1. The Problem, with the Solutions Currently Available

Most of the discussion about Chaucer's possible knowledge of the City of God has centered on what appears to be a direct reference to the book made near the start of the story of Lucretia in the Legend of Good Women. “The grete Austyn hath gret compassioun / Of this Lucresse, that starf at Rome toun,” Chaucer writes (1690–91); and as Augustine does not appear to comment on the story at any length elsewhere, it would seem that Chaucer has the prominently placed discussion in the City of God (book 1, chapter 19) in mind.Footnote ² There is, however, a difficulty: the treatment Lucretia gets there does not at first glance seem especially rife with compassioun. For one thing, part of its conclusion is that she has reacted badly to the awful events that have befallen her: Augustine argues at length that, barring the exceptional and hard-to-discern case of a direct command from God, suicide is always morally wrong. Still worse, in the course of the discussion he traps Lucretia in a piece of dilemmatic logic that has struck many readers as unkind at best, and at worst downright cruel. If Lucretia was completely innocent of any desire for the assault of the rapist Tarquin, Augustine reasons, then she has committed a different crime: that of putting to death an innocent person, namely herself. If she has not put to death an innocent person, then — since for Augustine there is no way to incur guilt by being purely on the receiving end of someone else's action — she must have harbored some hidden consent to the rape after all. In either case, the implication seems to be, she is not quite worthy of the unreserved praise that the Romans lavish upon her. And as if all these forms of disapproval were not enough, Augustine ends his consideration by saying that not only was the deed of suicide objectively unjust, but its motives were less noble than is sometimes supposed. The passage is frequently quoted:

According to Augustine it is not regret over lost chastity, but a bad reaction to the fear of shame, rooted in a yearning for praise endemic to Roman women (or wives), that has led to Lucretia's drastic action. One begins to see why so many recent writers have thought that an author working under the sway of gret compassioun, particularly in the etymological sense of a capacity to feel-and-suffer-with this victim of acquaintance rape, might say something a bit different.

Nearly everyone who comments on the “Legend of Lucrece,” in fact, attempts to explain the apparent discrepancy between Augustine's text and the two-word review Chaucer gives it. Ninety years ago J. S. P. Tatlock offered to dispatch the problem quickly, if wryly, by crediting it to a slip in the poet's memory.Footnote ⁴ Writers of the last few decades, on the other hand, have tended rather to believe that Chaucer's memory never had a chance to prove its mettle in this particular arena, because it never met Augustine's text in the first place: Chaucer likely drew instead on some intermediate source, they reason, whose abridgments or other changes gave the poet his mistaken idea about the compassioun of the original. Among proposed intermediaries the version of Lucretia's story in the Gesta Romanorum, which claims as its own source the City of God rather than the vastly more developed accounts in Livy and Ovid, turns up often; the late-fourteenth-century translation of Valerius Maximus into French by Simon de Hesdin, which supplements Valerius's short account with long passages from Livy and also briefly describes Augustine's treatment, is another candidate.Footnote ⁵ A second set of writers has been willing to credit to Chaucer an encounter with the original text but has suggested that deliberate irony is at play, whether in Chaucer's own voice or, in the case of at least one proposal, as a result of Chaucer's creation of an implied narrator who “ignores the substance of Augustine's lengthy commentary.”Footnote ⁶ Virtually all responses to Chaucer's text currently in circulation attempt to solve the problem by resort either to the hypothesis of an intermediary or to that of irony.

Neither route, however, promises a particularly smooth journey. The reasons for not adopting the second hypothesis are best postponed for the moment, as the detailed look at Augustine's text that follows will bring them to light without additional effort. As for the notion of an intermediate source, there is nothing implausible about the idea itself; but trouble arises quickly on closer examination, because the hypotheses taken from this group generally do not explain the “data,” the ascription of compassioun to Augustine, any better than does the hypothesis that Chaucer read the City of God directly. In the case of Simon de Hesdin, in fact, the opposite is true, as his single-sentence summary of Augustine's judgment seems considerably harsher than the original. “Saint Augustine speaks of this Lucresse,” he writes, “and says in the first book of the City of God, in the nineteenth chapter, that she killed herself more because of desire for praise than because of love of chastity, and that she preferred that it would become clear that she had been violated against her will rather than [preferring] that she would be defamed [viz., according to Tarquin's final threat, discussed below] without being violated.”Footnote ⁷ The attempt to trace Chaucer's comment to the Gesta Romanorum is at first glance more viable, in part because that text, instead of summarizing Augustine's judgments in the course of a longer retelling, simply attributes the entire story to Augustine and does not mention his judgments, pro or con, at all. The bare story that it then relates, lacking Augustine's observations about Lucretia's imperfect motives and also his eagerness to show the general sinfulness of suicide, does leave the reader with a more favorable impression than does the end of Augustine's nineteenth chapter. But Lucretia comes across in the Gesta not as an object of pity but as a noble heroine: a victim, to be sure, but one of impressively, if not frighteningly, stalwart fortitude. She is, moreover, in that text an allegorical Everysoul — explicitly so in some manuscripts, which append to the story a short exposition of its alleged hidden meaning. As one might expect in a retelling aimed at allegory, there is not a touch of pathos in the very brief account of her death. For all these reasons, even if a reader believed that the Gesta had given him or her an accurate view of Augustine's treatment of Lucretia, compassioun would be a peculiar word for describing the impression it makes.

A third, more intricate version of the intermediate-source idea deserves a somewhat lengthier look: this is Andrew Galloway's proposal to trace Chaucer's impression of Augustine-on-Lucretia to a group of fourteenth-century Englishmen, perhaps as many as five of them, who wrote commentaries on the City of God or passed along or translated the commentaries of others.Footnote ⁸ The first of these is Nicholas Trevet, a writer of whom Chaucer has demonstrable knowledge, as another of Trevet's works underlies most of the Man of Law's Tale. Trevet's full-length commentary then influenced the later efforts of his fellow friars Thomas Waleys and John Ridevall, and Ridevall's at least was a strong influence on Ralph Hidgen's universal history or Polychronicon, which often borrows directly from it — a fact that Galloway seems to have been the first to register. The Polychronicon was apparently among the fourteenth century's most widely disseminated texts, and was translated into Middle English by John Trevisa in 1387, making it possible that Chaucer read the translation before writing his “Legend of Lucrece” (though the timing, for which see note 5, would be tight). Galloway's article is not in the first instance concerned to explain Chaucer's perception of a compassionate Augustine; it is after a broader point of intellectual history, arguing that an appreciation of an ancient Roman ethics of fame and shame grew gradually across the fourteenth century in England just as it had in Italy, rather than appearing suddenly and with little English precedent, as some have argued, in Chaucer.Footnote ⁹ Nonetheless his article does along the way propose that the remark about gret compassioun — which he labels a “peculiarly inept characterization” of the material in the City of God — is “explicable if Chaucer did not bother to go beyond the version of Augustine's account of Lucretia” to be found in this group of English writers, particularly Ridevall and Higden.Footnote ¹⁰

