University of Virginia Library

Search this document 


expand section 
expand section 
expand section 
expand section 
expand section 
expand section 
expand section 
expand section 
collapse section 
John Webster and Thomas Heywood in Appius and Virginia: A Bibliographical Approach to the Problem of Authorship by MacD. P. Jackson
expand section 
expand section 
expand section 
expand section 
expand section 
expand section 

expand section 


Page 217

John Webster and Thomas Heywood in Appius and Virginia: A Bibliographical Approach to the Problem of Authorship
MacD. P. Jackson


John Webster's reputation rests upon two fascinating Italianate tragedies which, though displaying such baroque excess of local elaboration and such straining for sensational effect as to lack clarity of outline or obvious coherence of purpose or point of view, are pervaded by a distinctive and haunting atmosphere that blends savagery, melancholy, and a nerve-tingling sense of impending doom. And the "Tussaud laureate" who created The White Devil and The Duchess of Malfi set his seal no less surely on their dazzling tragi-comic counterpart, The Devil's Law Case. But Appius and Virginia, first published in 1654 with a title-page ascription to Webster, is less clearly a product of the same singular imagination; for this Roman tragedy is "striking in its almost classical simplicity of construction" (Lucas, Webster's Works, III, 146). In plot it is straightforward, direct, and unified, and in style it is restrained. Can such a work really be Webster's?

Rupert Brooke thought not, vigorously arguing in 1913 that Webster's sole contribution to Appius was the revision of I.i and IV.i, the play being Heywood's.[1] Most scholars have since apportioned Appius between Webster and Heywood, although Bentley was characteristically impatient at what he pronounced "impressionism rationalized," and saw "no grounds for any dogmatic assertions about the authorship" (Jacobean and Caroline Stage, V, 1247). F. L. Lucas could scarcely be accused of dogmatism on the subject. In his standard edition of Webster's works he hesitantly ascribed I.i, I.ii, III.ii, III.iii, IV.i, and V.i to Webster alone, considered II.ii largely his, and found traces of his work in every scene except (doubtfully) II.i, IV.ii, and V.ii. In his massive two-volume study of the playwright, Fernand Lagarde surveyed all the evidence for authorship of Appius and Virginia, and on the basis of his own intimate knowledge of the dramatist and of an analysis of orthographical and linguistic forms concluded: "Nous proposons d'attribuer sans réserve à Webster I.i., I.ii., II.ii., III.iii., III.iii., IV.i. et V.i., et de lui concéder


Page 218
un certain rôle dans la composition de I.iii., de I.iv., de II.iii. et de V.ii." Peter B. Murray, after counting some contractions, felt most confident of Webster's authorship of I.ii, II.ii, III.ii, III.iii, IV.i, and V.ii. And Sanford Sternlicht's thorough comparison of the play's imagery with that of the three undoubted Webster plays persuaded him that five scenes (I.i, I.ii, II.ii, III.iii, and IV.i) "belong fully and strongly to Webster," and that three others (IV.ii.1-44, V.i, and V.ii) show signs of his hand.

Lucas, Lagarde, Murray, and Sternlicht exhibit a fair measure of agreement. Yet none offers compelling evidence, and the problem remains intractable. Arguments from diction and imagery are, as Lucas was well aware, of dubious value, when the whole artistic impulse of Appius and Virginia is towards a far more chaste dramaturgy than that which served Webster in his Italian tragedies. Though Lagarde, unlike Murray, gives an accurate and fairly full tabulation of the linguistic forms in Appius and Virginia, Lucrece is his sole source of comparative material for Heywood. There is still much to be done along this line of enquiry. My purpose in this article is twofold: to provide, through bibliographical and, in particular, compositorial analysis of the 1654 Quarto of Appius, a sold basis for examination of those orthographical and linguistic minutiae which may aid author identification, and to undertake a more comprehensive examination of this sort of evidence than has been reported up till now.

The Quarto of 1654: Printer and Compositors

The sole Quarto of Appius and Virginia survives in no fewer than five issues, with successive cancels replacing the original title leaf (Greg, Bibliography, II, 733). The first issue, of which Greg records two copies, was printed for Richard Marriot, to whom it had been entered in the Stationers' Register on 13 May 1654, and is dated 1654, as is the second, of which there are eight copies; the third, extant in a unique copy in the Library of Congress, is dated 1655; the fourth, of which there are eight copies, is dated 1659 and was printed for Humphrey Moseley, to whom the rights had been transferred on 11 June 1659; and the fifth, of which there are four copies, is as late as 1679. The collation is A1 B-H4 I4 (—I4), the text beginning on B1 and ending on I3.

The printer, hitherto unidentified, was almost certainly Thomas Maxey (active 1637-1657), whose name as printer is on the title page of at least seven of Marriot's publications in the period 1651-1655,[2] including Isaak Walton's The Complete Angler (1653, and second edition 1655), Sir Henry Wotton's Reliquiae Wottonianae (1651, and second edition 1654), and John Donne's Essays in Divinity (1651).


Page 219

There is little about the Appius Quarto that might serve readily to identify its printer—only an arrangement of nine unusual fleurons on the title page and a double row of small acorn-like ornamental types, of a very common kind, above the head title on B1. Both these forms of ornamentation are, however, found several times in Reliquiae Wottonianae (1654, Wing W 3649) on A3, A5, B1, E2, F7, H1, and many other pages; in Edward Sparke's Scintillula Altaris, printed by Maxey (1652, Wing S 4807) on A3. A4v, A5v, A6v, A7, B2, C1, and other pages; and in Essays in Divinity (1651, Wing D 1861) on A1, A2, and K11v; and the more distinctive appears in The Complete Angler (1653, Wing W 661) on A5; The Complete Angler (1655, Wing W 662) on Q11; Reliquiae Wottonianae (1651, Wing W 3648) on a1, b1, A1, D1, F4, I5, and P4; and Sir Thomas Overbury's Observations upon the Provinces United (1651, Wing O 609), also printed by Maxey, on A4.

