University of Virginia Library

The Shares of Fletcher and his Collaborators in the Beaumont and Fletcher Canon (I)
by
Cyrus Hoy

I

The Beaumont and Fletcher canon consists traditionally of fifty-two plays, but it has long been recognized that of these only a small number represent the work of the two dramatists in collaboration.[1] The exact number has yet to be determined, but modern scholarship is agreed that less than twelve of the vast corpus of plays which are currently designated by Beaumont and Fletcher's names are indeed products of their joint authorship. Essentially, the some forty plays that remain represent the unaided work of Fletcher, or Fletcher's work in collaboration with dramatists other than Beaumont. Chief among these is Philip Massinger, whose share in the plays of the corpus can be demonstrated beyond any doubt, but there are others, and Beaumont-and-Fletcher scholarship from Fleay to Oliphant has suggested as candidates for the authorship of the non-Beaumont, non-Fletcher, non-Massinger portions of the plays in question, the names of virtually every dramatist known to have been plying his trade in Jacobean London. Among those whose names, with varying degrees of plausibility, have been advanced, are Nathan Field, William Rowley, Middleton, Shirley, Ford, Webster, Tourneur, Shakespeare, Ben Jonson, Chapman, Daborne, and Robert Davenport.

Any investigation into the authorship of the plays which comprise the Beaumont and Fletcher canon will not, in the nature of things, consist merely in separating the work of Beaumont from the work of Fletcher. Quite apart from the problem of determining which among the fifty-two plays of the corpus are indeed Beaumont and Fletcher collaborations, there remains the very sizeable task of distinguishing the work of Fletcher from

that of his various other collaborators apart from Beaumont. To distinguish any given dramatist's share in a play of dual or doubtful authorship, one must possess some body of criteria which, derived from the unaided plays of the dramatist in question, will serve to identify his work in whatever context it may appear. On this score, the question of authorship in the Beaumont and Fletcher canon is complicated at the very outset, for with the exception of his Masque, there is no play that can with any certainty be regarded as the unaided work of Beaumont. And while the Masque may afford a good enough indication of Beaumont's metrical habits, the poetic diction in which its verse is cast tends to preclude any widespread use of the linguistic forms—especially contractions—which comprise the particular body of criteria to be used as authorial evidence in the present study. Thus, in establishing evidence that can be used in determining the respective shares of the collaborating dramatists, it is necessary to proceed from the known to the unknown, the known in this case being the unaided plays of Fletcher (which, as will be seen, can be identified) and of Massinger (about which there is no problem of identification).

My purpose in the present study is to show (1) how the unaided plays of Fletcher can be singled out from among the other plays of the canon, and (2) how the pattern of linguistic preferences which emerges from Fletcher's unaided plays contrasts sufficiently with the language practices in the unaided plays of Massinger as to afford a basis for distinguishing the work of the two dramatists one from the other. It will be noted that the tests to be applied in this and subsequent studies tend not so much to overturn the usual assignment of shares in the plays of the canon as to confirm previous attributions by a more extensive use of linguistic evidence than has hitherto been brought to bear upon the works in question. This is particularly true of Fletcher and Massinger, whose shares have been assigned within reasonably specific limits since the days of Boyle and Oliphant; though it might be argued that tests of the present kind serve to base such assignments on rather more demonstrable evidence than has sometimes been used in the past, while they tend as well to define somewhat more precisely the extent of previous attributions. In the case of such dramatists as Field, Shirley, and Ford, it will be seen in a later article that linguistic evidence provides a more certain basis for assigning their share in the plays of the canon than has yet been available.

I

The criteria which I propose to apply in investigating the plays of the Beaumont and Fletcher corpus is of a linguistic nature. By linguistic criteria I mean nothing more complicated than an author's use of such a

pronominal form as ye for you, of third person singular verb forms in -th (such as the auxiliaries hath and doth), of contractions like 'em for them, i'th' for in the, o'th' for on/of the, h'as for he has, and 's for his (as in in's, on's, and the like). There is nothing particularly new in the use of criteria of this sort, and I can claim no originality for any of the linguistic tests that I apply in the course of this study. In 1901, A. H. Thorndike drew attention to the use of the colloquial contraction 'em as a possible test of authorship.[2] Thorndike found the form to occur frequently in Fletcher, and not at all in Massinger, but since his evidence for Massinger was based on Gifford's edition—wherein 'em is consistently expanded to them—his conclusions were vitiated, as he later pointed out in an errata slip. Nonetheless, the use of 'em as opposed to them can afford a significant clue to distinct linguistic preferences, and the relevance of Thorndike's evidence remains, though it does not apply in quite such a clear-cut fashion to Fletcher and Massinger as he originally believed.

