Agency and Ethics in The Spanish Tragedy

In his 2012 booklet Free Will, neuroscientist and 'new atheist' writer Sam Harris argues that 'free will is an illusion', and that 'if the laws of nature do not strike most of us as incompatible with free will, that is because we have not imagined how human behaviour would appear if all cause-and-effect relationships were understood'.1 Though he does offer some neurological and biochemical instances of causation, he does not claim to understand all causes, calling the most rudimentary self-analyses (why, he asks himself, have I resumed the martial arts?) 'utterly mysterious' and asserting that 'there is a regress here that always ends in darkness';2 consequently, we are leftwith little more than an assertion of faith in unspecified causality itself as the real truth underlying the illusion of autonomous human volition and agency. This is not a very serious book-it is thin and repetitive, contains numerous logical sleights of hand, and literally ends on page 66 with the author (apparently wearied by a long afternoon's work) wandering offin search of a snack-but it is connected to the extremely serious, and longstanding, philosophical problem of human agency, which addresses the relations of action, causation, and responsibility. In modern terms, 'determinist' philosophers argue that all human will and actions are fully caused, leaving no room for real choice or will; 'libertarians' contend that we do act by non-externally-dictated free will, thus disproving the fullest claims of determinism; 'compatibilists' aver that determinism and free will are logically reconcilable; and all parties argue over the degree to which moral responsibility requires the capacity to will and/or act independently of external compulsions or constraints.3 As a resolved skeptic and determinist, Harris is scornful of libertarians, who tend to 'invoke a metaphysical entity, such as a soul, as the vehicle for our freely acting wills', and compatibilism, whose discourse suspiciously 'resembles theology'.4

In this last point Harris is quite perceptive: while the secular version of this problem runs from Epicurus and the Stoics to the present day, the question of agency has over most of the last two millennia very often been approached as a theological issue5-even, and perhaps especially, in its determinist forms. Can humans, either wholly or partially, justify and save themselves, or is salvation entirely the work of divine grace? The conflict can be discerned in the New Testament, where in the first century James insisted on the saving importance of works, even as Paul reinterpreted the Law as an absolute marker of humanity's radical inability to be righteous before God. It was recapitulated in the early fifth century, when Augustine crushingly rebutted Pelagius' denial of original sin, and his contention that humans could live righteously under their own power, by arguing that the capacity of the human will to choose the good had been taken away by sin, and could only be restored by the sovereign intervention of God; in the ninth century, when Gottschalk of Orbais not only reasserted Augustine's teachings on grace and predestination, but extended them to positive reprobation (that is, the belief that God specifically ordains individuals to hell as well as heaven), and for this was degraded and convicted of heresy; in the thirteenth century, when Aquinas sought to reconcile divine sovereignty and human agency, and arguably came closer to succeeding than anyone before or since by contending that human freedom, while real, was secondary and subordinate to that of God; in the sixteenth century, when Erasmus and Luther famously argued over whether the human will was free or in chains, and Calvin reinstated the doctrine of double predestination, which even most Augustinians had denied, resisted, or quietly avoided; in the seventeenth century, when an intra-Protestant, anti-Calvinist reaction insisted, like Catholics, that people must contribute something to their salvation or damnation if that judgment is to be just; and so on and so forth. The history of this perennial issue has proven to be remarkably incompatibilist in theory, though often (perhaps necessarily) accommodationist in practice, and fundamentally conceived as an ethical negotiation between divine sovereignty and human capacity, between determinism and the possibility of action.

The subgenre of revenge tragedy that flourished in late-Elizabethan and early-Jacobean England, typically structured as a problematic contest between an individual who desires to act and some array of earthly and transcendent constraints, engages this issue deeply in its bloody focus on the problematics of action. Over the past eight or so decades, critics (including Campbell, Bowers, Hunter, Prosser, Broude, Diehl, Watson, Rist, Woodbridge, Kerrigan, and Greenblatt) have valuably considered various dynamics of individual action and constraint in these plays-and more specifically, whether particular plays or the subgenre as a whole were for or against revenge (that is, whether Christian teachings and God's prerogative justice prevail over pagan desires for equity), as well as whether the logic of the subgenre was more Protestant or Catholic.6 These are valid questions, but they have not always proven optimally productive, perhaps because of the binary and/or polemically fraught ways in which they are often framed and pursued, and perhaps because such readings often take an insufficiently synecdochic view of the subgenre's characteristic act. This essay will attempt to steer a course between the Scylla of confessional overinvestment and the Charybdis of widespread critical underattention to religion7 to argue that while the basic problematics of the genre are indeed fundamentally theological, they are so in ways that go well beyond simply asserting the contemporary force or impotence of Christian prohibitions of revenge, or recycling the polemical oppositions of the Reformation. The question is not whether a given play is for or against revenge (most are both), or whether it loves or hates Catholicism (it might do either, or neither, or both), but rather what compellingly unresolved conflicts underlie its interest.8 I will suggest that those motive issues might better be sought, and the plays more deeply and complexly understood, in perennial and foundational problems of agency-which Luther himself identified as the central issue of the Reformation9-than in delimited positions on confessional identity or revenge itself. Indeed, there is more at stake in revenge tragedy than just the nature or status of revenge; the genre's larger concern is the nature, status, conditions, and consequent ethics of human action itself, and of the agential subject that performs it. Revenge often functions in this genre as a condensed emblem of human agency in all its troubling glory, and is used to highlight its underlying theological problems. Only rarely does a play consistently declare an absolute position for or against revenge, and only rarely is it treated wholly as a reprehensible act against the divine or as a heroic assertion of its own autonomous necessity. Rather, revenge tragedies exist in a continuum between these poles, and articulate complex dynamics of agency that negotiate between oppositional, cooperative, and independent accounts of human and divine action in ways that are deeply (if not always explicitly) theological.

All of these models necessarily posit some relationship (or lack thereof) between the earthly and transcendent, but it is not clear that this link is best conceived as a mechanism of effectiveness or fairness. Linda Woodbridge is at least partially right to observe that 'radical Christians, like revengers, quested after fairness,'10 revenge being dependent on equity and reciprocity, but for Paul and his heirs, fairness is the last thing a sinner should want; nothing we can do can atone for our guilt, and our best hope is unmerited grace. Hamlet understands this: 'use every man after his desert, and who shall scape whipping?' (2.2.529-30).11 Indeed, while revenge tragedy offers a tantalizing prospect of action and equity, tragedy in its purest forms, from Oedipus to Lear to Stoppard's Rosencrantz and Guildenstern, is often about that which cannot be (though perhaps it could have been) helped: injustice, suffering, alienation, death, plots already written without our consent. In this sense it resembles the tradition from Paul to Augustine to Calvin, and between that inevitability and our ceaseless efforts to control or subvert it (that is, the tradition from James to Pelagius to Arminius and beyond) lies the poignance of the human condition outlined by theologians and tragedians alike. And this further suggests that revenge tragedy is about more than just revenge; it is about the problematic conditions, limitations, and responsibilities-in short, the ethics-of human action.