The claim is noteworthy and the article useful, both in drawing attention to the often overlooked material in these writers and in raising questions about the differences between a Roman cultural “ideology” based on shame and a Christian one more concerned with guilt. Nonetheless it must be said that some of the essay's claims, including this one, find themselves seriously undercut by problems with its representation of what Augustine actually wrote. Galloway believes that the five medieval Englishmen were more attuned to the differences between their culture and that of prerepublican Rome, and more impressed with the latter, than Augustine had been; and in arguing the point he says that the five made as they passed along Augustine's work a number of innovations that might lead a reader who did not know Augustine directly to believe that the fifth-century bishop had been as much of a “sympathetic historicist” as they were.Footnote ¹¹ The trouble is that many of the instances offered are not innovations at all, but the straightforward transmission of a plain-sense reading of Augustine. One example directly relevant to this present essay appears in the article's longest direct quotation, taken from Trevisa's translation of Higden:

The initial R, spelled out to Ranulphus in Higden's Latin, marks a transition between Hidgen's direct borrowing or paraphrase from an authoritative source and his own comment (though the latter may itself derive from another source, as Galloway has discovered that it does here; Higden is quoting Ridevall's City of God commentary). In this case the preceding paraphrase claims Livy and Augustine as its sources, though what Higden actually transmits there is overwhelmingly if not entirely from Livy. According to Galloway, R-marked comments like this one often serve Higden as opportunities to judge among or even correct the opinions of the preceding authorities; and in this particular case, he believes that that work begins with the second sentence. “The second ‘Bote,’” he writes, “ … marks [Higden's] characteristic oppositional posture.”Footnote ¹³ That claim fits the article's general theme: Galloway takes the passage's second sentence as evidence that fourteenth-century historicists like Ridevall and Higden soften Augustine's harsh conclusions as a result of their own greater willingness to take sympathetic account of the differences between ancient Roman culture and their own.

In fact, however, this passage contains no softening of, and no opposition to, its ancient source. Higden's (and Trevisa's) second sentence puts forward nothing beyond the ideas we have already seen in Augustine himself, and with diction sufficiently close to Augustine's to suggest that Hidgen, so far from meaning to oppose Augustine, was simply passing along the summary and paraphrase of the City of God that he found in Ridevall, and that neither fourteenth-century writer has done much beyond compressing the text and slightly simplifying its language. The similarity of all three sources becomes clear when one compares Higden's and Ridevall's Latin with that of the key passage from Augustine cited above; the choice of words is strikingly similar, nor can I detect any significant changes in Augustine's tone or implied conclusions.Footnote ¹⁴ The next sentence in the Polychronicon, moreover, is a direct quotation from Augustine that Higden explicitly flags as such — a peculiar move to make without comment if the previous several dozen words had been intended to oppose him. And this is just one of several points at which Galloway's article labels as a commentator's innovation a remark that in fact derives directly from Augustine.Footnote ¹⁵ In the end it appears that the proposed group of commentators intermediate between Augustine and Chaucer, like the earlier proposals for single mediators, does not explain Chaucer's remark about compassioun any better than does the hypothesis that Chaucer read Augustine directly: here as there, the compassionate tendencies available in the proposed intermediaries are already present in a straightforward way in Augustine himself.

If we turn now from the secondary literature to direct readings of Augustine and Chaucer, we will discover evidence of a more positive kind for this essay's two theses — evidence that builds up to a demonstration, laid out in section four below, that Chaucer's first-hand encounter with the City of God is much the best available explanation for the “data” that the two texts provide. That is, such an encounter would offer excellent explanations not only for Chaucer's attribution of compassioun, but for a number of other peculiarities of the “Legend of Lucrece”; and those explanations are stronger by a long way than any that emerge from the hypothesis that Chaucer relied primarily on commentaries and excerpts.

2. Compassion and Context

The demonstration in section four, if successful, will establish this essay's first thesis. The best way to reach that goal, however, is by way of the second thesis, which is to say by showing that many modern readings of Augustine on Lucretia misrepresent the text. In particular, a detailed and relatively extensive reading of the first book of the City of God, put forward in this section and the next, suggests a remarkably simple explanation for Chaucer's surprising line: perhaps he credits Augustine's text with compassioun toward Lucretia because Augustine's text contains considerable compassion toward Lucretia. Though the notion is almost heretical against the background of current critical consensus, it arrives as the straightforward conclusion of a plain-sense reading, provided that the reading does not limit itself to the single chapter of book one in which Lucretia's name appears but takes into account the book's much longer discussion of rape and suicide in general. Such a reading reveals not only that most of the compassionate moments evident in later authors derive (as already noted) from Augustine himself, but also that his text issues other, stronger imperatives toward compassion that later summaries and commentaries entirely omit.

Here it will be helpful to step back briefly to recall the context that prompts Augustine to discuss Lucretia's story in the first place. The first ten books of the City of God have, it is fair to say, two major goals, both of which were prompted by the sack of Rome by Alaric's Visigoths in the year 410. Taking up the pen two or three years later, Augustine is concerned, on the one hand, to rebut the accusation that the calamity befell the seemingly impregnable city because it had abandoned its real and powerful ancestral gods — whose worship had been banned by edicts issued by the emperor Theodosius in 391–92 — in favor of the weak or simply unreal god proclaimed by Christianity. On the other hand, he also writes to console (often in ways quite reminiscent of the tradition of philosophical “consolations” like those of Seneca) the Christians of Rome after the disaster, since they, no less than everyone else, may have gone hungry, lost possessions, seen people they loved being murdered or raped, been raped themselves, and on and on.Footnote ¹⁶ The question in the background is that of theodicy, a perennial one for the Abrahamic monotheisms: how could a just and omnipotent God allow such things to happen, particularly to faithful worshipers who, their coreligionists might presume, had done nothing to earn such treatment? These two broad motivations are closely intertwined, and many passages respond to both, and Augustine only occasionally stops to point out which he is pursuing at any given moment: his style throughout the treatise is the discursive one of an orator or rhetorician. That style itself has been responsible for some of the problems in recent interpretation, as we shall see: the treatise does not aspire to the linear, quasi-mathematical transparency of language and logic sometimes found in the more purely “philosophical” works of the following centuries, from Boethius to high-medieval scholasticism. As a result accurate summary is difficult and requires attention to relatively long pieces of writing: to infer Augustine's meaning from a short excerpt is to invite misunderstanding.Footnote ¹⁷

In the eighteen chapters that precede Lucretia's appearance, Augustine has been working through a short list of topics connected with the sack, keeping an eye on his two basic goals as he does so. The first seven chapters tell us that the (Arian Christian) Visigoths, against all expectation about the usual conduct of war, spared many Romans, both Christians and others, who had taken refuge in Christian churches, thus offering at least a partial bulwark against the accusation that the sack came about as punishment for the empire's conversion: without the conversion, Augustine argues, things would have been far worse. He then turns a bit more toward the work of consolation, taking on its unavoidable questions: why did many of the “just” suffer? Where there was relief from suffering, why did it help many of the unjust as well as the just? Can the just have benefited in any way from their own sufferings? Gradually the discussion takes the form of a list of specific calamities and the author's responses to them. What of the starvation that happened during the siege preceding the sack itself? What of the awful deaths that many suffered in the sack? What of the fact that many of the dead were not buried? What of the long captivities that many underwent?