Moreover, the Appius fleuron is used on A2v of the Quarto of Henry Glapthorne's Revenge for Honour, which has a publishing history similar to that of Appius, appearing in three issues, two of 1654 "for Richard Marriot," and a third of 1659 "for Humphrey Moseley" (Greg, Bibliography, II, 730). In the second issue, surviving in a single copy in the Pforzheimer Library, a leaf containing an Epistle Dedicatory has been inserted after the title. Greg notes: "While there is no proof that the leaf belongs to the book, there seems no reason to doubt it. According to the Pforzheimer Catalogue the printer of the leaf was Thomas Maxey (1647-57), but the printer of the play is not known." Greg does not say on what basis the Pforzheimer Catalogue assigned the printing of the second issue's additional leaf to Maxey, but that he was in fact the printer of the whole play is strongly suggested by the type ornamentation that the Quarto has in common with Scintillula Altaris (1652), The Complete Angler (both 1653 and 1655), and the other books printed by Maxey. Compare, for example, A2 of Revenge for Honour with A2, A8, and B1 of the first edition of The Complete Angler (1653), and with A2 and B1 of the second edition (1655). Or compare the opening A2v-A3 of Revenge for Honour with the opening 2C1v-2C2 of Scintillula Altaris. The habit of dividing a row of ornamental types by colons links Revenge for Honour with the books known to have been printed by Maxey. In Revenge for Honour, as in Appius and Virginia, only two leaves per gathering are signed, always in the form 'B', 'B2', without stops, and in neither play are lines of verse capitalized. Both Quartos seem to use the same fount of type, of which twenty lines measure approximately 81-82 mm.[3] Thomas Maxey is, then, the probable printer of both Revenge for Honour and Appius and Virginia.

Analysis of the headlines in the Appius Quarto reveals that a single skeleton was used for inner and outer formes of sheets B-E. This skeleton also machined both formes of G and the outer formes of H and I. However, a second


Page 220
skeleton was introduced to machine both formes of F and the inner formes of H and I.[4] Tabulation makes the sequence clear:      
This skeleton pattern implies an undisturbed relationship between composition and presswork over the first four sheets, but suggests that the first forme of F may have been prepared for the press before the second forme of E had been released. It would not be surprising to find some change in the compositorial pattern in sheet F.

A hint towards differentiation of compositors is given by the spellings of those useful test words do and go. On some pages the spellings do and go are preferred, on others doe and goe. Pages exhibiting a decisive preference for do and go are C1, C2, C4v, D4, E1, F2, F3, F4, G2v, G3, G3v, and I4v; and for doe and goe B2v, C3v, D1, D1v, E2v, E3, E3v, E4, and E4v.[5] When the twelve do/go pages are carefully compared with the nine doe/goe pages, further differences emerge, and these all prove statistically significant beyond the one per cent level, when tested by Yates's chi-square.[6] Table 1 presents these results:

Table 1

12 do/go pages  9 doe/goe pages 
(Compositor A)  (Compositor B) 
do/go   39  do/go  
doe/goe   doe/goe   31 
I'l   15  I'l  
I'le   I'le/Ile   10 
-ness   10  -ness  
-nesse   -nesse  
-l   22  -l  
-ll   -ll   22 


Page 221
Figures for -l and -ll endings exclude monosyllabic words (shall, will, well, still, and so on), of which the modern spelling predominates throughout, but the do / go speller, whom I have labelled Compositor A, nevertheless uses the -l ending in a monosyllable 9 times in his twelve pages, whereas the doe / goe speller (Compositor B) never uses it. Compositor A's I'le spelling occurs on a page (F4) on which I'l also appears. Compositor B's anomalous single -l spelling is in the last word of a full line, and is outweighed on the page (E2v) by four -ll endings. Compositor B's uncharacteristic do spelling is also in a full line, and there are eight doe / goe spellings on the same page (E3v).

From Table 1 we may safely conclude that at least two compositors set the text and that the spellings listed there will serve to distinguish them.

When the whole text is examined in the light of these key words, almost every page affords unequivocal evidence (presented in full in Table 2 at the end of this article) of one or other compositor's handiwork. The overall pattern suggests that the doubtful pages in sheet B are probably Compositor B's, those in F-H probably Compositor A's. The change at F in the relationship between composition and presswork, tentatively inferred from the skeleton pattern, is confirmed by Table 2, which shows Compositor A, who had shared equally with Compositor B the setting of sheets C and D, and also set the first two pages of sheet E, now taking over as sole typesetter. Presumably presswork was lagging behind composition, perhaps because the press was partly occupied with some other job, so that Compositor A could comfortably keep ahead of the press on his own.

Some confirmation of the proposed compositorial divisions is provided by the setting of words immediately following speech prefixes. Compositor A often fails to set adequate spacing between prefix and text. Compositor B is more consistent. There is doubtless subjectivity in my criterion of "adequacy" here, but on A's thirty-five unquestioned pages I count fifty-six instances of deficient spacing between prefix and text, whereas B's nineteen unquestioned pages afford only five. B1, B4, and B4v are all perfect in this respect, while F1, F2v, G4v, and H3v have two, two, one, and two examples of inadequate spacing. So my tentative assignment of the doubtful pages in sheet B to Compositor B, and of those in sheets F-H to Compositor A receives some support from this evidence.