Thomas Kyd's great play The Spanish Tragedy-the play that sparked the revenge craze in early modern England-is at first glance surprisingly non-sectarian and indeed even non-Christian. It is framed in an explicitly classical and pagan context; most of its references to an unmistakably Christian God are in the non-Kydian 1602 additions; no mention at all is made of 'Jesus' or 'Christ'; and most of its remaining references to the divine are either oaths or relatively neutral appeals to 'gods' and 'heavens'. Beyond the overtly political anti-Spanish dimensions that one would expect in the Armada years, the play in fact seems to deliberately sidestep direct involvement in the theological problems of Kyd's time. But Kyd either could not or simply did not altogether avoid those problems; in its displaced and masked way, The Spanish Tragedy is (by way of its interrogation of equity, action, and ethics) very deeply about contemporary theological issues of agency. In the end, I will argue that careful attention to both the play's main action and its pagan frame can teach us much about the theological nature and stakes of the play's agential concerns.

The long monologue of Andrea's ghost that begins the play is, despite being set in a thoroughly pagan context, riddled with theological tensions. In one sense these initially appear to be about identity (is he a lover or a fighter? where does he belong?), but in a larger sense they are about action and its relation to destiny (what kind of eternity have his earthly actions earned him?), here frozen into multiple forms of suspension. Charon at first refuses to even ferry Andrea's soul into Hades until he is properly buried; once that is done, the official judges of the underworld seem flummoxed as to where to file him, and send him on (carefully threading between the torments of hell and the stimulating fields of love and battle) to the boss for disposal, but even Pluto defers his judgment onto Proserpine, who in fact renders no decision but sends Andrea to watch a play with a new friend.

But this is neither the end of Andrea's predicament, nor its beginning. As he explains to his companion Revenge and us at the play's start:

When this eternal substance of my soul

Did live imprisoned in my wanton flesh,

Each in their function serving other's need,

I was a courtier in the Spanish court.

My name was Don Andrea, my descent,

Though not ignoble, yet inferior far

To gracious fortunes of my tender youth.

For there in prime and pride of all my years,

By duteous service and deserving love,

In secret I possessed a worthy dame,

Which hight sweet Bel-imperia by name.

But in the harvest of my summer joys,

Death's winter nipped the blossoms of my bliss,

Forcing divorce betwixt my love and me.

For in the late conflict with Portingale

My valour drew me into danger's mouth,

Till life to death made passage through my wounds.12

Note all the tense oppositions that Andrea has been subjected to: soul and body, death and life, winter and summer, love and battle, the self as active subject and passive object. Three features of the very first section of this speech exemplify and deepen these conflicts. First is the troubled relationship of soul and body in the play's opening lines, eternal substance imprisoned in wanton flesh, but the implicit pietude of this tension (which conventionally privileges the soul over its refractory incarnation) does not result in any visible sense of liberation or transcendence when the soul is freed; what we get instead after this speech is a hollow sense of spectral lostness in a pagan underworld, followed increasingly by a hunger for retaliation. These initial pressures adumbrate a central conflict between the expected Christian and providentialist understanding of wrong and transcendence on the one hand, and on the other a less sanguine but more bloody, activist and material impulse toward dissatisfaction and action that appears more immediately compelling.

Two further complications extend this basic problematic in interesting directions when Andrea's ghost posits a hierarchy between his 'descent' (that is, that inheritance over which he had no control) and the 'gracious fortunes of my tender youth'. His middling, non-royal but 'not ignoble' lineage is easy enough to understand, but what are we to make of the 'gracious fortunes' that outweigh its determinist influence? Do we take 'fortunes' to indicate the capricious favours of the blind goddess, and 'gracious' to intensify its arbitrariness by evoking, in a theologically-inflected term, the gifty gratuitousness of those favours? If so, then Andrea's micro-autobiography reads like a determined complex of effects, a shaped object of various external agencies. Or do we take 'fortunes' in a more Baconian sense, as things one makes for oneself,13 and 'gracious' simply as an indirect descriptor of how one does so? Such a reading obviously makes Andrea more of a self-made man, with a heightened form of agency that trumps the immutable circumstances of his birth. Perhaps the third pairing, of 'duteous service' and 'deserving love', will provide us with some direction. These appear to be the two conditions for Andrea's 'possession' of Bel-Imperia (which may be the 'gracious fortune' of the preceding lines), but here again, we run into some deep lexical and logical ambiguities, for these two terms are quite complicated. 'Duteous service' certainly implies obligation and subordination; the fulfilment of that obligation might be seen as meritorious (incurring some kind of debt owed to the server), or simply as a recognition of subordination and duty done (not creating any debt or inequity either way). 'Deserving love' is even trickier: is 'deserving' a participial adjective, grammatically aligned with 'duteous', or a participial verb, describing a state of earned desert between the service and the possessing? That is, is the love what he rendered deservedly to Bel-Imperia, or what she owed him?

These syntactical questions are not neatly resolvable, and they have significant consequences. Their accumulation generates a sense of a character in radical crisis, whose identity, place, destiny, mode of being, and capacity for action are in serious question. What we can particularly discern after attending closely to the play's first eleven lines is a foundational uncertainty about body and soul, action and quietism, merit and grace-in short, whether humans earn what they get, and get what they deserve, or whether they simply get what they get by a logic inscrutable and arbitrary.

This ambiguity turns out to be an important structuring principle of the play. Revenge proffers the symmetrical pleasure of his trademark mode at the end of that first scene, but the following scene reminds us that the questions are not that easy. In it, the captured Balthazar says stoically that 'cards once dealt, it boots not ask why so' (1.2.140) as he submits to his defeat and captivity. The King of Spain responds,

Meanwhile live thou, though not in liberty,

Yet free from bearing any servile yoke;

For in our hearing thy deserts were great,

And in our sight thyself art gracious. (1.2.147-50)

Of course, the King is not talking theology here, but he (and the play generally) is using its vocabulary, and recapitulating its tensions in a different context. Balthazar, like Andrea, exists neither in liberty nor in bondage, but somewhere in between. He has earned something, but it is not entirely clear what, and it is also unclear whether the King's seeing him as 'gracious'-i.e., possessing grace-is what was earned, or if the King (who has already asserted and foregone his right to reciprocate Portugal's transgression with 'evil measure' [136]) is simply imputing it to him. When Balthazar responds, 'And I shall study to deserve this grace', he simultaneously invokes the logic of deserving, and makes clear that that deserving is posterior to grace, a response to it rather than its precondition.