Here quite a number of passages do double duty, in that the consolations also advance Augustine's argument that the empire's conversion was not responsible for Rome's sack. His response to the question of captivities may illustrate the point. There Augustine invokes the story of the Roman general Regulus, who was taken prisoner by the Carthaginians and eventually sent by them to Rome to negotiate for the release of Carthaginian prisoners — swearing, however, before departure that he would return if unsuccessful. On arrival he argued against an exchange of prisoners, apparently purely on the grounds that he thought it disadvantageous to Rome; and having persuaded the Roman Senate not to do what would have freed him, he voluntarily returned to captivity with the Carthaginians, who tortured him to death in a particularly awful way. Augustine appears deeply impressed by this pagan's outstanding virtue, but his larger point is elsewhere: since Regulus, such a devoted worshiper of the pagan gods that he would accept torture and death rather than break an oath sworn before them, underwent such a terrible captivity, the captivities recently undergone by Christians cannot serve as evidence against the reality, strength, or providential involvement of the Christian god. Anyone tempted to invoke them in that way finds himself caught by one of the dilemmas on which Augustine's rhetorical style frequently relies: either we expect religious worship to provide the worshiper, here in this present life, with “external goods” like health, wealth, and long life, in which case the pagan gods of Regulus cannot be any more real, or more powerful, than those of the Christians who underwent less severe misfortunes; or we do not, in which case none of the misfortunes that befall individuals or cities can count as evidence for the reality or unreality, or the strength or weakness, of their gods.Footnote ¹⁸

Augustine then turns to the last item on his list of misfortunes, and the one that will inspire the appeal to Lucretia: not only were Christians held captive for long periods, but some were raped, and among the victims were not only married women and young women bound for marriage but “virgines sanctimoniales,” women much like the nuns of today, consecrated to lifelong virginity for the sake of their religion. “They,” that is the pagan opponents of Christianity, “suppose that they cast a great accusation against Christians” in relating those attacks, Augustine says (1.16), meaning that here they ask the skeptic's question with renewed force: surely a god who was real, just, and powerful would not allow this?

In the background of Augustine's response, which goes on for several chapters, is an idea about morality and suffering that underlies the rest of the “consolation” he offers, and indeed the whole tradition of consolation on which he draws: the idea that external goods or goods of the body are not what count as “true riches,” not what is most important about human life, not what must be most desperately mourned if lost.Footnote ¹⁹ That background belief is surely the reason that the first sentence of his response is a postulate about passivity, activity, and ethics: “Let it therefore in the first place be laid down and held firmly that the virtue by which one lives rightly governs the body's members from the seat of the mind [animus], and that the body is made holy by the habitual practice of a holy will — and that when that will is unshaken and remains stable, anything another may do, with respect to the body or in the body, that could not be avoided without a sin of one's own, is without guilt of the one suffering” (1.16).Footnote ²⁰ Whatever else the victims of rape may have lost, they did not lose “holiness,” because it is impossible for a person to lose that by anything that happens to her or him; they did not incur any guilt. And Augustine specifies, in a bolder claim to which he will return many times, that not even their bodies were damaged in a way that would matter most essentially. They too remain “holy,” since the body's holiness derives from the mind and persists, no matter what happens to the body, as long as the mind wills to act well. The next three chapters largely work out and emphasize the implications of this fundamental postulate: at least on the earthly plane, a person's most important attributes — goodness, holiness, presumably also virtue — are matters of what she does, not what she suffers.Footnote ²¹

It is the desire to support this postulate that first brings Lucretia into the discussion. At the beginning of chapter 19 Augustine asks rhetorically whether his imagined interlocutors — those “against whom we defend not only the minds but even the bodies of those Christian women raped in captivity as holy” — will “dare to contradict the clear reasoning” by which he has argued that in the rape of a person whose “intention of chastity” remained unchanged, the disgrace belongs only to the rapist, not the victim. But surely, he answers himself, the interlocutors could not thus contradict him, given that they raise up Lucretia, “the well-known [or noble] Roman matron of a former time, with praises of [her] great chastity.”Footnote ²² She too was raped, in other words, and Rome redounds with the praise of her virtue; that historical fact should suffice, even if reasoning has failed, to show that the Christians in a similar situation must be judged equally innocent, that they have lost nothing of the virtue and holiness that was their most important earthly possession. Augustine then retells Lucretia's tale in two sentences, following them with a striking affirmation of her chastity in the form of another rhetorical question:

Here again Augustine will answer his own question, this time by quoting the observation of an unknown commentator. At the moment of the rape, this commentator says, “there were two [present], and one became guilty of adultery” — a remark Augustine festoons with laudatory adverbs: it has been said “uncommonly well” and “truly,” “splendidly” and “most accurately.”Footnote ²³

From this brief summary of the context there already arises an obvious challenge to the more or less standard way in which recent readers have represented Augustine's treatment of Lucretia. The “cruel dilemma” on which he hooks her later in the chapter is actually not much of a dilemma, given that he has forcefully declared which side he is on long before mentioning the dilemma itself. Lucretia is innocent of any unchastity, in his view; the first aim with which he introduces her into his book would be impeded if she were not. He has brought her in, at least initially, for reasons more or less parallel to those that inspired the introduction of Regulus. Just as fair-thinking pagans could not assert that captivities inflicted on Christians demonstrate the falseness of the Christian god when their own stories show that most faithful of pagan worshipers suffering things still worse, so here: no pagan can suggest that the Christian rape victims have lost what is most important, their holiness or virtue, while still continuing to trumpet Lucretia's chastity. And in each case Augustine willingly ratifies the positive judgment that the pagans make about their moral champion.