Compositor A always sets speech prefixes for Marcus Clodius as Clod. (36 times), except for one Clo. in a full line on D4: the pertinent pages are D2v, D4, D4v, F1, F1v, F2, F2v, F3, G1, G1v, I1, I2v, and I3. Compositor B never sets Clod., but employs a great variety of other prefixes for Clodius. He begins with two instances of Marcus. Cl. on B1v, sets Marcus twice on B2-B2v, changes on B3v from Cl. to Clodius. (twice) and M. Clodius. (3 times), continuing to use M. Clodius. three times on B4. Then he follows one Clodius. on C2 with Clo. twice on C2 and three times on C3, reverting to M. Clodius. (twice) and Clodius. (6 times) on D2. In E2v-E4v he sets a single Clo. at the top of E3, but otherwise wavers between Clodius. (15 times) and Clodi. (5 times). The difference between the two men is striking:


Page 222

Speech Prefixes for Marcus Clodius

Clod.   other forms 
Compositor A:  36 
Compositor B:  48 
Three tentative attributions are thus further confirmed—B4 to Compositor B, and F1 and F2v to Compositor A.

The Manuscript Copy for Q

What sort of manuscript served as copy for the Quarto of Appius? R. G. Howarth claims: "Obviously the text was prepared for the press, though no title-page (with a motto, in Webster's usual fashion) and no dedication or preliminary address have been drawn up" (PQ, 46, 137). Nor is there any Dramatis Personae, such as heads The Duchess of Malfi and The Devil's Law Case. That the manuscript had been "prepared for the press" is not obvious to me, though it is divided into acts, not scenes, with the formula Actus Primus Scena Prima, and so on, and each act but the last (followed by FINIS) ends with an Explicit. The explicit, used by Heywood in his autographs of The Escapes of Jupiter and The Captives, reappears in the Quartos of several of his plays—The Iron Age (both parts), The Wise Woman of Hogsdon, and The Fair Maid of the West (both parts); but it can also be found ending Acts II and III of Webster's The Devil's Law Case.

The text is comparatively clean, requiring few substantive emendations, and the speech prefixes give little cause for complaint, though Lucas plausibly transfers five words from Numitor to Appius at F4v, IV.i.61, and there is ambiguity in the frequent abbreviations for Virginius and Virginia, both usually Virg. in the Quarto, even when present in the same scene. Corbulo, so named in the speech prefixes at the opening of II.i and III.i (C2, C2v, D4v, E1), abbreviated to Corb. in III.iv (F3, F3v, F4), is Clown. or Clow. in III.ii (E2v, E3, E3v). The Quarto's one Cl. and seven Clo. speech prefixes for Marcus Clodius could cause confusion with the Clown, but Compositor B may have been solely responsible for the abbreviations. Lucas adds a score of exit directions and supplements a few entries, while on F1v Enter Valerius is a minor error, as it is Sertorius who has been called for on E4v and who is dismissed on F2v. So prompt-book provenance seems unlikely. On C1v the stage direction has Virginius enter where Virginia is required, but this error may well have been the compositor's.

In the Quarto minor characters are often designated by a number: 1. Cozen., 2. Cozen., 1. Petitioner., 2. Petit., 1. Serving., 2. Serving. (for servingmen), for example. In three scenes numbers alone are used: in II.ii "six Souldiers" are given numerals as speech prefixes (22 times); in III.ii two of "foure Lictors" speak alternately with the numerals 1 and 2 (10 speeches); and in IV.ii "two Souldiers" are indicated in the same way (19 speeches). The use of unqualified numerals as speech prefixes is not a feature of either Webster's or Heywood's undoubted plays as they have come down to us, though


Page 223
in The Duchess of Malfi, especially, Webster has a great many prefixes of the type in which a numeral precedes a generic title—for two servingmen, four officers, two pilgrims, and four madmen; and Heywood also has many in his autographs of The Captives and The Escapes of Jupiter, and in other plays. The unaccompanied numerals occur in scenes of Appius that will be shown to be wholly or largely Webster's. Their presence suggests foul papers copy.

Perhaps pointing to the same conclusion is the fact that fifteen prefixes in Appius are assigned to Omnes, twelve in II.ii, one in II.i, one in IV.i, and one in IV.ii. Some of the speeches go far beyond simple exclamations of a united rabble, and in performance they would obviously need to be allocated to individuals among those assembled, rather than uttered in chorus. The designation Omnes is an author's conventional shorthand, common enough in dramatic texts of the time. Neither Heywood nor Webster is elsewhere given to such frequent use of collective speech prefixes, and Heywood normally prefers All when he does use them, though there are three examples of Omnes in The English Traveller and two in Webster's The White Devil. II.ii is the first scene in which a sizable group of soldiers must be incorporated into the dialogue and action: hence the special need for the group prefixes here.