In the remainder of this scene, the squabble over whose prisoner he is presents another version of the issues at stake in this play. Balthazar surrendered to Horatio because of what he had done, and to Lorenzo because of who he was: 'To [Lorenzo] in courtesy, to [Horatio] perforce' (162), acknowledging the competing and differing logics of choice and necessity, rank and action. Concerned about the false equity implied in this formulation, Hieronimo-who, the play never lets us forget, is in charge of justice-interjects on his son's behalf that 'Enforced by nature and by law of arms / My tongue should plead for young Horatio's right. / He hunted well that was a lion's death, / Not he that in a garment wore his skin' (168-71). Simply putting something on isn't good enough, and is in fact a kind of imposterism unless one has earned it through action. After reassuring Hieronimo of his impartiality, the King asks the disputants, 'Will both abide the censure of my doom?' and Lorenzo promptly responds, 'I crave no better than your grace awards.' It is Horatio, our temporary protagonist, whose response should be more troubling to attentive English Protestants: 'Nor I, although I sit beside my right'. The king, unfortunately, will attempt to Solomonically split the difference, thus fully satisfying neither Horatio's interest in desert and equity, nor Lorenzo's in entitlement by way of arbitrary dispensation.

The first-act interest that Hieronimo and his son demonstrate in rights of merited reward, and their corresponding scepticism regarding gratuitous bestowal, will play a crucial role in what happens after Horatio's murder. This commitment to equity (a word derived from the Latin aequus, meaning 'even' or 'fair', and thus a word that has balance, reciprocity, and symmetry built into its notion of right) is crucial to Hieronimo's remarkable suitability to the administration of justice: we are assured by a citizen that 'There's not any advocate in Spain / That can prevail, or will take half the pain / That he will, in pursuit of equity' (3.13.52-4). But this 'pursuit of equity' can lead to much darker places. It is undoubtedly what makes Hieronimo such a good executor of state justice, and in this role he performs what was widely considered to be the ordinary form of divine vengeance: the everyday working of God-ordained institutions of justice. Hieronimo's slide from this mode into the much dicier one of personal violence is in many accounts occasioned by his supposed inability to achieve justice institutionally, and by his resulting sense of irremediable inequity that requires him to colour outside of the lines. Both the disappointment and the improvisation are powered by a similar principle of symmetrical justice, but this should not obscure the fact that the changeover reveals important things about him. All of his appeals to heaven are appeals for justice-present justice, which is important, and which I will return to shortly-and when those don't pan out, he shows himself quite willing to take his business elsewhere and look for its evil twin in the other place ('I'll make a pickaxe of my poniard, / And here surrender up my marshalship: / For I'll go marshal up the fiends in hell, / To be avenged on you all for this' [3.12.75-8]). In other words, while equity underwrites state justice, it can also trump metaphysical affiliation, and indeed override right and wrong altogether. This change of offices appears to occasion none of the doubt or metaphysical distress that we see in Hamlet; Hieronimo's traumas appear to be limited to being bereaved of a son and unable to do his job, and in an inverted sort of way this is a religious problem, a lack of reflection that implies a critique or rejection of metaphysics.

Indeed, what emerges from this analysis is a recognition that Hieronimo's approach to justice and action has been solipsistic from the start. In an influential 1931 essay, Lily Campbell enumerated many Elizabethan articulations of the Christian anti-revenge theme to argue that it was the dominant view of this culture, and emphasized the importance of the principle 'that God's vengeance may be delayed but is nevertheless sure'.14 This problem of delay or suspension is central, because it raises questions about the certainty of God's justice, and enables us to see something useful about the Christian problem with private revenge: it implies a lack of faith in God. God's vengeance could operate through a large variety of channels in this world, but hell is its final guarantor, and God might perfectly well reserve his vengeance for the afterlife in any given case. Seen in this light, the revenger's unilateral insistence on present, worldly justice would seem to clearly imply not just impatience but a more comprehensive distrust of divine oversight, and a seizure of control through direct action-which is to say that it implicitly and inherently takes a rather clear position on the theological version of the agency question.15 One way around this is traditionally to understand oneself as an ordained instrument of vengeance, as Hamlet will a decade later, but in that case how does one know one's vindictive calling is real, and not a self-justifying delusion? Answering this might involve looking for evidence of heavenly leading or alignment, or considering whether the revenge is contrary to one's own will. But while Hieronimo's rhetoric occasionally hints at a belief in divine justice (see, for example, 3.7.56), and even a confederation with it (in 4.1 especially), he acts most often like heaven's judge, not its instrument; any alignment is on his terms, and he steps in to supplement heaven's imperfect activity in this life.

Actually, to say even that much about Hieronimo's positioning is a bit misleading. His sense of revenge as a response to state and divine failures to act is something that develops only over time in the play, and problematically; what is constant from the start is his desire for personal violence. When he discovers Horatio's dead body in the second act, Hieronimo's thoughts, after his initial outburst of grief, turn immediately to revenge-not divine or state justice, but bloody individual satisfaction.16

To know the author were some ease of grief,

For in revenge my heart would find relief...

See'st thou this handkercher besmeared with blood?

It shall not from me till I take revenge.

See'st thou those wounds that yet are bleeding fresh?

I'll not entomb them till I have revenged.

Then will I joy amidst my discontent,

Till then my sorrow never shall be spent. (2.5.40-1, 51-6)17

This is not a patient waiting for God, or an eager anticipation of the sure working of state justice; it is an almost instant hunger for first-person retribution as the only means to 'joy'. It is Isabella that has to remind her husband that 'The heavens are just, murder cannot be hid: / Time is the author both of truth and right' (57-8),18 but as his following Latin speech makes clear, he sees time less as something in which one waits for God's unquestionable justice than as something needed to act in. He first discusses divine justice, and proposes a prospective judgment of it, only in the third act.