Where, then, does the cruel dilemma come in? It is connected to Augustine's discursive style. Having introduced Lucretia to his argument with one apparent aim, he has no qualms about then developing quite another line of thought, with a different conclusion, from her presence. The second conclusion is simple enough: it is that the Christian victims under discussion are in fact more virtuous than Lucretia, not less. It would be fair, in fact, to characterize this development as a contribution to a third goal of the first ten books of the treatise, supplementary to the goals of consolation and of clearing Christianity of blame for the sack: Augustine simply wants to show that Christian morality is generally superior to pagan morality. Thus he will go on to argue that the Christian victims of rape who, “having suffered similar things, nonetheless live (quae passae similia vivunt tamen)” (1.19), have reacted in a better way than Lucretia. It is the motion toward this third use of the story that brings Augustine to his dilemma, which he presents, if not quite as a full-blown counterfactual, then at least as something like a thought experiment.Footnote ²⁴

The experiment works this way. After his forceful declaration of Lucretia's chastity, Augustine begins to muse: given that she is chaste, is it not strange and unjust that of the two people involved, it is the innocent one who was killed while the guilty Tarquin was merely exiled? Should you, O Romans, so proud of the justice of your laws and the equity of your legal system, not want to punish a person who inflicted such an extreme penalty, without authority to do so or due process, even on a guilty person, let alone on one who deserved no penalty at all? But that is what Lucretia has done to herself. Unless of course (and here arises the dilemma) you think that she was not innocent of adultery after all: “Or perhaps … she did not kill herself guiltless, but knowing evil of herself? For what if (which only she herself could know) — although to a young man who forced himself in violently — she also consented, allured by her own sensual desire, and, in avenging it on herself, she so lamented it that she supposed that it ought to be expiated by death?”Footnote ²⁵ One aim of the surrounding argument is to establish that Lucretia's killing herself was a moral mistake, an instance of wrong-doing. That it is for Augustine also an understandable instance we will see in the next section; but nonetheless the wrong-doing means, he will conclude soon after, that when the Roman heroine is held up for comparison with the recent Christian victims, the latter come out in the lead. And the only way to lessen the magnitude of Lucretia's crime against herself would be by supposing that there was after all a previous crime to avenge — in which case Lucretia would still be less admirable than the pagans say, though for a different reason. Hence, as Augustine puts it a few sentences later, “this case is hemmed in from both sides.”Footnote ²⁶

But immediately on making that observation he pulls himself up short, turning back from the comparison of moral systems to the general argument about chastity with which he first introduced Lucretia: “Nonetheless it suffices for us, in this case, so well known, of this woman — to refute those who, not understanding any deliberation about holiness, revile the Christian women who were lain with in captivity — that it has been said in the celebrated commendation of her: ‘There were two and one committed adultery.’ For by them Lucretia has rather been believed to be of a kind who could not stain herself by any adulterous consent.”Footnote ²⁷ Although Augustine does not here repeat the declaration from the chapter's beginning that he agrees with “them” about Lucretia, we do clearly find ourselves back in that space, where Lucretia is invoked as the example whose reputation for chastity will convince, and convince even those unreceptive to other arguments, that there is no guilt in being a victim. The dilemma has done its work and can be put aside, revealed as a bit of a red herring, more a matter of looking down every possible pathway than of a possibility that Augustine seriously entertains.

Thus the explicit treatment of Lucretia in the City of God is less cruel, less eager to entrap its hapless subject in one kind of condemnation or another, than it has often been made out to be. Augustine is even ready to praise her for the virtue of chastity, just as he praises Regulus for his steadfastness in keeping faith — a willingness made only more remarkable by the surrounding polemic against the culture in which both lived. Augustine is also adamant about the innocence of those victimized by another's will, a point that may strike us as obvious but that he clearly thought needed defending, given the amount of space he devotes to it. Nonetheless, however much less ferocious these bits of contextualization make him appear, one might fairly observe that they fall short of justifying a verdict of gret compassioun, which should require more than the simple absence of cruelty. To see what might inspire Augustine's reader to that verdict, we need to look more deeply at what he says not about Lucretia's particular case, but about rape in general, and especially at how he thinks it differs from other crimes.

3. What Augustine Really Said about Rape

Augustine's great concern to declare that no one can become guilty or lose “holiness” of mind or body by anything that happens to her or him has been stressed enough. But immediately after the above-remarked strong statement of that ethical postulate (1.16), Augustine adds that sexual assault raises a special complication in this regard. Because “not only that which pertains to pain but that which pertains to sensual desire can be perpetrated on another's body,” there is the risk that the victim will be thought to have been seduced by pleasure into a secret assent, thus getting her own will entangled in the crime and incurring guilt after all. It is that possibility, Augustine thinks, that gives rise to the shame that victims often feel, even when no such assent has happened: “Whatever of this sort [i.e., sexual assault] may be done, although it does not drive out chastity [or “shamefastness”: pudicitia] that has been preserved by dint of a most constant spirit, nonetheless inspires shame [pudor] lest it be believed that what perhaps cannot happen without some pleasure of the flesh happened also with the will of the mind.”Footnote ²⁸

That account of victim's shame is surely far too simple.Footnote ²⁹ Nonetheless it is worth pausing a moment to appreciate what Augustine has said, and even more what he has not said. He has not said that rape victims commonly do find themselves assenting to the criminal assault. He has not even said that such a thing has ever happened. It is merely a matter of the possibility that third parties learning about a particular rape may imagine that it has happened. And, to be still more precise, Augustine does not even directly assert that possibility: what he says is that victims often fear that third parties may imagine that a secret assent happened. Thus his concern is, as far as I can tell, entirely sympathetic with the plight of rape victims. And it is sympathetic both in a general sense and a precise etymological one: it is an attempt to understand and describe what a rape victim experiences, to feel-and-suffer-with her.

To this point his efforts have all been on the level of factual description. What follows next is the “ought,” the moral response expected of bystanders, and it is if anything still more remarkable. “But on account of this,” Augustine writes, “even those women who have killed themselves lest they suffer something of this sort — what human feeling would not wish to forgive them? And those women who declined to kill themselves, lest they avoid another's disgraceful act by a misdeed of their own — whoever imputes this to them as an accusation will not himself ward off the accusation of being a fool.”Footnote ³⁰ Strong words in themselves, but much stronger when one considers their full context: that of an already famous and authoritative bishop who is about to reinforce his own emphatic declaration that suicide is sinful with another eight chapters on the subject, drawing his conclusions there in large part directly from the Decalogue. And yet here, a few chapters before that discussion begins, he offers a qualification: even if some woman does commit this serious sin, if it comes about under the awful duress imposed by sexual assault, we will, and should, want to forgive her. More: to want anything else is inhuman. It is hard not to see here a clear and forceful expression of sympathy (in both the loose and the etymological sense) for the victim — sympathy and, yes, compassion. In fact the word I have translated as “feeling,” adfectus, generally has a positive rather than a neutral valence, and thus instead of “human feeling” the phrase could easily be rendered “human compassion.”

Augustine's sympathy does not, to be sure, alter his judgment that suicide — any suicide not divinely mandated, and thus of course Lucretia's — is sinful. But it does provide, for her case and others like it, a frame that cannot be overlooked.Footnote ³¹ It reminds the reader that to say that someone has committed a sin is not, for a Christian theologian, to condemn the person irrevocably. It is, in fact, merely to say that the person is human, as in general the religion takes rather seriously the various biblical declarations that “all have sinned” (“omnes enim peccaverunt” [Romans 3:23]; cf. Romans 3:10–18, a passage that reproduces several verses on the theme from the Hebrew scriptures, and also 1 John 1:8–2:2). Thus there is no contradiction between pointing out someone's sin and exercising compassion toward her or him; in fact if one is not going to be compassionate toward sinners, one is not going to be compassionate at all, as there is no one else available to receive the gift. The distinction between hated sins and beloved sinners makes possible the complexity of Augustine's reaction to Lucretia, the fact that he can praise her virtue and have compassion on her plight while still asserting that she sinned; critics who miss the distinction, or who do not believe that it could be sincere, will quite naturally also disbelieve the complexity. For that matter, a simple failure to read, or remember, these lines from Augustine's seventeenth chapter would have the same effect, as they do not merely mitigate the negative things said two chapters later; they transform their meaning.Footnote ³² After the directive that we should want to forgive, a condemnation of a particular victim's suicide cannot come across primarily as intended to condemn the victim. Quite the opposite: the condemnation itself can be read as yet another expression of compassion. It is at least partly aimed at keeping other victims from harming themselves further.

