Verbal Magic and the Problem of the A and B Texts of Doctor Faustus

[First published in the Journal of English and Germanic Philology 82.3 (July 1983): 324-46.] In the present version, I have made several small corrections, one of them noted in a new footnote 6; the text has not otherwise been altered.]


Readers of Marlowe's Doctor Faustus find themselves, at present, in a very curious position. The play survived the Elizabethan age in two versions, which were first printed in 1604 (the “A-text”) and 1616 (the “B-text”). These two quartos, from which all subsequent editions are derived, offer variant readings in many parallel passages, and also differ very substantially from one another in the last three acts. While accepting many of the A-text's readings, recent editors have been unanimous in making the B-version of the play the basis of their editions. (The last major edition to use the A-version as its copy-text was that of Tucker Brooke, in 1910). There is now, however, a near-consensus among critics and editors alike that although the A-text shows clear signs of textual corruption, it presents the play in a form that is both aesthetically preferable to the B-version and more authentic, in the sense of being closer to what Marlowe actually wrote. What is anomalous about the present situation is that Fredson Bowers, the scholar whose arguments in 1973 established this opinion (thus overthrowing the view, dominant since the 1940's, that the B-text corresponded closely to Marlowe's intentions), at the same time based his edition of the play upon the 1616 quarto, and relegated the third and fourth acts of the A-text to an appendix. Readers are thus informed that the A-version is more authentic—and are given the B-version to read.1



My purpose in this article is to review the textual analyses of Doctor Faustus which have led us into this very peculiar situation, to present some new arguments which confirm the decisive critical superiority of the A-version of the play, and to demonstrate that in one important respect—its handling of verbal magic—the B-version is fundamentally incoherent. The great soliloquies of Acts One and Five (speeches which are preserved in substantially the same form in both versions of Doctor Faustus) show clearly that the reversal of Faustus' aspirations with respect to verbal magic is an essential art of the play's meaning. Faustus' gloating anticipation of near-omnipotence—“All things that moove betweene the quiet poles / Shalbe at my commaund” (A: 86-87)—finds a distorted echo at the end of the play in the desperate invocations with which he attempts to stave off the recognition of his utter helplessness:

Ah Faustus, 
Now hast thou but one bare hower to live, 
And then thou must be damnd perpetually: 
Stand stil you ever mooving spheres of heaven, 
That time may cease, and midnight never come: 
Faire natures eie, rise, rise againe....   (A: 1450-55)

In the A-version of the play a consistent authorial attitude toward this reversal is evident throughout. In the B-version, however—and in precisely those passages which have been identified, on other grounds, as later and non-Marlovian additions—a second, conflicting attitude toward verbal magic is developed, with the consequence that one of the central patterns of the play is disrupted. Close analysis will show that this disruption is part of a general relapse from the tragic ironies of the A-version in the direction of the more grotesque features of the Faustbook.



The relative merits of the two texts can be briefly summarized: the B-text, in parallel passages, is generally better and more authoritative—although since it was evidently subjected to a process of bowdlerizing, the A-readings are often to be preferred; the A-text, however, presents the play as a whole in a form that is both more authentic—that is, closer to its original form—and also aesthetically superior.

The arguments of Leo Kirschbaum and W.W. Greg2 have made it probable that in many passages where the two versions run parallel the A-text is a “bad quarto,” a reported text which lacks manuscript authority; and yet, as Fredson Bowers has remarked, this bad quarto is “unusual ... in the apparent closeness with which much of the verse of the tragic action is recalled.”3 In addition to the corruptions introduced by memorial transmission, the A-text shows in some places signs of revision—which may indicate that the text was adapted from its original state, at a time when the London theatres had been closed due to plague, in order to conform to the simpler conditions of provincial performance.4 The B-text, which is some six hundred lines longer, is at many points derived from the 1611 reprint of A (known as A3); there is evidence, however, that its compiler had access to the promptbook of the play (probably damaged or incomplete) which enabled him to correct many of the errors of the A-text.5 (At the same time, though, he cut or altered lines that struck him as being obscure or dangerously profane.)

The textual variants of the A- and B-texts are of course of great importance. Far more arresting, though, are the passages in which the two texts diverge completely from one another. The B-versions of the episodes in the papal court, the imperial court, and the court of the Duke of Vanholt are all very much longer than the corresponding passages in the A-text, and contain incidents and whole scenes of which the A-text gives no hint. The B-text contains, in the fifth act, an “infernal conclave” in which the devils assemble to witness Faustus' end, a sequence of farewell speeches to Faustus by Mephostophilis and the Good and Evil Angels, and also a final scene in which the scholars discover and comment on his “mangled limbs” (B: 2110): none of this is present in the A-text. Less significantly, the scenes of farcical clowning involving Robin and Rafe (or Dick, as he is called in the B-text) are only loosely parallel in the two versions of the play. In one instance, the A-text gives a passage that has evidently dropped out of the other text: the chorus to Act IV, though misplaced in A, is wholly missing from B. But in return, fourteen lines of the chorus to ACT III are known to us only from the B-text.

Editors like A.W. Ward and C.F. Tucker Brooke used the A-text6 of the play as their copy-text—for knowing that in 1602 Henslowe had paid the substantial sum of four pounds to Samuel Rowley and William Birde (or Borne) “for ther adicyones in doctor fostes,”7 they not unnaturally believed the B-text to embody these additions. But a shift in scholarly and critical opinion was initiated by the appearance in 1932 of F.S. Boas' edition, which was based upon the quarto of 1616. Dwelling upon the corrupt and disordered state of the A-text, Boas argued that the comic prose scenes in the 1616 version constituted part of the original text, and he suggested Rowley as their author—thus involving him in the initial composition of the play, as Marlowe's collaborator, as well as in the 1602 additions.8 This argument altered the status of two extended passages which had previously been taken to be additions (and by no means very interesting ones): the scene in which the clowns recount Faustus' grotesque tricks and plot revenge on him, and their intrusion into the scene at Vanholt, where Faustus, striking them dumb at last, could perhaps be said to earn the audience's gratitude more than its laughter. Boas also drew attention to external evidence—passages in the anonymous The Taming of a Shrew (printed in 1594) and in Shakespeare's The Merry Wives of Windsor (probably written in 1600 or 1601)—which seemed to make it difficult to assign the 1616 scenes in the papal and imperial courts and the abortive revenge of the “injurious knight” Benvolio (III.i-ii and IV.i-iv in Greg's parallel-text edition) to the additions of 1602.9

Boas did not attempt to draw conclusions from this evidence, and his arguments were largely disregarded by critics, who continued to show a strong preference for the A-version. Kirschbaum and Greg, who independently made the important discovery of the memorial nature of the A-text, were less cautious. They decided that the B-text as a whole must antedate the 1602 additions—which had presumably been lost. The 1616 quarto, then, provided a text very close to the intentions of Marlowe and his collaborators, whoever they may have been. And the 1604 quarto, being a reported text, possessed little or no authority except insofar as it preserved lines which had been tampered with by the editor who put the B-text through the press, or which perhaps represented earlier or later stages of the initial process of composition and revision.10 The immense skill and authoritative minuteness of Greg's arguments in particular made this the standard view of Marlowe's Doctor Faustus for nearly a quarter of a century. Editors and critics, by the mid-1960's, were beginning to express dissatisfaction with the resulting interpretations, but faced with a choice between an almost complete text dating from 1592-93 and a memorial abridgement dating, on Greg's estimate, from 1594-95, they confined themselves in most cases to mutinous mutterings.11