O sacred heavens! If this unhallowed deed,

If this inhuman and barbarous attempt,

If this incomparable murder thus

Of mine, but now no more my son,

Shall unrevealed and unrevenged pass,

How should we term your dealings to be just,

If you unjustly deal with those that in your justice trust?

(3.2.5-11)

But it is important to note that, as the multiple conditional 'ifs' attest, neither heaven nor the state have even remotely failed him at this point-and in fact, the subsequent arrival of Bel-Imperia's letter from above might well be seen as providential in direction, timing, and content, identifying Horatio's murderers as if in direct response to Hieronimo's prayer. Of course, the letter exhorts him to 'revenge thyself', without any reference to institutional or transcendent forms of justice, but lines 10-11 suggest that those have already been dismissed and this conclusion has already been arrived at.

Seen in this light, Hieronimo's subsequent complaints contain, alongside the pathos, more insistent self-validation than is often recognized. He laments that 'unjustly we, / For all our wrongs, can compass no redress... / only I to all men just must be, / And neither gods nor men be just to me,' and doubts that 'by justice of the heavens' he will ever 'know the cause that may my cares allay' (3.6.1-10). But if that 'cause' is the identity of the murderers, he has already been given that in the first letter, and any residual doubts about that will be erased by Pedringano's corroborating epistle in the following scene-a letter whose delivery, like the first, occurs as if in answer to prayer, thus undercutting his claims of heavenly apathy (3.7.10-18). It is not until the end of that scene, hundreds of lines after Horatio's murder, and after two arguably answered prayers, that Hieronimo finally decides to 'go plain me to my lord the king, / And cry aloud for justice through the court' (3.7.69-70). When he does so in 3.12, he is thwarted with such ease by Lorenzo that more sceptical critical scrutiny is required; if Hieronimo has the needed status and respect and evidence to press his urgent claim to the King, why does he not do so more effectively, and why does he give up so readily? Most accounts of the play either ignore this problem or attempt to explain it away by blaming circumstances, or class, or madness, or the King.19 But by this point, having witnessed his initial (and still unabated and central) impulse for personal bloodshed, lackadaisical pursuit of justice through established channels, and disregard for what may be divine assistance, one might be forgiven for wondering whether he is simply undertaking a pro forma exercise in bureaucratic base-covering to clear the way for the bloody individual action he has been intending for some time.20

In a sense, though, it may not matter much whether Hieronimo's appeals are sincere or merely help rationalize his resolve, and the availability of both readings illuminates some important things about the dynamics of agency in this play. What seems most important is that he register alienation from both institutional and divine justice, and see them both as failed, in order to validate his resolve for personal revenge; he has been doing this, more on presumption than on evidence, at least since 3.2. But he also appears to need to experience some degree of identification with divine justice, even if as its judge and executioner (he has, after all, already concluded that Lorenzo and company 'did what heaven unpunished would not leave' [3.7.56]). Witness his famous speech in 3.13, which follows his first cryptic appeals to the king:

Vindicta mihi!

Ay, heaven will be revenged of every ill,

Nor will they suffer murder unrepaid:

Then stay, Hieronimo, attend their will,

For mortal men may not appoint their time. (3.13.1-5)

It would seem that one could hardly ask for a more direct or conventional articulation of the Christian prohibition of violence. But even here there are problems: by leaving out the crucial 'dicit Dominus' from his opening quotation of Romans 12:19, he ends up speaking in an ambiguous first person that could equally be ventriloquizing God's prerogative or staking his own.21 And in any case, the entirely orthodox sentiments of lines 2-5 give way abruptly, apparently triggered by his reading of Seneca, to an emphatic reaffirmation of his resolve to what he will soon call 'sweet revenge' (3.13.107). Hieronimo in fact never seems to very seriously consider using his position to effect institutional justice, and sees himself officially only as an instrument of a failed (though actually largely untried) system. Similarly, he complains a great deal about heaven's indifference, regardless of potential evidence to the contrary, because he has to in order to feel unheard. These senses of alienation and failure, that is, are required for him to feel legitimately vengeful, and for him to be able to effectively identify his own bloody desires and actions as the will of heaven; he has to understand the state and the divine as both negligent and commissioning (even if in a mode of default) in order to understand himself as legitimately qualified for satisfying individual action. 'I, heaven' might be read as a submissive affirmation, or as a coup. The semi-coherent nature of this speech-which claims to 'conclude, I will revenge his death!' (20), but logically does not conclude so much as lurch toward-might argue for the latter. But even here, what he seems to ramble toward is a sort of fatalist activism that prompts individual initiative under the unhelpful but usefully vague aegis of 'destiny' (fata). Hieronimo occasionally attempts to harmonize transcendent and personal agency, but whether these struggles are fully genuine or, as I have suggested, more of an exercise in self-legitimation, he doesn't exactly obsess over them, and he always in practice privileges his own desires. Though he may at times talk like a compatibilist or even a determinist, he quite consistently acts like a libertarian.

After this pivotal scene, Hieronimo becomes fully decisive, definitively declaring the exile of justice and craftily pursuing the bloodshed he has long ago resolved upon. Only in one scene of the final act does he revisit his relation to divine agency, and that is done to quite bizarrely interpret Bel-Imperia's decidedly non-theological scolding as a sign of heavenly endorsement: 'Why then, I see that heaven applies our drift/ And all the saints do sit soliciting / For vengeance on those cursed murderers' (4.1.32-4). After all the almost unbearable ironies of the play rehearsal with his victims, he smugly concludes, 'Now shall I see the fall of Babylon, / Wrought by the heavens in this confusion' (4.1.195-6). But the desire and the working are his, not heaven's, and neither the playlet nor its aftermath indicate any explicit concern with divine justice or vindication among the litter of corpses. Hieronimo's bloodcurdling speech of explanation is all about his identity, his loss, his grief, his vow, his bloody and satisfied heart, his appointing and determining power as playwright and agent, 'Author and actor in this tragedy, / Bearing his latest fortune in his fist' (4.4.147-8) rather than the other way around. When he does mention heavenly vengeance on the slaughtered Lorenzo and Balthazar, 'upon whose souls may heav'ns be yet avenged / With greater far than these afflictions' (174-5), his point seems to be the distinction between the two forms of revenge rather than their coordination. Divine payback may, he hopes, reiterate and amplify his own, but they appear to be independent functions, and if he really had full faith in the former, why would the latter be necessary except as self-gratification?