In editing The Spanish Curate for the Variorum Beaumont and Fletcher in 1905, R. B. McKerrow noted the marked preference for the colloquial form ye of the pronoun you in Fletcher's portion of that play, and W. W. Greg, in his Variorum edition of The Elder Brother, made the same observation with regard to that play. The extent to which Fletcher employs the pronominal form ye was noted independently by Paul Elmer More, who commented upon it in an article in The Nation in 1912.[3] In 1916, in an article in the Publications of the Modern Language Association of America, W. E. Farnham considered the use of such contractions as 't (for it, as in to't, on't, in't, etc.), 's (for his or us, as in on's, in's, to's, etc.), i'th', o'th', and the like, as a possible clue to authorship.[4] Most recently, in 1949, A. C. Partridge has applied linguistic evidence of this sort in his study of the authorship of Henry VIII, adding such additional criteria as is to be derived from the occurrence of the auxiliary do as a mere expletive in affirmative statements, and the use of the inflexional ending -th in the third person singular of notional and auxiliary verbs.[5] Linguistic tests of the sort that I have indicated have not, however, been hitherto applied to the question of authorship on any very considerable scale. The observations of both McKerrow and Greg were made

incidentally in the course of editing single plays, and neither ever pursued the matter further. Paul Elmer More, after examining the occurrence of ye in fourteen plays, and pointing to the possible value that such evidence might have as an indication of Fletcher's share in the plays of the canon, added that work of the sort required for any detailed study of the subject was not much to his taste, and must be left to another. Farnham, who did not consider at all the occurrence of ye, dealt with 't, 's and contractions involving the (i'th', o'th', etc.) in only eight plays. And Partridge, to the present time, has been concerned only with Henry VIII. Thus the various linguistic tests that have been proposed during the past half century have yet to be applied systematically to all of the plays which comprise the Beaumont and Fletcher canon.

From an examination of the language forms present in the plays of the canon, at least one distinct pattern of linguistic preferences is evident at once. This is chiefly marked by the widespread use of the pronominal form ye, together with the frequent use of such contracted forms as i'th', o'th', 'em, h'as, 's for his, and a markedly infrequent use of third person singular verb forms in -th. The pattern can be traced throughout fourteen plays: ye is used repeatedly from the beginning to the end of each, and this is enough to set them apart from every other play in the canon. They are: Monsieur Thomas, Rule a Wife and Have a Wife, Bonduca, The Chances, The Island Princess, The Humorous Lieutenant, The Loyal Subject, The Mad Lover, The Pilgrim, Valentinian, A Wife for a Month, Women Pleased, The Wild Goose Chase, The Woman's Prize. In no one of these does ye ever occur less than 133 times (in The Woman's Prize), and in the remaining thirteen plays its rate of occurrence is much higher than this, as high as 543 times (in The Wild Goose Chase). Elsewhere in the canon, ye never occurs with anything approaching this frequency. In certain plays (e.g., The Knight of the Burning Pestle, The Nice Valour, The Coxcomb, A King and no King), ye appears sporadically or not at all. In certain others (e.g., The Spanish Curate, The Prophetess, The False One, Barnavelt, The Maid in the Mill) the form appears, but it is to be found clustered in single acts or scenes, and does not occur throughout the length of an entire play. Thus, when ye is found to occur regularly throughout each of fourteen plays—and this in a manner that is not paralleled in any of the other thirty-eight plays of the canon—it seems reasonable to conclude that one is here in the presence of a distinct linguistic preference that can be of use in determining the work of the dramatist whose practice it represents.