In 1973, however, Fredson Bowers demonstrated this analysis to be incorrect, and his arguments have since been supplemented by those of Constance Brown Kuriyama.12 The external evidence adduced by Greg and Kirschbaum proved, upon examination, to be surprisingly fragile. The passage in The Merry Wives of Windsor which was believed to show the existence before 1600 or 1601 of the attempted revenge of Benvolio (and consequently of the preceding B-text scenes in the papal and imperial courts) is in fact by no means a precise allusion. Moreover, this passage does not occur in the 1602 “bad quarto” of The Merry Wives—it is present only in the 1623 Folio text, which, as Bowers remarks, shows signs of having been revised.13 And the parallel between A Shrew and two lines from this same Benvolio episode—a parallel which in Greg's own words “cannot be taken by itself to prove anything at all”14—becomes almost meaningless in face of the probability that both passages were written by Samuel Rowley.15 This rather dubious evidence is outweighed by the fact that B: 1200, “He took his rouse with stopes of Rhennish wine,” seems to be derived from Hamlet, I.iv.8 ff.: “The King doth wake to night and takes his rowse. / Keepes wassell and the swaggring up-spring reeles: / And as he draines his drafts of Rennish downe, / The kettle drumme, and trumpet, thus bray out....” As Bowers writes, “This is, in fact, better evidence for date than the Merry Wives reference since it is a quotation and not a rather vague allusion of doubtful specific application. The only alternatives to line [B: 1200] as written in 1602 are to speculate that Shakespeare was drawing on his memory of Faustus or to place the I.iv lines in the Ur-Hamlet, both rather desperate expedients to avoid a straightforward conclusion.”16 Kuriyama's discussion of further parallels with A Shrew and with Robert Greene's A Looking-Glass for London and England (1591-92, printed 1594) makes a somewhat stronger case for the comparative originality of the comic prose scenes in the A-text than Boas, Kirschbaum, and Greg had been able to make for the contrary opinion.17

This external evidence is strongly supported by other indications. It is noteworthy that Greg's reconstruction of the play's textual history, which rests on his conviction that the B-text does not contain the 1602 additions, involves a large number of speculative assumptions, none of which are necessary if the unique B-text passages are identified with these additions.18 This complex structure of hypotheses creates some embarrassing problems. For example, to explain the presence of what appear to be Rowley's stylistic traits in most of the material peculiar to the B-text by the hypothesis that he collaborated with Marlowe in 1592 or 1593, and that the additions for which he was paid in 1602 are lost, is to substitute speculation for one of the few available pieces of historical information about the text of Doctor Faustus. And if this hypothesis were granted, it would become difficult to imagine what the 1602 additions—which earned their authors a sum equal to what Henslowe sometimes paid for new plays—could have consisted of. It is curious, moreover, that while the reported A-text is missing a scene (presumably a comic one) between the signing of Faustus' pact and the next scene, which begins “When I behold the heavens, then I repent” (A: 628), the supposedly original B-text has a gap in the same place. These problems disappear if one accepts the alternative hypothesis that “the more 'original' text is not original at all, but a revision and expansion of the 'debased' version, or of a text very similar to it.”19 As Bowers has argued in detail, the passages peculiar to the B-text appear to be superimposed upon the tragic action rather than integrated into it: editors like Bullen, Ward, and Tucker Brooke were right in thinking them revisions rather than part of a collaboration.20

The A-text, then, preserves the last three acts of Doctor Faustus in a form which, although probably corrupt and possibly abridged as well, is as close to Marlowe's original as we can expect to get. But how corrupt is the A-text? To what degree should its versions of the non-parallel scenes be allowed by critics and editors to supplant those of the B-text? Fredson Bowers has followed the path of compromise: while arguing that passages added in 1602 cannot legitimately be used to clarify the meaning of passages common to A and B, he took the 1616 quarto as the copy-text for his edition. Given the arguments of Greg that in many parallel passages the readings of the B-text have manuscript authority, this editorial decision can be defended. But it represents a sacrifice of critical integrity not so much to bibliographical principle as to speculation. For Bowers' belief (which in his view made acceptable the decision to base his text upon the B-version) that the A-text passages which were supplanted or cut in the 1602 revision were almost exclusively the work, not of Marlowe, but a collaborator, seems to me highly questionable.21 Judgments on delicate matters of this kind are notoriously difficult to sustain in the absence of such clear-cut stylistic divergences as those which distinguish the 1602 additions. And once the obvious displacements of the A-text's farcical scenes are remedied, its structural and aesthetic superiority over the B-text can be easily recognized.



The denigration of one or the other version of the play in terms of its more or less glaring textual inadequacies can degenerate rapidly into a form of literary-critical table-tennis. But when one deals with the last three acts only, and on the level of their continuity, tone, and decorum, the issue is more one-sided. Bowers and Kuriyama have noted several points at which the awkward splicing of the B-text additions into the text has resulted in repetition or inconsistencies.22 This fumbling is most evident at the beginning of the material relating to the anti-pope Bruno in Act III. The writer (probably Rowley),23 whose fondness for couplets is one distinguishing feature of his verse, tumbles immediately into confusion. The occasion of the papal celebration becomes, by the addition of a slovenly triplet, not only “holy Peters feast” (B: 855), as in A, but also “the Popes triumphant victory” (B: 858). This may be a minor incongruity; and yet there is further evidence of carelessness in the stuttering repetition of “this day” in these same lines: “... holy Peters feast, / The which this day with high solemnity, / This day is held through Rome and Italy” (B: 855-57). What follows, to avoid charges of partiality, is best described in Greg's words:

To Mephostophilis' invitation to wait and see the pope, Faustus replies with an irrelevant declaration that (862-63)

My foure and twenty years of liberty
I'le spend in pleasure and in daliance,

ending with another couplet. Very well, replies Mephostophilis, here they come. “Nay stay”, says Faustus, one thing “and then I go”—he has not been asked to go anywhere, but only to “stand by me”. And then amazingly he launches into a description of their cosmic flight, previously narrated by the Chorus, as a reason for being allowed to be an actor in the papal “shew”. So be it, replies the spirit, “but first stay”—for the third time—and “view their triumphs”, and whatever pranks you like to play, I will see to it that they take effect—and here really does come the procession.24

As Bowers observed, “It is difficult to argue that this series of contradictions and non sequiturs marked the original composition of the scene and that the simple and unified A-text action was its memorial reconstruction.”25 Some readers may find it equally difficult to see why an episode that begins in such an unpromising manner (the unintentional burlesque of this mindless cross-talk goes on for more than thirty lines) should be given preference by editors over the alternative in the A-text.