These agential conflicts are echoed by several complex analogues in The Spanish Tragedy, which speak indirectly to the central dynamics I have sketched out. One is Bel-Imperia's suicidal off-script improvisation, contrary to the ending that Hieronimo had planned to spare her (see 4.4.135-45); in doing so she in effect refuses authorial grace and self-destructs in a supreme act of self-determination, thus leaving the godlike author's will effective to kill but ineffective to save. Another is the box which Pedringano believes to contain a letter of pardon for his crime. But the promised pardon is a sham that will send him to his doom with more guilt and swaggering confidence, and he, more opportunistic than bright, plays the part fully, confessing, wisecracking, and trusting Lorenzo until the moment his neck snaps. Kerrigan aptly compares Pedringano to 'an overconfident Calvinist' (179), but does not explicate the ironies of the box's almost complete inversion and perversion of Calvin's doctrine of grace: it represents a 'pardon' which has been earned, not bestowed, and not by righteousness or atonement but by sin and crime; which by its promise fills its recipient with gleeful assurance and joy in his misdeeds, thereby preventing him from taking any action that might save himself; and which proves in its nonexistence to be not only a con but the letter that kills. Here is an assurance that is problematic indeed, a box of death fatally misunderstood by a fool and criminal as a promise of life.22

I have, much like the plot of The Spanish Tragedy (the dislocation of which is one of the play's major critical problems), appeared to forget about the plight of Andrea with which I began. But of course, Pedringano is not the only bustler who is unaware that he is simply playing a pre-scripted role, and a tragic one at that; Revenge has in the play's first scene assured Andrea that he will see his lover murder his killer (1.1.86-9). Their periodic reappearance, while perhaps ironically funny, is a serious reminder of something very important: the actors in this play may think of and conduct themselves as fully autonomous agents, but the frame suggests that their apparently self-generated actions are subject to the authorial and directorial will of at least one supernatural agent within the world of the play (to say nothing of the text or of Kyd and whatever forces were acting in turn upon him). Andrea's rising dismay and panic at seeing his friends suffer and enemies prosper is met by Revenge's imperturbable calm-a sanguinity that evidently builds to fairly extensive napping. Upon being awakened after the third act, he reassures his impatient protégé:

Nor dies Revenge although he sleep awhile,

For in unquiet, quietness is feigned,

And slumbering is a common worldly wile.

Behold, Andrea, for an instance how

Revenge hath slept, and then imagine thou

What 'tis to be subject to destiny. (3.15.23-8)

What follows is, significantly, a prophetic allegorical dumb-show; Hieronimo is neither the only nor the greatest playwright in Kyd's play, and while Bel-Imperia may have improvised her way out of his script, she (and Balthazar) will not escape the immutable promptbook of Revenge.

The play's demonstrative insistence on 'destiny', however, does not resolve the problem of human agency by annihilating it. Revenge is an irresistible playwright, but a sketchy one. When we hear him promise Andrea in the opening scene that 'thou shalt see the author of thy death, / Don Balthazar, the prince of Portingale, / Deprived of life by Bel-Imperia' (1.1.87-9), we have already reached the limits of what we can confidently know about Revenge's absolute foreordination;23 he has determined a single specified outcome of the story he has been just been told. He later reassures Andrea that 'I'll turn their friendship into fell despite' (1.5.6), but since there has been no indication that this was part of the initial plan, it comes offas a bit of ad hoc plotting in response to human actions, an effort to direct and subordinate them to the destined goal, a promise of recuperation and endgame rather than part of a fully pre-scripted plan.

Such improvisational subordination of unpredicted human action to that guaranteed and single-minded master plot is (in its satisfaction of questions the frame seems designed to provoke) one of the chief structural pleasures of the play, but it is also the source of some of its most notorious critical problems. Critics have long noted, and fretted over, the shakiness of the originary claim to vengeance; the disjuncture of the play's shiftin focus from Bel-Imperia's revenge for Andrea to Hieronimo's revenge for Horatio; the odd indirectness and incoherence of Hieronimo's pursuit of conventional justice; and the wild excess of revenge in the final act, including the needless murder of Castile, who has apparently done nothing wrong. Typically, critics who focus on these problems have either lamented them as defects, or attempted to integrate them into some larger pattern of structural or thematic consistency (often having to do with the nature of justice or violence, or the corruptness of the Iberian peninsula).24 But perhaps we might learn something by resisting, even if temporarily, the impulses to dismiss or smooth out, and attending instead to what these difficulties might show us. One thing they might productively indicate is the complexity of the play's treatment of the conflicting imperatives of revenge. After all the prior dislocations have undermined the moral clarity of cleanly symmetrical and necessary revenge, Andrea's slavering bloodlust at the play's end prompts ethical and spectatorial reflection when he indiscriminately celebrates the slaughter of foes and friends alike as 'spectacles to please my soul' (4.5.12). He then doubles down on the Castile question by wishing on him the eternal torments of Tityus; Revenge in turn affirms this by including Castile-who, let us remember, is guilty of little more than questioning his daughter's choice of boyfriend-in the group headed for 'endless tragedy'. This is disturbing, but these things have, of course, also pleased and satisfied centuries of audiences and readers, and fulfilled their desires for action and closure and justice even as they provoke them to examine their complicity in the ambivalent and troubling dynamics of revenge.

Even more importantly, these disjunctions can illuminate how the play is working out the problems of agency that I have suggested are fundamental to it, and to revenge tragedy generally. The Spanish Tragedy begins with a backstory and a single, supernaturally-ordained end; what comes between the two, the Aristotelian middle, must follow that beginning and end in that end, but the means of that development are haphazard and, so far as we can tell, incorporated after the fact rather than predestined. We never, for example, see Revenge directly intervening into the play's action, and this suggests that Revenge's authorial control is real but not comprehensive; it involves coordinating and subordinating disorderly actions taken by other agents who are not under his power in every detail. And if this is correct, it follows that even if some of them (actually only one: Balthazar) are rigorously subject to a predefined end, the characters in the play have at least an epicyclical sort of agency within the plot, a freedom to act as they choose even if those actions may be pre-structured or put to unforeseen uses.25 This is evident in the motif of characters resisting or exceeding or escaping scripts written or demanded by others, which recurs repeatedly in the play, and takes on particular intensity in the final act with the ironic murder of those who submit to a script they don't understand, Bel-Imperia's script-violating suicide, the expansion of Revenge's hitlist, Hieronimo's refusal (after voluntarily explaining everything) to explain anything on command. Even G.K. Hunter, one of the best readers of the play and its genre, adopts a needlessly totalizing view of these dynamics in his essay on 'ironies of justice':

[T]he characters of the play...are not to be taken by the audience as the independent and self-willed individuals they suppose themselves to be, but in fact only as the puppets of a predetermined and omnicompetent justice that they (the characters) cannot see and never really understand. But we (watching the whole stage) must never lose sight of this piece of knowledge.... Hieronimo, for all his devotion to the cause of justice, is as much a puppet of the play's divine system of recompense as are the other characters in the action. He is stuck on the ironic pin of his own ignorance; we watch his struggles to keep the action at a legal and human level with involvement, with sympathy, but with assurance of their predestinate failure.26

This suggests a much more thorough sort of determinism than I believe the play actually exhibits. Such a formulation will have difficulty accounting for even unscripted or unexpected human action as anything much more than puppetry, and even a critic as capable as Hunter falls into distortion and overreading when operating under its terms.