Many of the entry directions seem authorial, most notably: Enter Virginius with his knife, that and his arms stript up to the elbowes all bloudy; coming into the midst of the souldiers, he makes a stand. (H1) There is a descriptive tinge to several: Appius enters melancholly on B3, as does the Clowne, meaning the character named Corbulo in speech prefixes, on F3. Iulius enters troubled on D2. Corbulo, called Clown again, enters whispering Virginia, after her M. Clodius with presents on C2: Clodius's entry here is inconsistent with the direction some twenty-five lines later, Enter Clodius and Musicians, which Lucas modifies to Clodius comes forward, with Musicians, supposing that Clodius "should lurk in the background" (III, 250). On F4 Virginius and Virginia each enter like a slave, and on H1 the first mutinous Souldier enters in haste. On I1 Enter Appius, and Marcus Clodius, fettered and gyved. Numbers of minor characters entering are precisely specified except on B4, where the stage direction ends with &c, C2v (Musicians), and D2v (Lictors). On B2v Enter a Servant, whispers Icilius in the eare, spread over three lines against the right-hand margin, begins at least a line too late and is moved up by Lucas.

Another feature of the Quarto that tells against a theatrical origin for the printer's copy is the tendency in some scenes for short speeches that continue a line of verse to occupy the same line as the preceding speech ending. As Greg explained, "In all manuscripts intended for stage use every speech begins a fresh line of writing, irrespective of the metrical division. The habit of writing speakers' names in the middle of a line would be distracting for prompt purposes, and may be taken to indicate a purely literary intention in the writer. The speakers' names are in the margin, and the beginning of speeches are not indented as is usual in print, but a short horizontal rule is


Page 224
drawn separating the speeches."[7] The untheatrical setting out of some short speeches in the Quarto[8] is unlikely to be a printing-house strategy for the saving of space, because it occurs in the stints of both compositors and in various positions on the pages, which are seldom unduly crowded. Moreover, the practice is characteristic of Webster in all three of his unaided plays, and in Appius it is confined to scenes in which there are other strong indications of Webster's authorship, as I shall show.

On the other hand, of the few directions for action that are set to the right of the page, one looks very like a book-keeper's annotation: on H3v Wine, against the right-hand margin, comes ten short speeches before Virginius actually calls "Wilt a, wilt a give me some Wine?" The direction A Shout on I2 might be from the same source.

The evidence suggests to me authorial papers, which may have received some attention from the prompter.

Linguistic Discriminators between Webster and Heywood

Attempts to determine the authorship of Appius and Virginia are complicated by ignorance over its date of composition. Metrical and stylistic considerations have led most commentators to conclude that "Appius lies at one end or other of Webster's development" (Lucas, III, 122), but there is no agreement over which end: it has been dated 1603-4 at the one extreme, 1626-34 at the other. Bentley finds the evidence of earlier scholars "at best inadequate, often absurd, and frequently tied to arguments for hypothetical collaborators," but judges that "the period from 1624 to Webster's death seems to have slightly more to be said for it than have the other dates offered" (V, 1246). Lucas favoured 1625-7 as the most likely date of composition, and 1624 is the year under which it is entered in the Schoenbaum-Harbage Annals. In view of the paucity of hard evidence on the matter, however, Heywood and Webster characteristics that fail to differentiate between the dramatists throughout the first quarter of the seventeenth century are of little legitimate use.

Another complicating factor is doubt over the extent of Webster's participation in the plays in which he is alleged to have collaborated, such as Anything for a Quiet Life, The Fair Maid of the Inn, and A Cure for a Cuckold—and again there are problems of chronology. Could we be certain of his shares in these plays, a wider range of data concerning his stylistic and sub-stylistic characteristics would be available to us, and we might better be able to gauge and understand the changes in his art. But only the early collaborations with Dekker, Westward Ho (1604) and Northward Ho (1605), can


Page 225
be apportioned between the parties with any precision or confidence.[9]

In pursuit of "linguistic" features that may serve to distinguish between Webster and Heywood, I have examined the three plays of Webster's undoubted sole authorship and fifteen plays known to be the unaided work of Heywood, including The Captives and The Escapes of Jupiter, extant in manuscripts generally agreed to be autograph. The minutiae studied were those that Cyrus Hoy, David J. Lake, and others have shown to fall into patterns discriminating neatly between the various Jacobean and Caroline dramatists—colloquial contractions such as 'em, I've, e'en, ha', i'th'; linguistic preferences for ye, you, or 'ee, hath or has; exclamations, oaths, and other expletives; affirmative particles (Ay, Yes, or Yea); connectives such as among or amongst, betwixt or between, while or whilst; and habits in the setting out, spelling, and punctuation of the text.[10]

In many of these respects the two authors are closely alike. Neither makes liberal use of a wide range of contractions and both eschew has and does in favour of the formal hath and doth,[11] except that Webster favours has and


Page 226
does in The Devil's Law Case: we cannot know whether the change reflects chronology (The White Devil is usually dated 1612, The Duchess of Malfi 1614, and The Devil's Law Case 1617), or was made in the interests of decorum, the colloquial forms being felt more appropriate to tragi-comedy than to tragedy. Another possible explanation would look to the probable origins of the texts. John Russell Brown judged that The White Devil was printed from the author's own papers, that The Duchess of Malfi was printed from a transcript, probably in the hand of Ralph Crane, and that while for The Devil's Law Case the indications were slight and inconclusive, they appeared to point to a literary rather than a theatrical manuscript.[12] Although Webster and Heywood avoid various forms common in the work of other dramatists —and these forms are absent from Appius—both men are rather fond of aphetic forms, such as 'fore and 'gainst, which occur several times in Appius; and even the exclamation Ha!, a recognized favourite of Webster's, is of very little value in discriminating his writing from that of Heywood, who uses it quite often in several plays. Murray (A Study of John Webster, p. 38) asserts that instances of 'tis in Heywood's work are "relatively few," but in The Captives, The English Traveller, and The Fair Maid of the West, Part II Heywood uses 'tis with a Websterian liberality, and in some other plays his rate of use is not significantly lower than Webster's.