The remainder of the papal episode in the B-text is competently written, but displays a certain poverty of imagination, a willingness to make repeated use of the same material. Thus “Saxon Bruno,” the anti-pope elected by the emperor, is ordered to “stoope, / Whilst on thy backe his hollinesse ascends / Saint Peters Chaire and State Pontificall” (B: 896-98). These last two words represent a mannerism of this writer, a trick of inversion which he uses to add weight to his verse (and which helps to identify him as Samuel Rowley).26 “State Pontificall” is also padding—though in a sense functional, for it makes possible a further repetition: “Sound Trumpets then, for thus Saint Peters Heire, / From Bruno's backe, ascends Saint Peters Chaire” (B: 903-04). The writer evidently felt that he could get more mileage out of the passage from Foxe's Book of Martyrs that seems to have been his source,27 for some forty lines later the Pope returns to the subject of human footstools. He will depose the Emperor:

And as Pope Alexander our Progenitour, 
Trode on the neck of Germane Fredericke
Adding this golden sentence to our praise; 
That Peters heires should tread on Emperours, 
And walke upon the dreadfull Adders backe, 
Treading the Lyon, and the Dragon downe, 
And fearelesse spurne the killing Basiliske: 
So will we quell that haughty Schismatique....   (B: 945-52)

While such bombast may be appropriate to this Pope, the verse comes closer, perhaps, to the style of Ancient Pistol than to that of Marlowe.

A sense of rhetorical strain, of verse being stretched to the writer's capacity by material which could more suitably be handled at a lower pitch, is more evident still in the B-text scenes at the imperial court. Even the elaborate dumb-show which Faustus provides—a contrast to the simple stage directions of the A-text—is an inevitable anticlimax after his announcement that

if so your Grace be pleas'd, 
The Doctor stands prepar'd, by power of Art, 
To cast his Magicke charmes, that shall pierce through
The Ebon gates of ever-burning hell, 
And hale the stubborne Furies from their caves, 
To compasse whatsoere your grace commands.   (B: 1254-59)

This writing might best be described as opportunistic in its striving for local advantages and neglect of larger structural values: its rhetoric, aspiring in all directions to exaggerated effects, deadens any but the most superficial responses. The Emperor greets Faustus in hyperbolic terms: “Wonder of men, renown'd Magitian, / Thrice learned Faustus...” (B: 1237-38). And he responds with a display of boot-licking: “These gracious words, most royall Carolus, / Shall make poore Faustus to his utmost power, / Both love and serve the Germane Emperour, / And lay his life at holy Bruno's feet” (B: 1250-53)—which is followed by the rhodomontade that I have already quoted. The interruptions of the sceptical knight are reduced to the level of clowning: “...he looks as like [a] Conjurer as the Pope to a Costermonger” (B: 1261-62); “...zounds I could eate my selfe for anger, to thinke I have beene such an Asse all this while, to stand gaping after the divels Governor, and can see nothing” (B: 1275-78). And when he follows the Emperor's line, “Be it as Faustus please, we are content” (B: 1286), with “I, I, and I am content too...” (B: 1287), he drags the whole scene with him into farce. The words with which he vows retribution for his horning are at least amusing: “But an I be not reveng'd for this, would I might be turn'd to a gaping Oyster, and drink nothing but salt water” (B: 1365-66). And they come appropriately from this hung-over clown.

Much on the same level is Falstaff's threat: “and I have not Ballads made on [you] all, and sung to filthy tunes, let a Cup of Sacke be my poyson.”28 It is therefore incongruous, to say the least, when this self-proclaimed ass, Benvolio, appears in the next scene with high words and a sense of honour:

Away, you love me not, to urge me thus, 
Shall I let slip so great an injury, 
When every servile groome jeasts at my wrongs, 
And in their rusticke gambals proudly say, 
Benvolio's head was grac't with hornes to day? 
O may these eye-lids never close againe, 
Till with my sword I have that Conjurer slaine.   (B: 1375-81)

But the tendency toward farce seems irresistible. Benvolio and his accomplices, standing over Faustus' beheaded body, decide to nail horns on his head, sell his beard to a chimney-sweeper, and put out his eyes: “and they shall serve for buttons to his lips, to keepe his tongue from catching cold” (B: 1439-40). Faustus interrupts their contemplation of this “excellent policie” (B: 1441): “Zounds,” cries Benvolio, “the Divel's alive agen.” “Give him his head for Gods sake,” shouts one of his accomplices (B: 1443-44). Elements of farce, as Shakespeare knew, can heighten the effects of a tragedy. But here the farce is out of control, and seems to result primarily from the writer's inability, when he sees an opportunity for a comic quip, to resist putting it in.

Other aesthetic weaknesses of the B-text deserve mention: its repeated anticipations of the opening of Faustus' great speech to Helen,29 the abruptness with which its version of the horse-courser scene begins, the feebleness of the clown's last appearance,30 and the anticlimactic quality of the scholars' final speeches. The speeches of the spirits that precede Faustus' confession to the scholars and his last appearance are less obviously at odds with the tragic structure of the play. These added speeches have a definite energy; they emphasize and literalize Faustus' dread and his sense of loss. But their very conventionality detracts from the power of the play's conclusion. Hell, in particular, becomes disappointingly specific. After two impressive lines—“Now Faustus let thine eyes with horror stare / Into that vaste perpetuall torture-house” (B: 2018-19)—the Evil Angel offers an increasingly unimpressive sequence of torments which make one forcibly aware of the inappropriateness to the play of such a concretized material hell. To make these things so relentlessly—grotesquely—physical, or perhaps even to name them at all, is to rob them of much of their horror. The writer (to alter slightly the words of Shakespeare's Lafew) has made ancient, and familiar (and moralistic as well), “things supernaturall and causelesse”; he encourages us “[to] make trifles of terrours, ensconcing our selves into seeming knowledge, when we should submit our selves to an unknown feare.”31

With one notable exception—the Old Man's first speech, the B-version of which is in my opinion decisively superior—analysis confirms Greg's observation that “The passages peculiar to B are not in any way organic....”32 The B-text additions tend throughout to emphasize and augment the element of supernatural spectacle in the play, and the inflated rhetoric with which such spectacle is presented, in the fourth act especially, is clearly meant to inspire awe. But the detachment of this rhetoric from psychological plausibility33 and from dramatic consistency means that, in the third and fourth acts at least, the awe that is inspired is “safe.” Because the additions function only on the level of disconnected dramatic artifice, the audience can maintain a secure awareness of the insubstantiality of the show, and can indulge its appetite for marvels without discomfort. Marlowe's own writing in this play, however, has a very different tendency.