It seems clear, rather, that Hieronimo's plot, like Revenge's, and like Kyd's, is working with some rather refractory materials, and these difficulties are important to a proper understanding of the play. The instance of Kyd is particularly fascinating, since he had the option of smoothing these problems out, and didn't. Are we to conclude that he was semi-competent, or that these problems are part of the point? One needn't be committed to Kydian infallibility to find the first option relatively uninteresting. Indeed, Hunter himself seems in a later essay on Seneca and English tragedy to take the latter, and I think more accurate, view: 'the revenge that takes place is wholly determined by the natural emotions of the characters: Revenge is given no direct relationship to Hieronimo, the hero and revenger...Kyd creates, in short, a gap between the supernatural concern in the play, and the more central matter, the individual lives, which exhibit that unexplained capacity to make free valid choices which is essential to a Christian view of the world and to modern drama'.27

That the 'capacity to make free valid choices' is 'essential to a Christian view of the world' (or to a scripted drama) is not self-evident, and in fact runs counter to a considerable amount of Christian theology, but Hunter's Seneca essay is right about this: the dynamics of The Spanish Tragedy are not the utterly determinist nightmare that he described in his earlier essay. Their divine intentions are determinist only in a supervisory, recuperative way that subordinates free actions to its ends-what we might call a determinism of ends and not means. We know the play's world is determinist because Revenge's first-act promise that Andrea will watch Bel-Imperia murder Balthazar is in fact fulfilled in the final act, and Revenge's breezily torpid placidity in the interim assures us that that fulfilment is never really in doubt. But we also strongly suspect that this determinism is not comprehensive, because Revenge never discusses his methods, never directly or visibly influences the characters' actions, and those actions are often emphatically erratic and unpredictable. And when he tells his protégé in 3.15 that 'though I sleep, / Yet is my mood soliciting their souls', far from being a ringing announcement of action, the import of 'my mood' is so nebulous that we can almost feel Revenge evanescing into his allegorical function as a representation of a deep human tendency. In this respect, he can seem less a puppet-master deity than a mythic distillation of the human.28

In the end, Kyd refuses to give us a comfortingly straightforward picture either way. His deities29 have effective foreordaining power, but not a total control over people and their actions, and this means that they have to improvise reactively and contingently in what I have called an agency of ends but not means. They're also indecisive, somnolent, and rather mean. On the other hand, Kyd's human characters appear to exercise real free will and agency, but those are subordinated to the larger plans of the gods, and thus people exercise an agency of means but not ends-or to put it another way, their ends are Revenge's means, and this relationship may not be comprehensible or even perceptible from the perspective of the human. The play, that is, resists the determinism of a Chrysippus, or Calvin, or Mulryne, or the earlier Hunter-but it equally resists the libertarianism of a Pelagius or a Hamilton (or the later Hunter). Its exclusion of these options leads us necessarily to some kind of dynamic compatibilism, and while I do not wish to exactly claim that Kyd was a Thomist, one might well see the play's structural dynamics of subordinated agency as resembling a blood-spattered and non-redemptive version of Aquinas, who argued that predestination

does not impose necessity of such a kind that its effect is realized through necessity... It does not exclude the freedom of the will, but realizes its effects contingently by means of it... [and this is possible because] Providence does not suppress secondary causes, but achieves its effects through subordinating their operation to itself.30

We might press this line of thought one step further in a play in which many of the major human characters are if anything more mean and manipulative than the gods. They plot and kill on slim, dubious, or no pretexts, and Hieronimo as much as Lorenzo tends to pursue his ends with violence and excess-even when, as in the former case, in service of an arguably good cause. Per scelus semper tutum est sceleribus iter: human action is portrayed as something that tends toward proliferative corruption (something that Kyd's contemporaries would have unblushingly labelled 'sin'), and this may suggest that the subordination of their purposes is in fact a positively good thing, at least insofar as it puts them to a use other than unrestrainedly widening circles of destruction and 'endless tragedy' on earth.31 Where this leads, eventually, is to the paradoxical conclusion that the divine will of Revenge, for all its consumptive violence, might actually play a beneficially limiting function on revenge in the hands of sinners-if not comprehensively and in practice, at least ultimately and in principle. In its allegorical function, Revenge embodies, prompts, and gratifies the deep human tendency to action and payback, but in its executive function it sets limits on that tendency, and these limits are both desirable and generative of meaning. An unkilled Lorenzo is an affront to our sense of justice and decency, a troubling emblem of unchecked human mischief; hence the value of having a Hieronimo to foreclose the horrifying possibility of unlimited and unpunished agency.32 But as commentators have always recognized, an unkilled Hieronimo then becomes in turn equally intolerable, particularly when there is no certain divine mandate for his excessive actions. Better, perhaps, to live under the teleological oversight of some transcendent agent-even one of dubious ethics, even one that may require our destruction-in relation to which humanity might understand itself and its actions as meaningful, not only because they are free, but also because they are bounded, subordinate, and accountable. Ethically responsible agency, like a liquid, and indeed like the blood in which tragedy trades, depends for its purpose and shape precisely on its containment.33 After all, even Revenge answers to somebody.

Footnote

1 Sam Harris, Free Will (New York: Free Press, 2012), pp. 5 and 11.

2 Ibid, pp. 43 and 39.

3 For overview collections of the philosophical literature, see Robert Kane (ed.), Free Will (Oxford: Blackwell, 2002); Derk Pereboom (ed.), Free Will (Indianapolis: Hackett, 2009); and Gary Watson (ed.), Free Will. Oxford: Oxford UP, 2003.