In the face of these several strong exhibitions of sympathy, it is fair to ask how so many contemporary readers have concluded that Chaucer's remark about compassioun is not only wrong but obviously and unquestionably so,Footnote ³³ and how it is that one infrequently finds critics saying anything else.Footnote ³⁴ As with the often incomplete understandings of his dilemma concerning Lucretia, at least part of the blame can be laid at the door of a discursive style that several times leads Augustine to the brink of saying something rather horrible to modern sensibilities — leads him, indeed, so close to the horrible thing that hurried readers, particularly if they do not drink in the entire context, may easily come away thinking it has actually been said. For example, one who reads in isolation the statement that “a woman violently overcome and oppressed by another's sin, without any consent of her own, does not have anything that she is avenging against herself by voluntary death” (1.18) may easily form the impression that Augustine intends also the logical “inverse,” namely that a woman who did secretly consent to an assault and then killed herself was avenging something against herself and was therefore justified in the suicide. But a more careful reading, taking into account the broader context, makes it abundantly clear that Augustine means nothing of the kind. Although, as already remarked, he does believe that the killing of an innocent person (oneself or another) is a worse crime than the killing of a wicked one, both kinds of killing — bracketing the highly regulated situations of military action and judicial execution, and the rare and risky possibility of direct divine command — remain clearly forbidden.Footnote ³⁵

A second example of a tempting misreading is important to note because, though small and subtle in itself, it is capable of discoloring all the light in which Augustine's attitude toward pagan Roman culture appears. Readers sometimes look at the first passage from Augustine quoted in the present essay, the one connecting Lucretia's suicide with pudoris infirmitas rather than pudicitatis caritas, and infer that the essential problem with Lucretia's suicide was that it was motivated by shame — as if, again, a suicide motivated by love for lost chastity would have been permissible. But here as in the previous case, Augustine has not said what has been heard, and a reading of the surrounding chapters makes clear that he intends something very different.Footnote ³⁶ The problem with Lucretia's suicide is simply that it is suicide, and for Augustine suicide, regardless of circumstances or motivation (barring his single exception, irrelevant here), is a sin. In fact it is not even clear that Lucretia's motivation by shame has made her crime worse in Augustine's eyes. Here the analysis becomes regrettably grim, but its logic seems to proceed as follows: since killing the guilty, while still criminal, is judged to be less bad than killing the innocent, a suicide motivated by an accurate perception of prior guilt would be a lesser crime than Lucretia's suicide. But it does not follow that the shame that (according to Augustine) motivated the latter is what has made the crime worse: Lucretia's innocence has done that. The shame, in fact, may well weigh in on the opposite side, by helping to give an account of why this woman, whose chastity Augustine has long since ratified and the awfulness of whose plight he has considered at length, has reacted wrongly to her wretched situation. In that case his notion about Roman eagerness for praise functions not as an argument for the prosecution, but as a mitigating explanation offered by the defense, a cultural background that tells why the wrong choice might come easily to this otherwise admirable person. Such an interpretation would not conflict with Augustine's clear interest in showing the superiority of Christian morality and culture to their Roman competitors; it would merely mean that he distinguishes between a culture and the people who inhabit it, and is quite capable of pointing to weaknesses in the former without issuing blanket condemnations of the latter. It would mean, in other words, that his invocation of Lucretia's fear of shame functions as yet another summons to compassion for her.Footnote ³⁷

It should now be clear that, so far from being ripe for out-of-hand dismissal, Chaucer's remark about compassioun is quite a plausible one; what he means is even evident on the surface of Augustine's text. All its discussion about rape victims in general, before Lucretia makes her entrance, lays a foundation for an account of Lucretia herself that is in many ways laudatory and sympathetic, despite the polemical context that helps inspire criticism along with the praise. It is worth calling again to mind the powerful sentence with which Augustine ends her story: “foedi in se commissi aegra atque impatiens se peremit” (grieving over and unable to bear the abominable thing committed against her, she killed herself). Even if “unable to bear” suggests criticism as well as pathos (that is, the criticism that she might have been able to bear it had she not been habitually overprotective of reputation), it is difficult to miss the sympathetic tone of the whole — a tone that reinforces one more time the thesis, this essay's second, that our contemporaries have largely misrepresented Augustine's attitude. And, of course, once that thesis is affirmed the hypothesis of deliberate irony in Chaucer's line more or less dissolves of its own accord: there is no need to posit irony in an author who is merely reporting what seems to be the case.

The last task before us is a return to the first thesis, the question of Chaucer's direct knowledge of Augustine. The evidence presented to this point should have effectively warded off any objection based on the mistaken but common idea that there cannot have been any such knowledge because Augustine's text cannot be understood as compassionate. But the evidence should also have done far more than that, because the large majority of the arguments for compassion just presented derive from parts of the text that are simply not reproduced in the proposed substitute sources — neither in the Gesta Romanorum, nor in Simon de Hesdin's Valerius, nor in the Polychronicon.Footnote ³⁸ As a result, if we find a later writer calling Augustine compassionate, the safer inference is that the writer probably did read Augustine directly rather than that he did not; there is simply more compassion to be found there than in any of the obvious alternatives. We turn now to a final step that offers evidence of a different kind to strengthen that inference still further in Chaucer's case: the above-mentioned peculiarities of the “Legend of Lucrece” that are easy to explain if Chaucer did have a direct acquaintance with Augustine but difficult to explain if he did not. These appear in two contiguous passages near the poem's end.

4. Chaucer's Familiarity with Augustine: Evidence in the “Legend”

To understand the first of the two passages, it is necessary to broaden the field of investigation in a second way: not only by taking into account the whole of Augustine's discussion, but also by briefly comparing parts of Chaucer's tale to the various renditions of Lucretia's story mentioned above as certain or possible sources — those in Ovid, Livy, the Gesta Romanorum, Simon de Hesdin, and Higden's Polychronicon. The basic question concerns the moment of the rape itself. In all these versions, as in Chaucer's, Tarquin, generally after having found a series of entreaties, bribes, and threats powerless to convince Lucretia to yield, hits upon the worst threat of all: he will not only ensure her death but defame her memory, by putting a murdered servant in her bed and claiming to have caught them in the act of adultery. At this point in all the possible sources just listed it is clear that Lucretia, overcome by horror at the possibility, gives in. But Chaucer gives her a different reaction:

It is not just a brief fainting spell. She is so deeply unconscious that nothing, not even ghastly violence to her body, could wake her. And given its novelty with respect to every work suggested as a source, it seems to be entirely Chaucer's invention.Footnote ³⁹ What has prompted such a drastic change?