Among the more striking idiosyncracies of Heywood's autographs, the only one to survive into several printed texts is the spelling Ey for the affirmative Ay (I in most dramatists), but its appearance in a Quarto as late as 1654 would be most surprising, and it is in fact absent from Appius.[13]

The main features of Webster's linguistic pattern were noted by Hoy. Remarking that, apart from the switch to has and does in The Devil's Law Case, the linguistic practices evident in Webster's plays comprise a pattern that is "extremely consistent," Hoy showed that, though Webster avoids most contractions, "his use of i'th', o'th' (which often occurs in his work as a'th'), and 's for his in each of his three unaided plays far exceeds the occurrence of these forms in any single play of such other dramatists whose unaided work I have examined as Fletcher, Massinger, Field, and Ford."[14] Heywood makes little use of i'th', o'th', or 's for his. The two dramatists also differ in their preferences over connectives, one or two expletives are favoured by one man but not the other, and there are a few pseudo-bibliographical markers. Figures for some discriminators that prove relevant to Appius and Virginia are presented in Table 3 at the end of this article. I have omitted good discriminators between Webster and Heywood that fail to appear in Appius, such as the locution Gramercy, which turns up in almost all Heywood's plays but is never used by Webster.


Page 227

The forms listed in Table 3 vary in their power to discriminate. Heywood uses more contractions in i'th', o'th', and to th' towards the end of his career, to which period Appius is perhaps most likely to belong, but the rate of occurrence in his plays always falls far short of Webster's. Most of Heywood's to th' contractions are merely variants of the use, which he favours, of th' before a vowel; only five examples in The Captives and three within the two parts of The Fair Maid of the West are the genuine to th' contraction before a consonant, which is the type employed by Webster. For't appears with Websterian frequency in 2 Fair Maid but none of Heywood's other plays, though Webster's use is itself variable. A few examples of i'th', o'th', to th', and for't in a scene of Appius would be perfectly compatible with Heywood's authorship. Far more distinctive is Webster's liking for of't, which is completely absent from the Heywood plays. Though I have not entered figures in the table, Webster's partiality for enclitic 't contractions generally is much greater than Heywood's, and his range is wider, including many more examples of 't appended to a verb—proclaim't, swear't, bring't, whisper't, and the like. Heywood usually restricts himself to the more common 't contractions, such as is't, in't, do't, to't. The total numbers of enclictic 't contractions in Webster's plays, according to my rather quick counts, are: The White Devil 68, The Duchess of Malfi 102, The Devil's Law Case 76. There are only four Heywood plays in which the total falls outside the range 5-25; in 1 Fair Maid, 2 Fair Maid, and The Captives it is about 50, and in The English Traveller about 60. Again, in three of the four later Heywood plays the figures are high and the rate of use Websterian. Clearly the number and type of 't contractions in a scene of Appius can be used only as slight supporting evidence. The most obvious discrepancy between Heywood and Webster in their use of contractions concerns 's for his: any instance of this type of contraction in Appius must constitute strong evidence for Webster's hand.

The connectives ought to supply useful testimony. Whereas Heywood greatly prefers betwixt (83 times) to between (22),[15] and whilst (150 times) to while (10), Webster employs between exclusively (28 times), and prefers while (19 times) to whilst (6). Though in The White Devil and The Duchess of Malfi Webster shares Heywood's preference for amongst over among, in The Devil's Law Case this preference is reversed. Thus while, among, and between will, with varying decisiveness, point to Webster, whilst and betwixt to Heywood.

A further pointer to Webster's authorship has not been tabulated. In The White Devil he adopts the classical and Jonsonian practice of using small capitals for speech prefixes and setting them within the line when a new speech continues a line of verse. In The Duchess of Malfi and The Devil's Law Case small capitals are replaced by lower-case italic (the names beginning with a capital, of course), but speeches continuing a line of verse still often begin within the line. This practice is entirely alien to the Heywood


Page 228
autographs; it occurs sporadically in the Quartos of The Royal King and the Loyal Subject and The English Traveller, and a few times in The Rape of Lucrece, The Golden Age, and (on a single page) The Silver Age, but it is not otherwise a feature of Heywood's plays.

Linguistic Evidence in APPIUS AND VIRGINIA 1654

Table 4 (at the end of this article) presents the incidence per scene of the discriminators mentioned so far. As Heywood favours no contractions avoided by Webster, Webster markers predominate, and their absence from certain scenes must be an important part of the evidence for Heywood's hand. As far as Table 4 goes, the surest signs of Webster's presence are in II.ii, III.ii, and IV.i, with II.iii, the very short III.iii, IV.ii, and V.i also offering something of value; in V.ii a solitary Webster marker, stop's for stop his, occurs in the third speech from the end of the play. The frequent occurrence of betwixt and whilst confirms that Webster is not the sole author of the play, and is consistent with Heywood's responsibility for much that scholars have assigned him. The divisions suggested by Table 4 are quite independent of the compositorial stints. For example, of the eighteen speech prefixes set within the line, ten belong to Compositor A's pages, eight to B's. All but one of them (one of the two on E3) prefix a speech that continues a line of verse. However, it will be convenient to discuss the evidence of Table 4 section by section.