4 Harris, pp. 16 and 18.

5 William Hasker contends that 'the fullest and richest development' of the problematics of divine power and human freedom 'has occurred in the Christian theological tradition, beginning at least as early as Origen and reaching a climax in the debates of the sixteenth and seventeenth centuries'. See God, Time, and Knowledge (Ithaca: Cornell UP, 1989), p. 1. Hasker cites Anthony Kenny's assertion that 'nineteenth- and twentieth-century treatments of these matters have added very little to the work of earlier philosophers and theologians'. See The God of the Philosophers (Oxford: Oxford UP, 1979), p. 8.

6 See Lily B. Campbell, 'Theories of Revenge in Renaissance England' Modern Philology 28:3 (1931), 281-296; Fredson T. Bowers, Elizabethan Revenge Tragedy (Princeton: Princeton UP, 1940); G.K. Hunter, Dramatic Identities and Cultural Tradition: Studies in Shakespeare and His Contemporaries (New York: Barnes & Noble, 1978); Eleanor Prosser, Hamlet and Revenge (Stanford: Stanford UP, 1967); Ronald Broude, 'Vindicta filia temporis: Three English Forerunners of the Elizabethan Revenge Play', Journal of English and Germanic Philology 72:4 (1973), 489-502; Broude, 'Revenge and Revenge Tragedy in Renaissance England', Renaissance Quarterly 28:1 (1975), 38-58; Huston Diehl, Staging Reform, Reforming the Stage (Ithaca: Cornell UP, 1997); Robert N. Watson, 'Tragedy' in The Cambridge Companion to English Renaissance Drama, ed. by A.R. Braunmuller and Michael Hattaway (Cambridge: Cambridge UP, 1990), pp. 301-51; Thomas Rist, Revenge Tragedy and the Drama of Commemoration in Reforming England (Aldershot: Ashgate, 2008); Linda Woodbridge, English Revenge Drama (New York: Cambridge UP, 2010); John Kerrigan, Revenge Tragedy: Aeschylus to Armageddon (Oxford: Clarendon Press, 1996); Stephen Greenblatt, Hamlet in Purgatory (Princeton: Princeton UP, 2001).

7 Tanya Pollard's overview essay on revenge tragedy, for example, discusses politics, law, justice, emotions, Seneca, Aristotle, Aeschylus, Italian tragedy, foreignness, gender, violence, tragicomedy, and metatheatricality, but religion is conspicuously absent. See Pollard, 'Tragedy and Revenge' in The Cambridge Companion to Renaissance Tragedy, ed. by Emma Smith and Garrett A. Sullivan (New York: Cambridge UP, 2010), pp. 58-72.

8 The theological dominance of Calvinism (which firmly reasserted a general individual non-agency, both politically and soteriologically) in the Elizabethan Church clearly did not result in plays unremittingly characterized by passivity or inaction; on the contrary, secular drama was often a site for exploring or contesting stark theoretical formulations of many sorts.

9 In his De servo arbitrio, he commends Erasmus for correctly identifying 'the real issue, the essence of the matter in dispute...the question on which everything hinges'. See Luther and Erasmus: Free Will and Salvation, ed. by E. Gordon Rupp and Philip S. Watson (Philadelphia: Westminster Press, 1969), p. 333.

10 Woodbridge, p. 37. I would add that Erasmus and the Catholic tradition and non-radical Arminian Protestants did so too, in rejecting the more totalizing forms of determinism and insisting that humans' eternal fate must in some way involve and reflect their own choice and action.

11 So, in his way, does Clint Eastwood's Unforgiven character Will Munny, who, after being told by Gene Hackman's corrupt lawman Little Bill that 'I don't deserve [to be killed],' says bitterly, 'Deserve's got nothin' to do with it'-and then shoots him dead to avenge the murder of his friend. The fuller version of this exchange, which is of considerable interest, goes like this:

Little Bill Daggett: I don't deserve this... to die like this. I was building a house.

Will Munny: Deserve's got nothin' to do with it. [aims gun]

Little Bill Daggett: I'll see you in hell, William Munny.

Will Munny: Yeah. [fires]

12 1.1.1-17. All references will be taken from J.R. Mulryne's New Mermaids edition (London: A&C Black, 1989).

13 In his essay 'Of Fortune', Bacon famously writes that 'chiefly, the mould of a man's fortune is in his own hands. Faber quisque fortunæ suæ [Every one is the architect of his own fortune], saith the poet....All wise men, to decline the envy of their own virtues, use to ascribe them to Providence and Fortune; for so they may the better assume them: and, besides, it is greatness in a man to be the care of the higher powers.' See Francis Bacon, 'Of Fortune' in Essays, Civil and Moral (The Harvard Classics, 1909-14), <http://www.bartleby.com/3/1/40.html> [accessed: 21 November 2015].

14 Campbell, 285.

15 Lorna Hutson similarly claims that Hieronimo 'cannot imagine justice without his collaboration, justice as taking place after death'-though she is discussing a relatively late point in the play, whereas I will argue below that this trait is visible much earlier. See The Invention of Suspicion: Law and Mimesis in Shakespeare and Renaissance Drama (Oxford: Oxford UP, 2007), p. 284.

16 De Chickera and Broude, while well worth reading, take what is in my opinion an overoptimistic view of Hieronimo's relation to divine justice. See Ernst De Chickera, 'Divine Justice and Private Revenge in The Spanish Tragedy' The Modern Language Review 57.2 (1962), 228-32.

17 He reminds us of this constant resolution in 3.13.86-9.

18 She also articulates a notion of afterlife much more Christian than the rest of the play generally allows in 3.8.

19 The continuing prevalence of these assumptions can be seen, for example, in Michael Neill's introduction to a brand-new edition of the play, where he describes the King as 'coldly indifferent' to Hieronimo's plight (p. xxxi) even though the play itself does not offer much evidence for this, and the King is in fact quite solicitous and concerned: 'What accident hath happed Hieronimo? / ...Believe me, nephew, we are sorry for't / ...let him have his due / ... We shall increase his melancholy [if we force him out of office]. / Tis best that we see further in it first' (3.12.83, 90, 93, 99-100). See Thomas Kyd, The Spanish Tragedy, ed. by Michael Neill (New York: Norton, 2014).

20 Bowers insists that 'Hieronimo first endeavors to secure his legal rights before taking the law into his own hands' (p. 69), but this is only technically true, in the sense that he undertakes these endeavors before taking action. I've argued that the impulse, the will, to take the law into his own hands precedes his legal endeavors, and they themselves might well be seen as belated, feeble, and too easily thwarted. Hutson's account of Hieronimo's forensic procedure (pp. 277-287) is insightful and smart-she recognizes that Hieronimo 'implicitly assumes... heavenly agency' and 'has evidence and the ear of the king' (pp. 279, 284)-but even though she implicitly identifies the reasons why Hieronimo's appeal to the King should not have failed, she does not sufficiently interrogate that failure.