A number of possible answers have been helpfully laid out in Lisa Kiser's treatment of the story, already mentioned with respect to the question of compassioun. In line with a wider set of similarities that Kiser notes between the Legend of Good Women and late-medieval hagiography, Lucretia's swoon fits the convention in some saints’ lives of declaring the saint miraculously spared from the pain of some of the worst tortures (cf. St. Cecilia's imperviousness to the heat of the hypocaust in which the Romans try to roast her — a feature of her story in the Golden Legend and elsewhere, preserved in the Second Nun's Tale, line 521). It also could be taken to foreshadow, in a way that would further mimic some hagiography of the time, the approaching real death of the “saint.” And it adds another degree to the depravity of the aggressor, in that he is willing to rape a (temporarily) lifeless body.Footnote ⁴⁰ While all these thoughts are reasonable enough, a fourth explanation is possible — one which Kiser also mentions, but only briefly and parenthetically. To my mind, however, if we follow the hypothesis that Chaucer may have been familiar with all the relevant passages of the City of God and not just the chapter that explicitly names Lucretia, this fourth explanation easily eclipses all the rest.

It is at root quite simple: it is merely the idea that Chaucer, in the context of writing a Legend of Good Women, wanted to create a Lucretia who was as “good” as possible. There is strong independent evidence for that intention. There are, first of all, the similarities to hagiography Kiser highlights for many of the legends, capped in Lucretia's case by the poem's declaration that after the resulting demise of the monarchy she “was holden there [sc. in Rome] / a seynt, and ever hir day yhalwed dere / As in hir lawe” (1870–72).Footnote ⁴¹ More tellingly, there is the conceit according to which the writing of the Legend has been urged upon Chaucer as a kind of penance for previous works that impugned the faithfulness of women — a conceit that makes considerably more specific the kind of “goodness” to be imputed to the lives he will narrate. Queen Alceste gives the poet his orders: Chaucer is to spend his time “In makyng of a glorious legende / Of goode wymmen, maydenes and wyves, / That weren trewe and lovyng al hir lyves; / And telle of false men that hem bytraien …” (F 483–86, G 473–76; the entire surrounding conversation with Alceste and Cupid is worth rereading). In the story of Lucretia, at any rate, the poet has followed his fictional directive closely, so often going out of his way to point out her “trouthe” (e.g., at lines 1843, 1860, and 1874–75) that its establishment as her outstanding characteristic is arguably the poem's most evident goal. And that goal, whether we infer it from the Prologue or the story itself, makes quite clear why Chaucer would wish to expunge from the tale any possibility of an unchaste complicity on Lucretia's part. In fact he appears to have taken other steps in that direction: consider the fact that he has Tarquin arrive by night and steal secretly into Lucretia's house, rather than being received by her, as he is in all the potential sources, as a guest. It is a striking innovation, almost certainly motivated, as Edgar Shannon observed ninety years ago, by Chaucer's desire to make Lucretia's “perfect chastity and innocence appear in greater relief.”Footnote ⁴² Her newly invented unconsciousness at the moment of the rape simply moves further in the same direction. Changing the story so that Lucretia no longer welcomes Tarquin for an overnight stay erodes the basis on which an ill-disposed onlooker might generate suspicions of complicity; rendering her unconscious during the attack more or less washes any such basis entirely away.

A second question immediately arises, however. Nearly all the possible sources for Chaucer's story answer the question of Lucretia's possible culpability in the rape with a rather clear negative. They thus let the entire weight of blame fall on Tarquin; and they do so without resort to any such drastic expedient as depriving Lucretia of consciousness. Augustine merely states her chastity as a fact, offering no more argument for it than his agreement with the apparently commonplace refrain about there being only one adulterer in the room. Livy and Ovid treat the case in more detail, but for establishing her innocence it suffices them to have her husband, father, and the other sympathetic men to whom she tells what has happened comfort her with unequivocal declarations that she, as the one on the receiving end of the crime, bears no guilt. Why, in the face of so much authoritative declaration that Lucretia is indeed a “good” (that is, a faithful) woman, did Chaucer feel the need to safeguard her innocence further by a major change in the story?

One answer that looks promising at first, but proves imperfect in the end, is that Chaucer may have felt the need to defend Lucretia also against a second kind of accusation, one that may have become a more pressing concern for him than it was for his sources, even though the groundwork for it exists in Ovid and Livy. That groundwork consists in the already noted fact that in their versions of the story (as in most subsequent ones) Tarquin perpetrates his crime not by brute force but by the threat to which Lucretia yields; to some later authors, despite the contrary declarations of the sympathetic males in the Roman sources, her yielding has seemed a culpable defection from the truer path of virtue, which would have been, they say, allowing herself to be falsely defamed and killed. There is little clear evidence of such an accusation in the works commonly suggested as Chaucer's sources: though the Gesta Romanorum, for example, does in various manuscripts declare that Lucretia concessit (yielded) or consensit (consented) to Tarquin, it juxtaposes that affirmation with the statement that she was coacta (compelled), and the combination leaves the reader uncertain about whether some degree of responsibility is being imputed to her.Footnote ⁴³ Augustine, for his part, goes to the opposite extreme, rendering any such accusation impossible by a bold move at the beginning of his account: he removes any possibility of a choice between evils on Lucretia's part by eliding all mention of threats, bribes, and promises, and simply stating without ambiguity, in a line already quoted, that the deed was a matter of one-sided violence.Footnote ⁴⁴ Nonetheless the idea of a cowardly Lucretia who should have accepted murder was abroad in at least some circles by Chaucer's time: in particular a number of treatises on canon law suggest that because the compulsion (coactio) directed at Lucretia was “conditional” rather than “absolute,” some degree of consent and therefore guilt (as regards the rape) should be attributed to her. Some went so far as to declare her guilty of adultery and of mortal sin. If Chaucer had gotten wind of accusations of that sort, one might imagine, perhaps he could have felt sufficiently alarmed to want to guarantee Lucretia's innocence against them; and having her swoon rather than yield at the crucial moment would meet the need.Footnote ⁴⁵

It is important to keep these two possible accusations distinct in one's mind, and modern readers have not always done so.Footnote ⁴⁶ The first of them (I will call it Accusation A) is that Lucretia, at least in part or “at some level,” secretly welcomed Tarquin's attack — that if she was not pursuing active adulterous desires, she at least exerted some volition that would substantiate a verdict of some degree of collusion or consent. The second (Accusation B) has nothing to do with any desire or welcoming on her part; it allows that Tarquin's attack was to her, as her interaction with her relatives the next morning suggests, a horror almost literally unspeakable. But it maintains that faced with a choice of two horrors, rape and murder-with-defamation, Lucretia chose badly in a morally culpable way. The plausibility of either accusation is not what is currently at issue: in fact, given the basic events narrated in all versions of the story discussed here, each seems difficult to assert, not to mention simply cruel. What is at issue instead is the degree to which various authors’ tellings remove or leave intact the possibility of each charge. Livy's version and Ovid's are alike in apparently sparing her the guilt of either accusation by having seemingly authoritative characters declare her innocent; but they are also alike in leaving open the possibility that later readers will think otherwise — not least by putting the declaration of innocence merely in the mouths of characters rather than in the tale-teller's own voice. Augustine's version, as just remarked, forcefully does away with the necessary precondition for Accusation B, namely an account of choice on Lucretia's part; but it does remain open to A as a logical possibility, though one that Augustine himself disavows. Chaucer's drastic innovation entirely excludes both, which is why it would be thinkable to conclude that a newly active concern about type-B accusations had prompted the change.