I.i—II.i: There is very little bibliographical or linguistic trace of Webster in the first Act or in II.i. The five instances of the less distinctive Webster contractions, and two 't contractions, within 514 lines of the first five scenes, contrast markedly with the nine Webster contractions, including the highly distinctive of't, from's, and in's, and four 't contractions, within 250 lines of II.ii, and the change is accompanied by a shift from Heywood's connectives (3 betwixt, 3 whilst) in the first 514 lines to Webster's (3 while, 1 among) in II.ii. Yet every scholar who has divided the play between the two dramatists assigns Webster I.i, and even Rupert Brooke nominated this as one of the two scenes that Webster might have revised, while Lucas, Lagarde, and Sternlicht all give Webster I.ii. There are passages in I.i that certainly seem to me more in Webster's than in Heywood's manner:

My Lord, my Lord, you dally with your wits.
I have seen children oft eat sweet-meats thus,
As fearfull to devoure them:
You are wise, and play the modest courtier right,
To make so many bits of your delight. (I.i.19-23)
I have heard of cunning footmen that have worne
Shooes made of lead some ten dayes 'fore a race
To give them nimble and more active feet:
So great men should, that aspire eminent place,
Load themselves with excuse and faint denyall,
That they with more speed may performe the trial. (I.i.55-60)


Page 229
In the first of the above passages, lines 20-21 are in fact repeated from The Duchess of Malfi, I.i.533-534. I would tentatively suggest that Heywood may have begun copying out the play, and that the first five scenes of the printer's manuscript copy were in his handwriting. Thereafter the handwriting of the authorial papers may have alternated with dominant authorship. In some scenes the work of the two dramatists would appear to be thoroughly mixed.[16] The short I.ii does carry at least an orthographical hint of Webster's presence in the two instances of to th' followed by a consonant. As we have seen, this type of contraction is confined to The Captives and 2 Fair Maid among Heywood's plays, and used in them only eight times altogether, but it is common in Webster's work; and of the first five scenes I.ii is the only one without any indications of Heywood.

There are some oddities about II.i. Hath and doth predominate throughout the play, except in this scene where, besides one hath, occur two examples of has and one of does (C2v-C3). There is one more has in IV.ii (G4). The opening of II.i between Corbulo and Virginia seems to be Heywood's; it is there that hath appears. The two instances of has and one of does are within the encounter between Clodius and Virginia in the second part of the scene. Lucas notes that metrically "it is peculiar owing to its high percentage of feminine endings and quite unlike anything else in the play" (III, 141). He perhaps exaggerates its metrical eccentricity: there are patches of verse elsewhere with many feminine endings, including the monosyllabic endings found here. But it is interesting that the encounter between Virginia and Clodius is introduced with the stage direction that Lucas found difficult to reconcile with that which opens the scene. Moreover, this episode has a heavy concentration of Clo. speech prefixes for Clodius: five of seven Quarto instances congregate here; Clodius is first given his full name, but is Clo. in his five other speeches in this scene. It is just possible that the Clodius-Virginia encounter, which is graced by two songs, is a relic of some third writer's participation. It may have been specially added to accommodate the musicians.

II.ii-II.iii: As noted above, II.ii is unmistakably Websterian, with a steady stream of his linguistic markers. Even the two instances of to th' are significant, each being followed by a consonant. So too is the instance in II.iii, where the evidence for Webster is sparse, but is countered by nothing pointing to Heywood. The contraction At's=at his on D2 is an especially clear pointer to Webster. Moreover, the contraction to'ave, which I have found in neither Webster nor Heywood, links the strongly Websterian II.ii (D1, Compositor


Page 230
B) with the less Websterian II.iii (D4v, Compositor A). Another possible trace of Webster in II.iii is the spelling I'ld for I should on D3v; this occurs several times in The Duchess of Malfi (where it may be Crane's, though you'ld and yee'ld appear in The White Devil, C3v and G4), but not in Heywood's plays.

III.i: Here the absence of Webster markers suggests Heywood's hand, and the single occurrence of whilst tends to confirm it. The 't contraction in this scene is the very common is't, which could as well be Heywood's as Webster's.

III.ii-III.iii: Act III, scene ii exhibits a wealth of diverse indications of Webster, including eight examples of i'th', two of 's, one of yon's, two of for't, the Websterian colloquialisms look you and by my life, and the connective while. The 't contractions are quite varied, including fear't, defer't, hear't, and obtain't. Supporting this evidence is a variation in the naming of Corbulo. In the first twenty lines of II.i (C2) and the first seventeen lines of III.i (D4v-E1) he is Corbulo. in speech prefixes, and in III.iv (F3-F4) his prefixes are abbreviated to Corb. Table 4 supports the generally held view that II.i, III.i, and III.iv are Heywood's scenes. When Corbulo speaks in III.ii, in contrast, he is Clown. or Clow. (E2v-E3v). Admittedly, pages E2v-E3 were set by Compositor B, while all but one of the Corbulo. pages were set by Compositor A, but B did set Corbula. (sic) in II.i at the top of C2v. The difference in naming is probably authorial, rather than compositorial, Webster not caring about this minor character's name. It seems to me obvious enough that the Clown of III.ii has a markedly more satirical, Websterian vein than the amiable punster of the other scenes.

In the very short III.iii deliver't and i'th' point to Webster, and perhaps outweigh the single Heywood marker, tush.

III.iv: The return to the name Corbulo is almost the only clue to the authorship of this scene. There is nothing to suggest Webster; the one instance of for't could easily be Heywood's.

IV.i-IV.ii: The Websterian markers in IV.i are not particularly frequent: i'th' three times, to th' before a consonant, and the expletive by my life. The 't contractions are of the kind that Heywood might have used. However the six speeches beginning within the line associate this scene with the markedly Websterian II.ii and III.ii, and the formality of this trial scene might have inhibited the use of many contracted forms.