21 McAdam argues that 'the Reformation emphasis on an unmediated relationship with God could blur the distinction between the self and its divine other, in such a way that the true meaning of 'Vengeance is mine' becomes psychologically problematic, as it does for Hieronimo, who comes to see himself as divinely assisted' (35). See Ian McAdam, 'The Spanish Tragedy and the Politico-Religious Unconscious', Texas Studies in Language and Literature 42.1 (2000), 33-60 (p. 35). McAdam reads Protestantism primarily in terms of radical, indeed antinomian, individual authorization, though it could also work in precisely the opposite direction to decouple and polarize divine and individual agency.

Interestingly, though this is not the kind of fact one would want to build an argument on, in the 1592 edition, the second line continues with a first-personal pronoun that is always emended to 'Ay': 'I, heauen will be reuenged of euery ill.' This is grammatically hard to bear out except as, possibly, a clumsy sort of apposition, and is in all likelihood a case of simple phonetic equivalence, but it does gesture nicely at the dynamics under discussion here.

22 In a differently sceptical vein, Katharine Eisaman Maus says of this, 'The box that pretends to contain an authoritative, salvific text may be understood as a figure for the opaque, perjured subjectivity of the machiavel, but also, perhaps, as a comment upon the hollow promises of a Christianity The Spanish Tragedy both evokes and renounces'. See Inwardness and Theater in the English Renaissance (Chicago: University of Chicago Press, 1995), p. 66. The first reading is far more persuasive than the second, which is difficult to accept without either de- or hyper-ironizing the box. She goes on to suggest that 'If a beneficent providence does not exist, there is little hope for the redress of injustice in this world or the next; at the same time, divine punishment for self-assertion is less automatic and thus perhaps less fearsome. Instead of reassuring his audience with a theologico-theatrical fiction of beneficent omniscience, Kyd acquaints it with the disquieting possibility that it is caught in the same ironies that doom his characters, victims of powers that are not necessarily just or merciful, and whom they are incapable of understanding' (p. 71). This is provocative but contradictory: is the problem the nonexistence of providence-which would seem to leave considerable space for individual action and worldly redress-or its inscrutability and possible non-beneficence? Is Christianity here empty, or just fatally misappropriated?

23 This is, in Rumsfeldian terms, a 'known unknown': we know, or at any rate should realize, that we don't know the extensiveness of Revenge's plotting, the degree to which he has prewritten or even foreseen the many things we see unfold in the play.

24 Empson takes a novel tack in his two essays on the play, beginning with these legitimate critical concerns but addressing them with what I think can be fairly called wild surmise: postulations of censorship, overdetermined readings, and even the hypothesizing (and writing!) of a 'missing scene' which conveniently makes his theory work (58-9) but for which there is no clear evidence in the surviving text. See William Empson, Essays on Renaissance Literature, ed. by John Haffenden (Cambridge: Cambridge University Press, 1994) 2: pp. 58-9. Even Empson's highly sympathetic editor acknowledges that readers may find his 'creative interpolation' an 'utterly illegitimate critical trick' (p. 4).

25 Joel Altman, in a thoughtful and influential consideration of the play, offers a dramaturgical and epistemological solution to the 'apparent discrepancy between the determinism assumed initially and the autonomy suggested by the extended dianoia of the major characters' which is a useful model for my thinking about the play's ethical and agential dynamics. He suggests that The Spanish Tragedy 'invites us to respond in several different ways: to enjoy the visceral pleasures of a well-deserved blood revenge, to ponder more carefully the problematic situation of the man who would be just, and to reflect upon the simplistic judgment of the frame as an aspect of his problem. The first two affective imperatives are familiar enough from our experience with the digressive plays of Kyd's predecessors. But the last is original. Evidently sporting Kyd, assiduously studying English Seneca by candlelight, discovered what earlier playwrights had failed to see: that the contrasting perspectives of frame and action proper might be made to reflect ironically upon one another'. See Joel Altman, The Tudor Play of Mind (Berkeley: University of California Press, 1978) pp. 269 and 270-1.

26 Hunter, pp. 218 and 222-3.

27 Ibid, p. 195.

28 Donna Hamilton is quite right to describe Revenge at this moment as speaking of himself 'as a quality in men, not himself as a creature in control of men,' but takes this too far when she flatly asserts that 'Revenge does not direct or control anything' (206, 204). See 'The Spanish Tragedy: A Speaking Picture', English Literary Renaissance 4 (1974), 203-17 (pp. 206 and 204). Mulryne, meanwhile, analogously overstates the opposite case: 'We know that the play's outcome will be disastrous for anyone who opposes Andrea's revenge....all action takes place within a determined framework to which we, but not the actors, hold the key' (p. xxiii).

29 I am thinking mainly of Revenge here. Pluto has delegated responsibility for this situation, and Proserpine's sinister smile and whispered commission in 1.1 are so enigmatic that they provide no clarity on the problems addressed in this essay. Presumably she has told Revenge something, but we can have little firm idea of what beyond mandatory play-watching; again we can only guess at the degree to which divinity controls or foreordains human action-which may lead us to reasonably conclude that it is not total.

30 Summa Theologica I, Q. 23, Arts. 6, 8.

31 Michael Neill usefully observes that 'by pitting the nightmare of no end against the formal closure of the catastrophe, The Spanish Tragedy deliberately unsettles the consolations of tragic design, and the effect is to question the framework of cosmic and human order as Kyd's contemporaries were taught to understand it' (p. xxxvi). I agree with this, except insofar as Neill's concerns are largely sublunary and sociopolitical; my contention here is that the transcendent order, problematic as it may be, is what finally contains and orders earthly chaos.

32 Consider, for example, the disillusioning meditations on unrestrained agency we find in Plato's fable of Gyges' ring (Republic, II) and Hobbes' appalling state of nature (Leviathan I.13.13): both cases argue that ethics and civilization fundamentally require limitation on human action.

33 'Necessity is not only compatible with autonomy', writes American philosopher Harry Frankfurt (The Importance of What We Care About [Cambridge: Cambridge University Press, 1998], p. ix); 'it is in certain respects essential to it. There must be limits to our freedom if we are to have sufficient personal reality to exercise genuine autonomy at all. What has no boundaries has no shape'.