It would, however, also be somewhat less than convincing. First of all there is the precedent of the ancient authors. Since Livy's story and Ovid's both leave open the possibility of type-B accusations from later readers, but also successfully fend it off for themselves while leaving Lucretia conscious, it is hard to see why Chaucer could not have done the same — first of all simply by repeating what Livy and Ovid say, and then, if the actual presence of such accusations here and there in high-medieval culture had reached Chaucer's ears, perhaps by strengthening it. After all, Chaucer is happy enough elsewhere to make all sorts of choices about the materials he inherits, sometimes selecting one auctoritee over another, sometimes bucking several to create his own new version of events, occasionally even breaking into the narration to point out what he is doing. It is not difficult to imagine his asserting himself similarly here in order to forestall any notion of “consent to conditional compulsion.”

But there is also a second reason to be skeptical about the swoon as a response to Accusation B, and to believe that Chaucer's main motivation lay elsewhere: there is a competing explanation that would not only make the need for the swoon more compelling, but would render extremely significant some of the language in which it is expressed — language that otherwise looks unimportant or even puzzling. The competing account is simple: Chaucer would have had special reason to introduce the swoon if he had read and been influenced by the City of God.

It is an answer that will gain more strength the more of book one we reread ourselves, because the fear of Accusation A — the fear that rape victims may be suspected of having been seduced by pleasure into secret consent — will reveal itself more and more as a central element of the author's thinking. Whether one finds that the fact speaks well or ill of Augustine's ideas about rape, it is impossible not to be struck by the frequency of his reversion to this concern: besides its most deliberate treatment, already discussed, in chapters 16 and 17, it arises as a logical possibility in chapter 19, in the context of the “dilemma”; then returns in chapters 25 and 27, after a pause while suicide and killing in general are discussed, as soon as sexual violence is taken up again; and it is present explicitly in chapter 28, and implicitly in 18, in the course of further emphasis upon the central proposition that a victim who does not consent is guiltless. Virtually none of this material appears in any other writer, earlier or later.Footnote ⁴⁷ And because Augustine accompanies all this worrying about the accusation with his own affirmations of Lucretia's chastity, the practical impact of his text is a peculiar hybrid: it manages simultaneously to strengthen his reader's impression of Lucretia's innocence and to inculcate into that same reader a nearly unshakeable fear that other hearers of the story might think otherwise. He has, after all, both declared his own belief that her guiltlessness with respect to Accusation A is so certain as to require no belaboring (text to note 23) and declared that he, and everyone else, will be eternally unable to prove that guiltlessness, since the logical possibility of hidden consent (“which only she herself could know”) acts as an all-but-uncloseable trap door in any attempt at proof (text to note 25).Footnote ⁴⁸ Given that irksome joker in the pack, according to which some undetermined third parties in the indefinite future may always find themselves tempted to think Lucretia guilty, it would not be surprising if a reteller working under Augustine's double influence were to go looking for ways to settle the question more decisively — particularly if the reteller had reason to want his character to fall beyond all suspicion of unchastity. The swoon that Chaucer invents meets the requirements of such a settlement quite handily, destroying as it does the possibility of either kind of accusation, A or B. In fact some such revision to the story might be the only way of definitively closing the trap door that Augustine's account has left open.Footnote ⁴⁹

The suspicion that this result is in fact a purpose, that Chaucer has made his change precisely in order to do away with Augustine's tenacious worry, grows still greater when one looks again at the language the poem uses for the swoon. It will help to recall first the strangely lyrical repetitive construction, almost a pun, with which Augustine concisely describes Lucretia's plight: she was led to suicide, he writes, out of fear “that it would be supposed, if she lived [si viveret], that she underwent freely that which she underwent violently while she lived [cum viveret].”Footnote ⁵⁰ Whatever Augustine intends by that rhetoric, one of its effects is to emphasize that the fact that Lucretia is alive during the rape is a presupposition necessary for the worry that she secretly assented. Might not a poet prompted by that observation remove the worry by temporarily depriving her of life-force? Is it not remarkable, in that context, that Chaucer describes the swough as one in which Lucretia wex as if ded (1816)? And if we also bear in mind the basic mechanism behind Augustine's worry, expressed three chapters earlier and so constantly returned to, the final line of the swoon's description becomes still more remarkable. In itself — which is to say, if Augustine's concerns are not in the picture — it appears to add nothing, even to be merely a kind of poetic filler. But in the context of a persistent fear that rape victims will be thought complicit because of physical pleasure that might attend the act, the poet suddenly has a very precise reason for telling his readers that Lucretia “feleth no thyng, neyther foul ne fayr” (1818). The line not only allays Augustine's besetting concern about Lucretia but even recalls the structure of his introductory thoughts on rape in general: in the terms quoted above from chapter 16, Chaucer is telling us that the unconscious Lucretia escapes both sorts of passivity that may attend a sexual assault, both the experience of “quod ad dolorem pertinet” (that which pertains to pain) and that of “quod ad libidinem pertinet” (that which pertains to sensual desire). No other proposed source mentions such a distinction at all. Moreover, only two others, namely Ridevall's commentary and the Polychronicon, reproduce any trace of Augustine's repetition of viveret, and they include only the remark about Lucretia's survival after the rape, not the one about her being alive during it (cf. note 14). Thus it is only the hypothesis of Chaucer's direct familiarity with Augustine that could invest these two lines from the poem with such robust significance. If instead Chaucer was seeking merely to head off the accusation that Lucretia culpably “yielded” to Tarquin out of fear of his threat, the special significance vanishes: the poem's remarks about her inability to feel pleasure and her deathlike state, responsive as they are to Augustine's concerns and diction, add to the defense against that type-B accusation nothing beyond what simple unconsciousness has already provided.