The A-text is also not without weaknesses. Two of its comic scenes are displaced, and the second of these (Scene ix in Greg's parallel-text edition) has two distinct endings which follow one after the other: sure evidence of corruption and revision. Moreover, the scenes in the imperial court, with the horse-courser, and with the Duke of Vanholt (which together make up Act IV) are linked by what Greg calls dramatic enjambement: Faustus remains on stage throughout, and the shifts in scene are marked by the exits and entrances of secondary characters. Even with these weaknesses—and some might argue that this dramatic enjambement (which is a prominent feature of The Jew of Malta, occurring in II.iii, IV.i, and V.i) is actually quite effective—the A-text is decisively superior to the B-version of the play in its “aesthetic integrity.” As Kuriyama (whose expression this is) has observed, the A-text employs different levels of verse and prose in a consistent and significant manner.34 This is more than a matter of stylistic decorum, and the structural importance of these differentiations extends beyond Kuriyama's confessedly schematic statement that prose is generally used “to lay bare the unpleasant facts which are glossed over and gilded by the verse.”35 It touches, in fact, one of the basic patterns of meaning in the play.

Realist theories of language were common in the Renaissance, and in the hands of Hermetists and Neoplatonists usually involved claims about the magical power of words. Marlowe's interest in the dramatic potential of such claims is evident in his first work for the public stage: the words of Tamburlaine are “strong enchantments”; his vaunts substantiate themselves, words becoming things.36 But while Tamburlaine's verbal magic has a “transitive”37 effect on other characters within the play-world, a different pattern is developed in the A-text of Doctor Faustus: Faustus' incantations may have a transitive effect upon the audience, yet within the play they work only on his own self. In the following pages, after identifying the kind of magical approach to language that Marlowe in this play quite deliberately inverts, I shall attempt to show how important a feature of the A-text this structure of inversion is, and how seriously it is disrupted by the B-text additions.

In Act I the speeches of Faustus and his two friends both imply belief in the magical power of language and convey a sense of that power. Even by casting his words into the despised form of syllogistic logic, Faustus has (so he says) achieved quasi-magical effects, making “the flowring pride of Wittenberg / [Swarme] to my Problemes, as th'infernall spirits / On sweet Musaeus when he came to hell” (B: 136-38). But he wants to literalize this figure of speech, to put his words into forms that will give them, and the will from which they spring, a more coercive power. The appearance of Mephostophilis in response to his invocation seems to fulfil this desire: “I see there's vertue in my heavenly words” (B: 255), he says—the key word in this sentence conveying, like the Italian virtù, the meaning of “power” of “efficacy,” and only secondarily, as an ironic pun, the modern sense of “virtue.” Mephostophilis, however, is not at his command, but is “a servant to great Lucifer” (A: 285): the “conjuring speeches” (A: 290) of Faustus were only per accidens the cause of his appearance.

The language in which Faustus and his friends anticipate the ability to translate will directly into power, as C.L. Barber noted, is insistently blasphemous: Valdes and Cornelius will make Faustus “blest with [their] sage conference” (B: 126), and Valdes tells Faustus that “these bookes, thy wit, and our experience, / shall make all ations to Canonize us” (B: 141-42). Barber associated this tendency to blasphemy with the tension in Elizabethan religious thought between a Protestant rejection of the idea that power inheres in the physical and gestural aspects or the verbal formulas of the sacraments, and a residual impulse to give independent meaning to these very aspects of worship.38 I would suggest that it can also be linked, along with Faustus' more general tendency to ascribe magical power to language, to the tradition of Hermetic magic.

The fourth and thirteenth tractates of the Corpus Hermeticum have been described by A.D. Nock as “mysteries of the word, sacramental acts of baptism and regeneration described and explained.”39 And as recent scholarship has shown, Christian Hermetists like Ludovico Lazzarelli and Cornelius Agrippa attempted to make these Hermetic patterns—the central one of regeneration especially—into mysteries of the Word, the Christian Logos.40 Or—from another point of view—they tried to appropriate the language and doctrine of the scriptures to express and justify their own theories of magic. The Logos, which is both the Word of the Scriptures and Christ, the co-eternal creator of the world, is central to Agrippa's view of the magical power of language. In De occulta philosophia (I quote for convenience from the 1651 translation) he advances the opinion of the “Platonists ... that the power of the thing as it were some kind of life, lies under the form of the signification. First conceived in the mind as it were through certain seeds of things, then by voices or words, as a birth brought forth, and lastly kept in writings. Hence Magicians say, that proper names of things are certain rayes of things, every where present at all times, keeping the power of things....”41 His explanation of this is constructed on the level of natural causes:

the vertues of this terrene orb proceed from no other cause then Celestiall. Hence the Magician that will worke by them, useth a cunning invocation of the superiors, with mysterious words, and a certain kind of ingenious speech, drawing the one to the other, yet by a naturall force through a certain mutuall agreement betwixt them, whereby things follow of their own accord, or sometimes are drawn unwillingly.... Now the words of men are certain naturall things; and because the parts of the world mutually draw one the other, therefore a Magician invocating by words, works by powers fitted to nature, by leading some by the love of one to the other, or drawing others by reason of the following of one after the other, or by repelling by reason of the enmity of one to the other, from the contrariety, and difference of things, and multitude of vertues; which although they are contrary, and different, yet perfect one part; sometimes also he compels things by way of authority, by the Celestiall vertue, because he is not a stranger to the heaven.

The magician's knowledge enables him to manipulate the occult forces in nature. He can also compel things because his words are in fact (though Agrippa seems to resist this conclusion) supernatural as well as natural. Celestial “vertues” originate with

the word of God, which word the wise Chaldeans of Babylon call the cause of causes, because from it are produced all beings, ... and that by reason of the union of this word with the first author, from whom all things are truely produced; The word therefore is the image of God, the acting intellect the image of the word, the soul is an image of this intellect; and our word is the image of the soul, by which it acts upon naturall things naturally, because nature is the work thereof.42

Agrippa's return to an insistence on the natural quality of verbal magic cannot disguise the obvious implication of this passage, which is that the magician can get in at the top of the hierarchical structure of the world because his “mysterious words” and “ingenious speech” draw upon the power of the Word of God: Christ, the Logos, the holy scriptures.

This syncretism was attacked by orthodox Christians because it applied, to this-worldly magic, language which owes its special qualities to the fact that it has been used by the Holy Spirit to express the sacramental links established by Christ between God and man. This appropriation of the Word was blasphemy, just as surely as the use of a stolen Host for magical purposes was blasphemy.

It is, I think, to these blasphemous implications of Hermetic magic, as well as to the tensions defined by C. L. Barber, that Marlowe is alluding in the opening scenes of Doctor Faustus. But only in Faustus' anticipatory imaginings are words an instrument by which his will can dominate over the world. The blasphemous rhetoric which he uses is not a net to catch the world, but a web in which he himself becomes entangled. To gain the obedience of Mephostophilis, Faustus must “binde [his] soule” (A: 490) in a form of words, “in manner of a Deed of Gift” (B: 447). There may be slightly alarming as well as agreeable implications in Mephostophilis' promise to give him “more than thou hast wit to aske” (A: 487)—but the real importance of these words is that they reveal the notion of magicians that words constitute a causal link between will and the external world to be an illusion. Even with a devil at his command, Faustus will not get what he orders, but something “more”—whatever that may mean.