To identify the dramatist whose linguistic practice is marked by the widespread use of ye is not difficult. He is clearly not Beaumont. The plays of the canon with which Beaumont's name is most closely associated—plays

like Philaster, The Maid's Tragedy, A King and no King, The Knight of the Burning Pestle—are precisely those in which ye seldom or never occurs. Nor is Massinger the dramatist in question. An examination of Massinger's fifteen unaided plays shows that, in all of these, ye occurs but twice; in all other instances, Massinger employs the pronominal form you. And the contracted forms (i'th', o'th', and the like) which are found to accompany the use of ye in the plays of the canon, are like ye itself conspicuous by their absence in the unaided work of Massinger, whose use of contractions is remarkably conservative. The assumption—a virtually inescapable one—is that the linguistic pattern characterized by a superabundance of ye's must represent the pattern of Fletcher. For three of the fourteen plays in question (The Loyal Subject, A Wife for a Month, and Rule a Wife) there is external evidence for Fletcher's sole authorship,[6] and I have no hesitation in regarding them all as his unaided work.

That they are unaided work can, I think, be demonstrated by comparing the manner in which ye occurs in them with its occurrence elsewhere in the canon. As I have already observed, in these fourteen plays the occurrence of ye, and all the linguistic phenomena that accompany its prevalence (absence of third-person verb forms in -th, frequency of such contractions as i'th', o'th', h'as, 's for his), is constant in its appearance through every act and virtually every scene. In plays of the type of The Spanish Curate and The Prophetess, however, the linguistic pattern established by the occurrence of ye is to be found only within single acts, or within individual scenes within acts, at the end of which it is abruptly broken off. In such cases, it is usually preceded or followed by a pattern of a quite different sort: one in which, first of all, the occurrence of ye is sharply reduced, and in which a decrease in the occurrence of other contracted forms is accompanied by an increased use of the verb form hath. In a very great number of cases, the linguistic pattern which accompanies the pattern established by ye is that of Massinger. A comparison of the first two acts of The Spanish Curate, the first two acts of The Prophetess, and the first act of Barnavelt, to cite but three examples, will indicate the manner in which the two linguistic patterns alternate within the same play.

It is, I think, valid to conclude that when a play, of the type represented by The Spanish Curate, demonstrates in consecutive acts and scenes two such sharply opposed linguistic patterns as those characterized by the prevalence and the absence of ye, then that play must represent the work of two separate dramatists. On the other hand, when in a play

of the sort represented by The Loyal Subject or Monsieur Thomas a single linguistic pattern is found to be maintained through virtually every scene of its five acts, there is I think no real room for doubt that that play is the work of a single author. Regarding the fourteen plays of this sort in the canon, the linguistic pattern which links them together as the work of a single dramatist is far too distinct in itself, and far too evident throughout each, to admit the possibility of a second hand intervening in their authorship. When a second hand appears in a scene that has been formerly dominated by the Fletcherian linguistic pattern, its presence is noticeable at once. If the second hand is that of a collaborator, then the pattern will be immediately interrupted, and will appear but sporadically throughout the play, as it does in such plays as The Spanish Curate and The Prophetess. If the second hand is that of a reviser, then the whole pattern will be obscured: ye's will, for the most part, disappear, or their number will be greatly reduced, and the whole texture of Fletcherian accidence is altered. The canon affords an illustration of this in The Night Walker, originally one of Fletcher's unaided plays, but revised in its extant text by Shirley.

Since the Fletcherian linguistic pattern is so pronounced and so discernible wherever his unaided work is present, I cannot consider his unaided work to be in fact represented in any play where this pattern is not evident. Thus I cannot agree with all those who have previously studied the Beaumont and Fletcher corpus in placing Wit Without Money among the plays of Fletcher's sole authorship. The linguistic pattern that emerges from this play resembles far more closely the pattern to be found in The Night Walker than the pattern which prevails in such plays as Monsieur Thomas or The Wild Goose Chase.