That concludes the consideration of Lucretia's swoon — one of the promised two passages providing internal evidence for Chaucer's direct reading of Augustine. The other, which can be handled much more quickly, is the single couplet immediately preceding the swoon. It too is an innovation on Chaucer's part: the two taken together, and completed by a short trailing apostrophe that condemns the “vileyns dede,” make up a single thirteen-line interruption in a flow of events that otherwise adheres closely to Ovid. The addition begins immediately after Tarquin issues his threat to frame Lucretia as an adulteress, which will mean, he concludes by telling her, that “thow shalt be ded and also lese / Thy name” (1810–11). Chaucer not only declares that the swoon comes about in response to that threat, but also specifies why: because “These Romeyns wyves lovede so here name / At thilke tyme, and dredde so the shame” (1812–13) that the combined threat of sclaunder and deth knocks this particular Roman wife out cold. The lines contain a subtle, but striking, parallel with the City of God. While nearly all the versions of the story here considered make it clear that Lucretia fears shame, it is Augustine alone who delivers any such comment about Roman wives as a class: his remark, again contained in this essay's first long quotation, associating Lucretia's shame with her status as a “Roman wife [mulier], excessively eager for praise” (1.19). Livy and Ovid make no such comment, and the later retellers writing in Augustine's wake generally omit it — with the exception, of course, of Chaucer himself. The only other widely known writer who comes close to preserving it is Higden, but he changes it to “gens Romana maxime erat avida laudis” (the Roman people was eager for praise) — thus providing yet another reason to think that however much Chaucer may have learned about Augustine from the Polychronicon, he all but certainly learned about Augustine from Augustine as well.Footnote ⁵¹

5. Chaucer and the City of God: The Evidence in Review

We have seen, by my count, six pieces (or in some cases “groups”) of evidence suggesting Chaucer's direct knowledge of Augustine's treatise; three derive from the treatise and three from Chaucer's poem. The first is that the instances in which certain later writers repackaged the contents of the City of God in a way allegedly more compassionate toward Lucretia turn out on closer inspection to be illusory, as such compassion as they possess derives quite directly from Augustine himself. The second is that the treatise shows several kinds of compassion that are routinely overlooked, including an effort to imagine the mental world of rape victims and an explicit statement that suicide in such a case, though wrong, should prompt a desire to forgive rather than to condemn — and also including some affirmations of Lucretia's chastity so strong as to make clear that the text's apparently cruel moral dilemma is a matter of covering logically possible bases rather than of anything that Augustine seriously entertains. Third is the observation that the passages containing those outstanding expressions of compassion are omitted, almost without exception, by the later authors commonly proposed as intermediates between Augustine and Chaucer.

The fourth piece of evidence is that one of Chaucer's most dramatic innovations in the story — the fact that Lucretia passes out rather than “yielding” in response to Tarquin's last threat — would be well explained as the reaction of an author primed to fear that some might suspect her of a kind of guilt that would be hard to disprove in any other way; and it is the City of God, rather than any other relevant source, that thus primes its readers. Fifth, Chaucer's language for the swoon encourages the inference that the innovation was meant to abrogate Augustine's concern: the poem's observations that Lucretia wex as if ded and felt nothing fayre both negate conditions (respectively life and pleasure) that Augustine notes as requirements for the third-party accusation of secret assent that so worries him. And finally, Chaucer's nearby assertion that Roman wives as a class dreaded shame seems to appear in no relevant source but the City of God itself.

The case is strong: even if it stops short of demonstrative proof, it seems to me easily sufficient to make the odds that Chaucer knew the treatise at first hand — or at least knew the substantial section of its first book that concerns itself with sexual dishonor and suicide — far stronger than the odds that he did not. And that conclusion, once established as the more probable of the two, should affect further thinking about Chaucer and his work in at least two ways. First of all, it suggests that Chaucer not only knew Augustine's treatise but understood it well enough to shape one of his poems partly in a targeted response to the detailed thinking about human psychology he found there — thus engaging with Augustine in a quite complex, human-centered, and, as it were, colorful way. It is particularly gratifying to see that thoughtful craft at work in the Legend of Good Women, an unfinished collection whose handling of its heroines has sometimes been experienced as so painfully monochrome as to suggest that Chaucer (or the work's fictional “narrator”) must have abandoned it out of sheer boredom.Footnote ⁵²

The second set of consequences is, in a way, simpler. It comes merely from noting that if Chaucer knew well and took seriously a substantial section of the first book of the City of God, the effects of that knowledge will likely appear in other poems than this one, and our readings will be stronger if they take account of the likelihood. Some complication enters with the question of how much more of the treatise Chaucer may have read. It seems probable that if he knew these middle chapters, he also knew the rest of book one: as noted above, medieval manuscripts separated the individual chapters much less than modern typography does, and I have seen no evidence that the thirteen chapters on rape and suicide were ever extracted as a unit. It is also reasonable to consider, though with a proportionately increased admixture of caution, whether Chaucer may have been familiar with the entire treatise. That interesting possibility will gain in likelihood if one recalls that the City of God was enjoying a kind of renaissance of interest in the fourteenth century, that the renaissance flourished among humanist authors (like Trevet and Francis Petrarch) well known to Chaucer, and that it involved a special fervor for working from the original text rather than from excerpts and summaries.Footnote ⁵³ But regardless of which of these levels of knowledge of the text one finds plausible, there will be consequences for Chaucer's poetry. Even the few sections investigated here with respect to the “Legend of Lucrece” harbor questions of evident concern to other poems — most clearly, perhaps, to the end of the Franklin's Tale, whose interpretation will depend very much on what ideas about the morality of suicide under duress (and under sexual threat in particular) we imagine to be presupposed, or aired for debate, there. If we then broaden the scope of Chaucer's likely knowledge by just a few pages, it will take in Augustine's treatment of the question of death without burial, considered among the ravages of war in chapters 12 and 13: and that question not only appears, but appears with a repetitive insistence Chaucer adds to his sources, in the mouth of the Clerk's Tale's Griselda. Augustine, of course, argues strongly (as many pagan philosophers also did) that lack of burial is not to be feared. How, then, might our understanding of Chaucer's character change if, as now appears likely, her creator has simultaneously emphasized her Christianity, relative to his sources, and knowingly made her fret at great length over a question that a leading Christian authority of the age has forcefully declared to be of no ultimate importance? Finally, if Chaucer knew the remainder of the City of God, the number of possible connections with his work grows large indeed: the extensive considerations of Stoic philosophy in books nine and fourteen, for example, may well be able to cast light on the several works (most patently the Clerk's, Man of Law's, and Knight's Tales and the Melibee) in which Chaucer meditates at length on the questions about human suffering and the relationship between passivity and activity that are so central to the ethics of that school.

There are, in a word, many possibilities for further work. There is also, of course, the need for caution: surely any critic will have to acknowledge the ease and frequency with which apparent evidence can offer itself in support of patterns and connections that he or she has in mind, but that later prove illusory. But a need for care is not the same as a signal to abandon the enterprise — nor even as a signal to abandon working from patterns and connections, if only they can be treated as hypotheses to be tested rather than as foregone conclusions to be shored up at any cost. Here I have attempted to establish such a hypothesis about one poem of Chaucer's, and I hope that the relatively restricted scope of the claim has helped make the argument convincing. If the claim holds, my next hope will be that others, fuelled by a similar blend of eagerness and care, will find here tools and materials useful in putting together their own, perhaps more expansive, efforts at understanding.Footnote ⁵⁴