He seems in fact to get rather less: his first demand—for “a wife, the fairest Maid in Germany” (B: 532-33)—is met by a practical joke; and later, Mephostophilis' refusal to answer the provocative question “who made the world?” (B: 636) leads to Faustus' angry response: “Villaine, have not I bound thee to tell me any thing?” (B: 640). There is no trace of deference in the retort: “I, that is not against our Kingdome. / This is: Thou art damn'd, think thou of hell” (B: 641-42). The magician whose “servile spirits” (B: 124) were to be the instruments of his wildest whims is being ordered what to think. And Faustus is quickly terrified into compliance by the appearance of Lucifer and Belzebub: he vows “never to looke to heaven, / Never to name God,”—an intention that contradicts itself—“or to pray to him, / To burne his Scriptures, slay his Ministers, / And make my spirites pull his churches downe” (A: 725-28). There is a grave irony to the B-text version of Lucifer's reply: “So shalt thou showe thy selfe an obedient servant...” (B: 667).43



Marlowe is clearly redefining the sense in which words can be said to have magical power, and he is doing so in a manner that assimilates magic to witchcraft. Verbal magic appears to work for Faustus only in reverse: his imagination is captured by words, and then bound by them, with alarming consequences. There remains a sense in which blasphemy can achieve magical effects for him. His mocking use, in the flourish with which he signs away his soul, of Christ's last words on the cross—words which bear the whole weight of the mystery of redemption—has an immediate result:

Consummatum est, this Bill is ended, 
And Faustus hath bequeath'd his soule to Lucifer
But what is this inscription on mine arme? 
Homo fuge, whither should I flie? 
If unto God hee'le throwe thee downe to hell, 
My sences are deceiv'd, here's nothing writ, 
[O yes, I see it plaine, even heere is writ] 
Homo fuge, yet shall not Faustus flye.    (A: 515-20, B: 468, A: 522)

But although the writing on the wall that answered the blasphemies of Balshazar was visible to all, and proclaimed a sentence that was executed that same night, the hallucinatory inscription which his own blood forms on Faustus' arm has power only in his imagination; his death-warrant, though signed in that same blood, is post-dated. On a deeper level than that of verbal magic, Faustus' blasphemies do seem to have a kind of transitive power: he manages to strike a better bargain with hell than witches were supposed to be able to negotiate,44 and he does actually get his agreed twenty-four years of demonic service. The manner in which he sheds his own blood may perhaps explain what happens. Faustus performs what could be described as a symbolic suicide:

Loe Mephastophilus, for love of thee, 
I cut my arme, and with my proper blood
Assure my soule to be great Lucifers
Chiefe Lord and regent of perpetual night, 
View heere the blood that trickles from mine arme, 
And let it be propitious for my wish.    (A: 493-98)

By this act he becomes, in effect, his own ritual murderer: though he has killed someone closer to himself than a brother, he bears (one might say) the mark of Cain.

Faustus is set apart for a punishment greater than he can bear; at the same time—and in part because of his inability to stand up even to the threat of immediate punishment—he seems to enjoy for the period of his compact an immunity from bodily harm. Faustus' shedding of his own blood is both a deliberate self-exclusion from the number of those for whom Christ's blood was shed and also a kind of blasphemous anti-Christian sacrament by which he becomes an anomalous or mediating term between the present state of life on earth and the eternity in hell to which he vows himself. In mythical structures, mediating terms between incommensurable categories or states of being are commonly felt to have a numinous, daimonic, or sacred quality.45 To say that Faustus is protected from external harm by having entered a taboo or sacred realm would be to exaggerate; and yet it would appear that he could not die within the time he has bargained for, unless by his own hand, in a real, not symbolic suicide. In the B-version of Act IV, this aspect of he blood-pact is made grotesquely explicit, and in a manner that puts Faustus' human nature in doubt; but it is also implicit in the A-text. The devils seem to respect this deep blood-magic, and so perhaps do we: few readers, I believe, find it odd that Mephostophilis does not wring Faustus' neck at the earliest opportunity, or drop him from the air, in a state of damnable pride, on the first leg of their Grand Tour.

On the level of verbal magic, however, it is Faustus himself, not the spirits or the external world, that is influenced and swayed. He is able to make use of the incantatory power of language only in a subjective, not a transitive, manner. The “sweete pleasure” of words can, at least temporarily, conquer “deepe despaire”: “Have I not made blind Homer sing to me / Of Alexanders love, and Oenons death?” (B: 595-96). He has also distracted himself with the music of him “that built the walles of Thebes, / With ravishing sound of his meodious Harpe” (B: 597-98); but the reference to Amphion only brings out by contrast Faustus' lack of such power.

The transitive impotence of verbal magic is underlined in Act III by the ludicrous failure of the monks' attempt to exorcize Faustus and Mephostophilis, and in the following scene of the same act by the consequences of Robin's reciting an invocation from the book he has stolen from Faustus. If this was the book which Lucifer gave to Faustus to enable him to take on any shape (and which Faustus promised he would keep “as chary as my life” [B: 739]), then the transformation of the clowns into animal shapes is appropriate—as is the fact that in the A-text we do not encounter them again.46

In the fourth act of the A-text this pattern is continued: there is no direct link through words between the imagination and will and the transitive effects of magic. The Emperor tells Faustus, “I have heard strange report of thy knowledge in the blacke Arte, how that none in my Empire, nor in the whole world can compare with thee, for the rare effect of Magicke: they say thou hast a familiar spirit, by whome thou canst accomplish what thou list...” (A: 1040-44).47 And Faustus accepts, with a modest disclaimer, this assessment of his powers: “My gratious Soveraigne, though I must confesse my selfe farre inferior to the report men have published ..., yet ... I am content to do whatsoever your majesty shall command me” (A: 1051-55). Faustus does not lack confidence or pride; he “doubt[s] not” that the spirits' representation of Alexander and his paramour “shal sufficiently content your Imperiall majesty” (A: 1090). But he is not making things happen by the triumphant power of his imagination and his will, working through words; everything is done by his attendant: “Mephastophilis be gone” (A: 1097). The words with which he introduces this magic thus belong at the level of prose. Similarly, it is appropriate that the Emperor should speak in prose of Faustus' reputation—which he wants to see substantiated—but that his language should rise to verse when he reveals his desire to see

                                         Alexander the great, 
chiefe spectacle of the worldes prehemin[en]ce, 
The bright shining of whose glorious actes
Lightens the world with his reflecting beames, 
As when I heare but motion made of him, 
It grieves my soule I never saw the man....    (A: 1063-68)

The Emperor does indeed ask Faustus to use the “cunning of thine Art” (A: 1069); but Faustus' willingness to comply, “so farre forth as by art and power of my spirit I am able to performe” (A: 1078-79), provides an exact statement of the manner in which he operates: the power, and probably the art as well, is not his but his spirit's.