One further link between presumptively Websterian scenes is the exclamation by the gods (10 times) or O the gods (4 times) or variant (4 times). This represents a rather perfunctory contribution to the Roman flavour of a play decorously devoid of the more colourful Jacobean expletives. The scenes in which evocations of "the gods" appear are II.ii (7 examples), III.ii (3), IV.i (5), IV.ii (1), and V.i (2). So this is another feature of IV.i connecting it with II.ii and III.ii.

Uncertain signs of Webster in IV.ii are limited to the opening and closing lines. The scene begins with a dialogue between two soldiers, in which 'tween occurs as well as betwixt, and there are three 't contractions suggestive


Page 231
of Webster—see't, hav't, and know't. Some thirty lines from the end of the scene occurs for't, and the last seven lines have while and Be't.

V.i-V.ii: The final Act is scarcely more Websterian than the first in its linguistic forms. The connectives are Heywoodian. The sprinkling of Websterisms in V.i are not among his most distinctive, but i'th', a couple of uses of to th' before a consonant, two instances of for't, and the colloquialism in sooth combine to make Webster's participation in the scene probable. Mixed authorship seems to be indicated. In V.ii the sole sign of Webster, but a clear one, is Stop's sixteen lines from the end.


To summarize: The original Quarto of Appius and Virginia was probably printed by Thomas Maxey. Henry Glapthorne's Revenge for Honour can be identified as likely to have been another product of his shop. The stints of two compositors in Appius can be distinguished with some precision and confidence, and the division is consistent with the skeleton pattern. When linguistic and bibliographical discriminators, derived from study of Webster's three unaided plays and fifteen plays by Heywood, are applied to Appius and Virginia, substantial evidence of Webster's hand is found in II.ii, III.ii, and IV.i, and less persuasive evidence in I.ii, II.iii, III.iii, IV.ii, V.i, and the very end of V.ii. The data concerning authorship are not appreciably affected by contrasting compositorial preferences. It is clear that Webster is not the sole author of the play. The data presented here cannot demonstrate that the second playwright is Heywood, who has no really distinctive linguistic preferences, so that his hand manifests itself in the absence of Webster's forms rather than in features uniquely his own, but they are consistent with Heywood's authorship and part-authorship of several scenes.

Lucas (III, 137) lists twenty-two words in Appius that are typical of Heywood's latinate diction. These are distributed thus: I.iii (2), I.iv (4), II.i (1), II.iii (2), III.i (1), IV.i (1), IV.ii (3), V.ii (8). Thus the scenes that are most strongly Websterian on the linguistic and bibliographical evidence, II.ii, III.ii, and IV.i, contain only one such word in 1012 lines, leaving twenty-one within the remaining 1579 lines. The three Heywoodian words in IV.ii, which I have claimed to be at least partly Webster's (pointers to his hand accumulate within the first and last forty lines or so), are confined to a short passage, lines 79-102. It cannot be coincidental that Heywood's diction is prevalent in portions of the text free from Webster's linguistic traces.

Appius and Virginia is, then, a joint composition. We cannot know whether Heywood and Webster worked in collaboration, or one dramatist revised the other's play. But the final product is an interesting illustration of a phenomenon that has been observed in other Elizabethan and Jacobean plays of dual or multiple authorship—the partial submerging of the more idiosyncratic traits of markedly different dramatists as they contribute to a worthwhile drama with its own distinctive quality. Heywood is an unsophisticated dealer in broad humour and in emotional and moral blacks and


Page 232
whites, "a facile and sometimes felicitous purveyor of goods to the popular taste," as T. S. Eliot described him (Elizabethan Essays [1934], p. 102). He is also jingoistic, and has the popular writer's talent for offering titillation and orthodox piety at the same time. His best play, A Woman Killed with Kindness, is a competent bourgeois melodrama, pregnant with potential ironies and subtleties which a Middleton, for example, would have exploited to the full, but of which Heywood seems blithely unaware. In Appius his essential simple-mindedness, which is not without its positive aspects, moderates his cerebral partner's tortuous art. As Lucas remarked (III, 146), under the influence of Heywood and with his help, "a better brain than Heywood's has here produced a work that at times seems to revert to the manner of a quarter of a century before, when its part-model, Julius Caesar, first appeared on a stage that had not yet lost its sense of directness and simplicity."


Page 233



Page 234


Page 235



"The Authorship of the Later Appius and Virginia," MLR, 8 (1913), 433-453; expanded as an appendix to his book, John Webster and the Elizabethan Drama (1917). Other references in this paragraph are to Lagarde, John Webster (Toulouse, 1968), I, 299; Murray, A Study of John Webster (The Hague, 1969), p. 238; and Sternlicht, John Webster's Imagery and the Webster Canon (Salzburg, 1972), pp. 192-193. Murray's table of contracted forms in Appius is wildly inaccurate, partly because the columns for i'th' and 'tis have been transposed.


The total seven is derived from examination of Paul G. Morrison's Index of Printers, Publishers and Booksellers (1955). In the years immediately after Maxey's death, John Grismond ("J.G.") appears to have been Marriot's main printer: five Marriot publications were printed by "J.G." in 1657-58. Thomas Roycroft also worked quite frequently for Marriot within the period in which Appius was printed. Maxey's books are the only ones showing typographical links to Appius.


The mixture of ordinary and swash italic capitals for speech prefixes is very similar: note the swash M, swash A, and occasional large C in each. A narrow lower case roman "a" appears frequently in both plays.


