, I believe he misreads Augustine; he does so, it appears, because he believes, as Augustine does not, that we are responsible only for what we voluntarily choose.
30. For a somewhat different take on the idea of dreams as narratives, see William Mann, “Dreams of Immorality,” Philosophy 58 (1983): 382. Unlike Augustine, Mann is exploring the possibility that dreams may not be experiences; even if so, he argues, we can nevertheless be responsible for our dreams, insofar as we discover our dreams within ourselves as narratives for which we have editorial responsibility.
31. See, e.g., O’Daly Augustine’s Philosophy, 117; Rist, Augustine, 86 n82; and Sorabji, Emotion, 415.
32. My thanks to my wife, Amy Tsou, for this example, and for her comments on the rest of this paper.
33. Cf. Joel Feinberg, “Collective Responsibility,” in Doing and Deserving (Princeton, NJ: Princeton University Press, 1970), 250–1.
34. G. W. Domhoff, Meyer-Gomes, K., and Schredl, M., “Dreams as the Expression of Conceptions and Concerns: A Comparison of German and American College Students,” Imagination, Cognition and Personality 23, no. 3 (2006): 269-82; G. W. Domhoff, “Dreaming as the Embodiment of Thoughts: A Widower’s Dreams of his Deceased Wife,” Paper presented to the annual meeting of the International Association for the Study of Dreams (Rohnert Park, CA, 2007); G. W. Domhoff, “Realistic Simulation.”
35. G. W. Domhoff, “Realistic Simulation.”
36. Domhoff et. al. “Dreams as the expression of conceptions”; cf. Domhoff, Dreaming. Other conceptions of dreams can also be compatible with this conclusion; cf. Owen Flanagan, “Self-Expression in Sleep: Neuroscience and Dreams,” in Self-Expressions: Mind, Morals, and the Meaning of Life (New York: Oxford University Press, 1996), 32–52 and Mann, “Dreams of Immortality.”