The B-text substitutes a wholly different pattern. We hear that “Faustus at the Court is late arriv'd, / And at his heeles a thousand furies waite, / To accomplish what soever the Doctor please” (B: 1209-11). And “The Emperour is at hand, who comes to see / What wonders by blacke spels may compast be” (B: 1225-26). His greeting, and Faustus' reply, try but fail to be more impressive and dignified than their prose counterparts in the A-text. Both speeches in the B-text emphasize a verbal incantatory magic in which words form a causal link between the ruling will and what it dominates—precisely that kind of magic that Marlowe has shown Faustus to be incapable of. The liberation of Bruno, says Carolus, “Shall adde more excellence unto thine Art, / Then if by powerfull Necromantick spels, / Thou couldst command the worlds obedience” (B: 1241-43). Faustus reassures the Emperor, by a display of abject humility, that he has no such political aims. But in lines which I have already quoted, he seems to imply that by “power of Art” and “Magicke charmes” he is capable of anything—though he puts this power at the Emperor's disposal.

The presence in the B-version of two quite incompatible views of Faustus' magic creates obvious inconsistencies. If Faustus after all possesses powers of a kind that the first two acts have shown him not to have, why then should the man who has anticipated, before discovering his limitations, that “The Emperour shall not live, but by my leave, / Nor any Potentate of Germany” (B: 335-36) now present himself to “most royall Carolus” as “poore Faustus” (B: 1250-51)? This is an unprofitable question—which the A-text does not tempt us to ask. For its fourth act does not clash with what we have been shown in Act II—a man who is deflected by despair from any sustained enterprise, and whose power over Mephostophilis is not a matter of command, but of permission.

More disturbingly, the B-text episodes of Benvolio's revenge and the horse-courser's leg-pulling and its consequences make Faustus no longer human, but a kind of monstrous amphibian—part demon, part animal—which can magically reconstitute its dismembered body and devour a whole load of hay. (“O monstrous” [B: 1608], say the clowns on hearing of this last feat.) Faustus is not interested in the return of his severed head—“Nay keepe it” (B: 1445)—for “had you cut my body with your swords, / Or hew'd this flesh and bones as small as sand, / Yet in a minute had my spirit return'd, / And I had breath'd a man made free from harme” (B: 1449-52). And the detachable leg, which in the A-text seems to be a demonic illusion48—though the audience's senses are deceived, as well as the horse-courser's—becomes in the B-text a part of Faustus' body. The horse-courser keeps it “at home in mine Hostry” (B: 1629), and for some thirty prose lines in the Vanholt scenes the clowns twit Faustus, on the assumption that he must now have a wooden leg; learning the truth they exclaim in chorus, “O horrible...” (B: 1755).



This disruption in the B-text of the play's patterns of meaning and its rhetorical decorum has serious consequences. For if Faustus is going to become once more in the last act of the play a human being, a tragic figure, then a major re-adjustment of the audience's responses is necessary. The moralistic or mocking tone of much recent criticism based on the B-version of Doctor Faustus suggests that this adjustment is not an easy one to make.

The great speeches of Act V—in particular, Faustus' rhapsodic address to Helen and his last soliloquy—are permeated by a verbal magic which is transitive only in its effects upon the audience. We are not conjured to do Faustus' will; rather, we are drawn into empathy with his predicament. His expressed intentions in these speeches are defeated, and yet the glamour which his words cast tends to remove one's desire, and perhaps one's capacity, to make detached moral judgments of him. Consider the speech to Helen. Although Faustus has asked for her in the hope that her “sweet embraces may extinguish cleare, / Those thoughts that do disswade [him] from [his] vow ... to Lucifer” (B: 1867-69), the knowledge of his predicament burns up through the very words that are intended to repress it:

Brighter art thou then flaming Jupiter
When he appear'd to haplesse Semele....    (B: 1889-90)

But who in the audience remembers, by the end of this speech, that Faustus came to Helen direct from his betrayal of the Old Man, and with the visible stain of his cowardice still on him, in the form of the blood shed in the renewal of his pact with hell? The self-inflicted wound of Faustus' perverse atonement has been transmuted, through his blasphemous adoration of Helen, into an aspect of his playful chivalry:

And I will combat with weake Menelaus
And weare thy colours on my plumed crest. 
Yea, I will wound Achilles in the heele, 
And then returne to Hellen for a kisse.    (B: 1883-86)

In the last scene of the play, this audience-directed verbal magic is still stronger. The effect is paradoxical, for Marlowe's magic works on us through the utter failure of his protagonist's. Faustus cries out in his agony: the stars are not held back, nor the mountains moved; but the audience is riveted.

Yet can an audience that has arrived at these scenes by way of the cheap tricks of the B-text Act IV respond fully to this magic? I doubt it. There is all the more reason, then, to wait impatiently for a new edition of Doctor Faustus based upon the A-version of the play.





1  All quotations from the play are from W.W. Greg, ed., Marlowe's Doctor Faustus 1604-1616 (1950; rpt. Oxford: Clarendon Press, 1968). Quotations are identified by Greg's line numbers; u/v and i/j have been silently normalized. It will be noticed that my usage distinguishes between the terms “version” and “text”: “version” refers to the general shape of the play in its 1604 or 1616 form, and “text” to the actual texts themselves. My demonstration of the superiority of the A-version is based on an analysis of passages in which the two versions diverge. In parallel passages I have felt free to quote from the B-text whenever I felt its readings to be preferable.

2  Leo Kirschbaum, “The Good and Bad Quartos of Doctor Faustus,” The Library, n.s. 26 (1946): 272-94; Greg, pp. 1-150, 295-405.

3  Fredson Bowers, ed., The Complete Works of Christopher Marlowe, 2 vols. (Cambridge University Press, 1973), II, p. 139.

4  Greg, pp. 60-62.

5  Greg, pp. 63-97. Greg's argument on this point is accepted by Bowers, II, p. 142.

6  My published text incorrectly reads “the A-version.” I have corrected this small error here.

7  Greg, p. 11.

8  F. S. Boas, ed., The Tragical History of Doctor Faustus (1932; 2nd ed. London: Methuen, 1949), pp. 21-28.

9  Boas, p. 30, n.1.

10  Bowers writes: “It was particularly unfortunate that Kirschbaum felt the demonstration of the memorial nature of the A-text involved the hypothesis that it was a redaction of the whole of the present B-text so that one in effect proved the other and that this linking of two essentially discrete problems powerfully affected the thinking of Sir Walter Greg.... The two problems are in fact independent; and where the evidence appears to overlap in certain comic scenes, the hypothesis of B-text revision as synonymous with the 1602 additions is as satisfactory as A-text memorial corruption and shortening where this evidence concerns the identifiable revisions” (II, p. 127 n.).

11  Roma Gill objected to certain of Greg's hypotheses (Gill, ed., Doctor Faustus, New Mermaids [London: Benn, 1965], pp. Xv-xvi), and J.B. Steane argued that the A-version is “artistically stronger” (Steane, ed., Christopher Marlowe: The Complete Plays [Penguin, 1969], p. 261); both editors, however, based their editions upon the B-text and the B-version of the play.

12  Bowers, “Marlowe's Doctor Faustus: The 1602 Additions,” Studies in Bibliography 26 (1973): 1-18; Bowers, Works, II, pp. 123-55; Constance Brown Kuriyama, “Dr. Greg and Doctor Faustus: The Supposed Originality of the 1616 Text,” ELR 5 (1975): 171-97.