There is a head title, instead of a headline, on B1, and as the text ends on I3 this page is the last to have a headline. But I(o) takes over its three headlines from H(o), and I(i) its two from H(i). Continuity from F(i) and F(o) to H(i) is not certain. Differences in the heaviness of the inking make identification difficult, but despite a major alteration to the spacing of the headline, the lettering on H3v is probably the same as that on F1v and F2v, while the headline of F3 and F4 appears to have been used for H2, and that of F3v and F4v for H1v. If this is so, F(i) was probably printed before F(o), as on F4v, H1v, and I1v the running title is slightly closer to the left-hand margin than on F3v.


My rule-of-thumb criterion for "a decisive preference" was as follows: the test words should occur at least twice and the ratio of preferred to unpreferred spellings should be not less than three to one.


For -ness/-nesse in two cells the expected frequencies are below 5, so that chisquare probabilities, even when Yates's correction is used, are suspect; but the lowest expected frequency is above 3.5 and the chi-square of 7.8, 1 d.f., is well beyond the 0.01 level of significance, which requires a chi-square value of at least 6.635. There is a description of the chi-squared test in Anthony Kenny, The Computation of Style: An Introduction to Statistics for Students of Literature and Humanities (1982), pp. 110-119.


Dramatic Documents from the Elizabethan Playhouses . . . Commentary (1931), p. 207. Greg notes John of Bordeaux as a possible exception to these general rules.


I do not, of course, mean to imply that marginal speech prefixes and horizontal rules between speakers are to be expected in a Quarto printed from a prompt book. These features would be transformed by standard printing-house practice.


See Peter B. Murray, "The Collaboration of Dekker and Webster in Northward Ho and Westward Ho," PBSA, 56 (1962), 482-486; David J. Lake, The Canon of Thomas Middleton's Plays (1975), pp. 47-49; Tony Halsall, "The Collaboration of Dekker and Webster in Westward Ho and Northward Ho," PBSA, 72 (1978) 65-68. The linguistic and bibliographical evidence cited by these scholars apportions the plays in much the same way as Frederick E. Pierce's "three-syllable Latin word test," described in The Collaboration of Webster and Dekker (1909). Our ignorance about chronology and about Webster's part in the later collaborate plays is far from absolute. Lucas does not indulge in what Bentley stigmatizes as "sage pronunciamentos about Webster's artistic development." He simply argues, through ordinary procedures of criticism, reinforced by metrical data, that Appius fits more comfortably at the beginning or end of Webster's career, especially the end, and finds allusions in the play that appear to confirm a late date of composition. Ants Oras, Pause Patterns in Elizabethan and Jacobean Drama (1960), accepting Lucas's assignment of scenes to Webster in Anything for a Quiet Life, The Fair Maid of the Inn, A Cure for a Cuckold, and Appius and Virginia, remarks that Lucas's dating of these plays is supported by examination of the line-split patterns of Webster's share of their verse: "if Lucas's attributions and chronology are accepted, Webster's line-split patterns are seen to evolve in a strikingly consistent and natural manner" (p. 32). This claim acquires decided significance within the context of Oras's subtle and sensitive prosodic analysis. Examination of Oras's figures and graphs (pp. 58, 86) strongly suggests that Appius belongs to the end, rather than to the beginning, of Webster's career. Oras's data concern quite different metrical features from those considered by Lucas. We should not be unduly distrustful of arguments from metre: each one of Shakespeare's plays can be placed within its right period on metrical evidence alone. The main features of Webster's linguistic pattern, as described below, recur in the scenes that Lucas assigns to him in the collaborate plays.


Hoy, "The Shares of Fletcher and his Collaborators in the Beaumont and Fletcher Canon," SB, 8 (1956) -15 (1962); Lake, The Canon of Thomas Middleton's Plays (1975). See also MacD. P. Jackson, Studies in Attribution: Middleton and Shakespeare (Salzburg, 1979), where reference to other studies of this kind may be found.


Heywood's avoidance of has and does is not, of course, complete, but no play has more than a few instances, and hath and doth always greatly predominate. In Webster's share of the Ho plays he shows the strong preference for hath and doth that can be seen in WD and DM.


"The Printing of John Webster's Plays (I)," SB, 6 (1954), 117-140.


Heywood's peculiarities are described in the introductions to the Malone Society Reprints of The Captives, ed. Arthur Brown (1953), and The Escapes of Jupiter, ed. Henry D. Janzen (1977).


SB, 12 (1959), 103. There are 3 i'th' and 2 's within the 330 lines of Webster's elegy to Prince Henry, A Monumental Column (1613).


The exclusive use of between in The Silver Age and The Brazen Age suggests that both plays may have been set from transcripts in the hand of a scribe prejudiced against betwixt—unless Heywood passed through a brief phase of preferring between.


The rarity of Webster's o'th' and a'th' in the Quarto requires explanation. As o'th' occurs in the second line of the play, and at th' in the fourteenth, it seems possible that Compositor B, after showing initial tolerance of o'th' and partial tolerance of a'th', normalized to of the, and that Compositor A normalized throughout his stints. Late Quartos often fail to preserve earlier seventeenth-century forms. Another possibility is that the whole manuscript was a transcript of the joint papers, made either by Heywood or by a scribe. As amongst outnumbers among in all Heywood's plays, while Webster's preference for this form may have been less compulsive, it should be recorded that amongst occurs in Appius in I.ii, III.iv (5 times on F3v), and IV.ii (twice).