13  Bowers, Works, II, pp. 136-37; cf. Greg, p. 28.

14  Greg, p. 28.

15  Kuriyama, 176, 191-96.

16  Bowers, Works, II, p. 138. (I have quoted the lines from Hamlet from the Second Quarto.) Greg, oddly enough, was well aware of this parallel: “...the only point that might make me suspect that any part of the B-text was as late as 1602 is a possible echo of Hamlet in l. 1200” (p. 29 n.). This can serve as a warning against the workings of prejudice in such delicate matters as the assessing of external evidence: had Greg quoted this parallel in his text, rather than relegating it to a footnote, he would have been unable to press his arguments for the priority of the unique B-text passages.

17  Kuriyama, 174-76, 181-95. The parallels with A Looking Glass would seem to indicate an early date (before 1592) for Doctor Faustus.

18  Bowers, Works, II, p. 138 n.: “Analysis suggests that every major difficulty in Greg's reconstruction of the history of the texts ultimately refers back to the denial of the evidence of revision as applying to the Rowley-Birde additions. Acceptance of these, for example, removes all need to argue that Marlowe must have reworked the play in the promptbook, that the passages peculiar to B in Act V are original drafts discarded in the final make-up for performance, that—contrary to the evidence of the A-text—Rowley was probably Marlowe's original collaborator, and so on.”

19  Kuriyama, 179.

20  Bowers, “Marlowe's Doctor Faustus: The 1602 Additions,” 5-8.

21  As Bowers observes, the choice of a copy-text, and the question of whether the edited text should be based in its structure on the A- or the B-version of the play, are separate matters (Works, II, pp. 142-44). It is the latter decision that I am questioning. I have seen no adequate demonstration that such passages as A: 1057-76, 116-19, and 1134-42 were not written by Marlowe; indeed, I suspect that such a demonstration would be impossible. If the stylistic and structural integration of these passages makes the question of authorship difficult to resolve, that should itself serve as a guide to editors in their choice between the A- and B-versions of the play.

22  These consist of the lines which prepare for the Saxon Bruno material (discussed below), the horse-courser's repetition of the story of his dousing (compare A: 1175-89 with B: 1553-61 and 1611-22), and the repetition, though in different form, of Wagner's response to Faustus' bequests to him (compare A: 1267-71 with B: 1777-82 and 1915-19).

23  H. Dugdale Sykes, in The Authorship of “The Taming of a Shrew,” “The Famous Victories of Henry V,” and the additions to Marlowe's “Faustus” (London: Shakespeare Association, 1920), was the first to offer stylistic evidence for Samuel Rowley's authorship of the greater part of the B-text additions. Sykes' arguments, which are supported by L.M. Oliver's discovery that the Saxon Bruno episode in the B-text is derived, like parts of Rowley's When You See Me You Know Me, from Foxe's Book of Martyrs (cf. Oliver, “Rowley, Foxe, and the Faustus Additions,” MLN 60 [1945]: 391-94), were largely accepted by Greg (pp. 133-35). Kuriyama's observations (191-96) refine and strengthen this evidence. Rowley's mannerisms include a fondness for postpositive adjectives and an unusually frequent use of the oath “swouns” and of phrases like “I warrant you,” “I promise you,” “I can tell you,” and “trust me.” Kuriyama's remark that these phrases produce an effect of “vague emphasis” (192) can be applied with equal justice to Rowley's other mannerisms.

24  Greg, p. 114.

25  Bowers, Works, II, p. 134.

26  Cf. Greg, pp. 133-34. It should be noted that a similar postpositive construction (“demonstrations magicall”) occurs at A: 183 (B: 172) in the first scenes of the play, in a passage that is otherwise suspect: “some lustie grove” in A: 184 (B: 173: “some bushy Grove”) is possibly a reviser's awkward anticipation of “some solitary grove” (A: 186, B: 175). The four-line speech by Faustus which contains this postpositive adjective and this possible anticipation may be a 1602 revision by Rowley of an original speech that is now lost. See note 46 below for another instance of 1602 revisions finding their way into the A-text. Although such passages should prevent us from making extravagant claims about the integrity of the A-text, they do not affect my argument that the A-version of the play possesses a structural and thematic integrity that shows it to be closer to the original than the B-version and preferable to the B-version as the basis for a conflated text.

27  Greg (p. 352) quotes the source passage from John Foxe's Acts and Monuments of these Latter Days (i.e., The Book of Martyrs), 1563, sig. F3: “The proude Pope setting his foote upon the emperours necke, saide the verse of the Psalme: Super aspidem et basilicum ambulabis et cõcaltabis [sic for 'concalcabis'] Leonum et draconem: That is, Thou shalte walke upon the adder, & the Basiliske: And shalt treade downe the Lion and the dragon. &c. To whom the Emperour answering againe, saide: Non tibi, sed petro: that is: not to thee, but to Peter. The Pope again, Et mihi, et Petro. Bothe to me and to Peter.”

28  I Henry IV, II.ii.41-43 (First Folio).

29  A pair of such anticipations is produced by Frederick and Martino, the accomplices of Benvolio, when they have struck off Faustus' head:

     Fred.   Was this that sterne aspect, that awfull frowne, 
Made the grim monarch of infernall spirits, 
Tremble and quake at his commanding charmes? 
     Mar.   Was this that damned head, whose heart conspir'd
Benvolio's shame before the Emperour.    (B: 1422-26).

The second scholar responds to the sight of Helen with another anticipation of the same lines: “Was this faire Hellen, whose admired worth / Made Greece with ten yeares warres af[f]lict poore Troy?” (B: 1804-05). Frederick's anticipation is part of the B-text's disruption in Act IV of the play's treatment of verbal magic. The other anticipations are purely parasitical.

30  Boas' view that this episode is a “clever adaptation of Faustbook material,” and that it “forms an effective finale to the humorous episodes” (The Tragical History, “Introduction,” p. 26), has not been widely shared by other critics.

31  All's Well That Ends Well, II.iii.3-6 (First Folio).

32  Greg, p. 23.

33  See Frank Manley, “The Nature of Faustus,” MP 66 (1968-69): 218-31.

34  Kuriyama, 189-91.

35  Kuriyama, 190.

36  I Tamburlaine, I.ii.419 (“What stronge enchantements tice my yeelding soule?”), 407-408 (“Nor are Apollos Oracles more true, / Then thou shalt find my vaubts substantiall”); quoted from The Works of Christopher Marlowe, ed. C.F. Tucker Brooke (1910; rpt. Oxford: Clarendon Press, 1941), pp. 19-20.

37  My use of the word “transitive” follows D.P. Walker's important distinction, in his Spiritual and Demonic Magic from Ficino to Campanella (1958; rpt. South Bend, Ind.: University of Notre Dame Press, 1975), pp. 76-82, between the “subjective” and the “transitive” effects that were attributed to different kinds of magic by Renaissance writers. Subjective effects “remain within the operator or those taking part in the operation,” while in transitive operations “the operator imposes an effect on someone else without undergoing it himself” (p. 82).

38  C. L. Barber, “'The form of Faustus' fortunes good or bad',” Tulane Drama Review 8 (1963-64): 96-97.

39  A. D. Nock, “Early Gentile Christianity and its Hellenistic Background,” in his Essays on Religion and the Ancient World, ed. Zeph Stewart, 2 vols. (Oxford: Clarendon Press, 1972), I, p. 61.

40  See, for example, E. Garin et al., ed., Testi umanistici su l'ermetismo (Rome: Fratelli Bocca, 1955), pp. 23-77, 107-62; P.O. Kristeller, “Marsilo Ficino e Lodovico Lazzarelli” and “Ancora per Giovanni Mercurio da Correggio,” in his Studies in Renaissance Thought and Letters (1956; rpt. Rome: Edizioni di storia e letteratura, 1969), pp. 221-57; D.P. Walker, Spiritual and Demonic Magic from Ficino to Campanella, pp. 64-72, 90-96.

41  Three Books of Occult Philosophy, written by Henry Cornelius Agrippa, of Nettesheim, tr. by J.F. (London, 1651), I.lxx, p. 153. Agrippa's text is as follows: “Dicunt iccirco Platonici in hac ipsa voce, sive verbo, sive nomine, iam suis articulis formato, ipsam vim rei sub significationis forma, quasi vitam aliquam latere: primò ab ipsa mente quasi per semina rerum conceptam, porrò per voces sive verba quasi partum editam, postremò etiam scriptis servatam. Hinc dicunt Magi, propria rerum nomina esse quosdam rerum radios ubique semper praesentes, rerumque vim servantes...” (quoted from R.H. Popkin, ed., Heinrich Cornelius Agrippa von Nettesheim: Opera, 2 vols. [c. 1600: facsimile rpt. Hildesheim: Georg Olms, 1970], I, p. 140, De occulta philosophia, I.lxx).

42  Three Books of Occult Philosophy, II.lx, pp. 337-39. Agrippa's words are as follows: “Non enim ab alia causa quàm à coelesti, terreni orbis virtutes proveniunt. Hinc Magus per illas operaturus, utitur invocatione astuta superiorum, verbis mysteriosis, & locutione quadam ingeniosa, trahens unum ad aliud: vitamen naturali per quandam convenientiam inter illas mutuam, qua res sponte sequuntur, sive quandoque trahuntur invitè.... Verba autem hominum, res quaedam sunt naturales: & quia partes mundi naturaliter se invicem trahunt, & in se mutuò agunt, idcirco magus invocans per verba, operatur per vires naturae aptas, quaedam amore unius ad aliam ducendo aut trahendo propter sequelam unius rei ad alteram: aut repellendo propter odium unius ad aliam, ex rerum contrarietate & differentia, virtutumque multitudine: quae licet sint contrariae aut differentes, perficiunt tamen partem unam: quandoque etiam dominio quodam cogit res virtute coelesti, quoniam non est alienus à coelo.... sed & virtute à Deo per verbum eius insita, quod verbum Chaldaei Babyloniae sapientes, vocant causam causarum: quoniam ab eo producuntur entia, etiam ipse intellectus agens ab eo secundus. Id autem propter unionem verbi huius cum autore primo, à quo omnia existentia verè producuntur. Verbum igitur, id est, simulacrum Dei: intellectus agens, est simulacrum verbi: anima est simulacrum intellectus: verbum autem nostrum est simulacrum animae, per quod agit in res naturales naturaliter, quoniam natura opus illius est.” Opera, I, pp. 302-304, De occulta philosophia, II.lx.

43  In both texts Faustus is told not only what to think, or not to think, but also what not to say: Lucifer responds to his amusingly naïve remark, “That sight will be as pleasant to me, as Paradise was to Adam the first day of his creation” (B: 672-73, cf. A: 733-34), with the stern admonition: “Talke not of Paradice or Creation, but marke the shew” (B: 674-75, cf. A: 735-36). The effect of this is weakened in the A-text when Lucifer goes on to tell Faustus what to say: “talke of the divel, and nothing else” (A: 736). Faustus' position is clear enough without this extra emphasis.

44  Reginald Scot's description of “such as are said to bee witches” includes many of the features of the witch stereotype: they “are women which be commonly old, lame, bleare-eied, pale, fowle, and full of wrinkles.... These go from house to house, and from doore to doore for a pot full of milke, yest, drinke, pottage, or some such releefe: without the which they could hardlie live: neither obtaining for their service and apines, nor by their art, nor yet at the divels hand (with whome they are said to make a perfect and visible bargaine) either beautie, monie, promotion, welth, worship, pleasure, honor, knowledge, learning, or anie other benefit whatsoever” (The Discoverie of Witchcraft, ed. Brinsley Nicholson [1886; rpt. East Ardsley: EP Publishing, 1973], I.iii. pp. 5-6). It is typical of Scot that he also identifies another category of 'witches'—the “couseners”—and that he emphasizes the paradoxical nature of this pact which offers nothing substantial to the witch.

45  See Edmund Leach, Genesis as Myth and Other Essays (1969; rpt. London: Jonathan Cape, 1971), pp. 9-15.

46  The A-text at this point shows signs of revision. It seems probable that the scene ended with Mephostophilis' entrance, with the terrified lines spoken to him by the Vintner, Rafe, and Robin (A: 1014-19), and with his command: “Vanish vilaines, th'one like an Ape, an other like a Beare, the third an Asse, for doing this enterprise” (A: 1021-22). There is no indication in these lines that Mephostophilis was compelled by Robin's conjuration; the clown's jumbled nonsense, like Faustus' invocation, is only efficacious per accidens. The stage direction “Enter to them Meph.” (A: 1020) is redundant in the text as it stands: he is already on stage. It also appears to be misplaced, for Mephostophilis' lines A: 1032-33 contain an alternate version of his transformation of the clowns (though not, in this case, the unoffending Vintner). The direction at A: 1020 was presumably written to precede the alternative ending of the scene, which begins with lines A: 1023-28:

Monarch of hel, under whose blacke survey
Great Potentates do kneele with awful feare, 
Upon whose altars thousand soules do lie, 
How am I vexed with these vilaines charmes? 
From Constantinople am I hither come, 
Onely for pleasure of these damned slaves.

The view of verbal magic contained in these lines is distinctly that of the 1602 additions: the devil as come not by Lucifer's command, or of his own accord (cf. A: 285-91), but because “these vilaines charmes” have compelled him. If these lines are in fact derived from the 1602 revision, it is curious that they are better preserved here than in the B-text, where (presumably to avoid the fines imposed for blasphemy under the 1606 Act of Abuses) lines A: 1023-25 are replaced by B: 1159 (“You Princely Legions of infernall Rule”).

47  This passage is very close to the wording of Ch. 29 of the English Faustbook.

48  The leg business seems, in the A-text, to be an example of the kind of demonic deception of the senses that was commonly attributed to those magicians known as praestigiatores. Johann Godelmann, in his Tractatus de magis, veneficis et lamiis (Frankfurt, 1591), discusses under this category the magicians of Pharaoh, Simon Magus, and “celebris Ioan. Faustus superiori seculo” (I.iii, pp. 25-28).