2

In evaluating linguistic criteria as a test of authorship, it is obvious that no linguistic form can be regarded as distinctive of a particular dramatist in any absolute sense; the extent to which he employs a given form may distinguish sharply enough his practice from that of two other dramatists, but not necessarily from that of a third. Thus emerges the necessity, in determining linguistic criteria for the work of any one dramatist, of singling out forms which are at once representative of his language preferences, while serving to differentiate his work from the maximum number of his known or supposed collaborators. The value to be attached to any piece of linguistic criteria is, in the end, completely relative: all depends upon the degree of divergence between the linguistic patterns that are to be distinguished.

With regard to the linguistic patterns which distinguish respectively

the work of Fletcher and Massinger, these, as has been observed, and as will be seen readily enough from the tables at the end of this study, are composed of language preferences which are of an essentially opposite nature. From this it is to be concluded that, in distinguishing the grammatical usage of two dramatists, a given linguistic form need not be present in an author's work to afford evidence for determining his share of a collaborated play. On the contrary, when his collaborator is found to employ that form, its absence in the work of the dramatist in question affords the best possible evidence for distinguishing the work of the two. In a play of Fletcher and Massinger's joint authorship, the fact that Massinger is known to make little or no use of the pronominal form ye constitutes evidence just as positive for his work as Fletcher's known preference for the form constitutes for his. Evidence of this sort is of the best, precisely because here the degree of divergence between the linguistic patterns that are being distinguished is as great as it can well be. The one pattern is marked by a strong preference for ye, with the use of the form averaging fifty per cent; the other reflects a tendency to avoid the form altogether.

Such clearly opposed linguistic preferences are, unfortunately, rare. The extent to which the work of two such collaborators as Fletcher and Massinger can be distinguished by the presence or the absence of a single linguistic form—pronominal ye—is, indeed, quite exceptional in the annals of the Jacobean collaborated drama. More often, such linguistic preferences as can be shown to exist in the work of two dramatists are of a more quantitative sort, with a given linguistic form present in the work of both, but present at a higher rate of occurrence in the work of one than in that of the other. In such a case, the value to be attached to any single linguistic form as evidence for authorship must depend upon the extent to which, in their unaided work, the one dramatist will tend to employ it and the other to eschew it. The less the degree of difference in the use which two dramatists make of the same linguistic form in their unaided work, the less will be its value as evidence for distinguishing their shares in a play of divided authorship. As two dramatists tend to approximate each other in their use of a given language form, the evidential value of that form is accordingly diminished.

Fortunately for any attempt to determine authorship on the basis of linguistic preferences, a single language form may be used by both of two dramatists and yet be of value in distinguishing their work in collaboration, provided only that that form can be shown to occur at a consistently higher rate in the unaided work of one dramatist than in that of the other. The value to be attached to the verb form hath, as it occurs in the unaided work of Fletcher and Massinger, is a case in point. Hath is to be

found in the unaided plays of both dramatists, yet its occurrence in any single play of Fletcher's never equals its occurrence in any one of Massinger's plays. Similarly with ye in the work of Fletcher and Field: ye occurs with some regularity in Field's unaided plays, but its occurrence there never approaches the extraordinary frequency with which Fletcher employs the form. The evidence to be derived from linguistic preferences as sharply opposed as these is second in importance only to that which is the most significant of all: the evidence that is based upon language preferences which reveal themselves in the prevalence of a given form in the work of one dramatist and its absence in that of another.

Thus far, in considering the factors that must be taken into account in evaluating linguistic criteria, I have tried to emphasize the necessity for determining the extent to which a given language form does indeed point to a clear and unequivocal linguistic preference that will serve in distinguishing the work of two dramatists. It need hardly be said that no single linguistic preference will serve equally to distinguish the work of a given dramatist from that of all others. As I have already observed, a grammatical or linguistic practice that may tend to set a particular dramatist apart from two of his fellows will not necessarily set him apart from a third. It should be obvious that no piece of linguistic criteria can be evaluated in isolation; the significance which a single form may possess for distinguishing the work of any one dramatist will derive directly from the extent to which that form is present in the work of his collaborators. The frequent use of ye, hath, i'th' or whatever in the plays of any dramatist is of no value in distinguishing his work from that of dramatists who employ such forms with equal or even approximate frequency. And no importance can be attached to the absence of a particular form from the work of any one dramatist unless it is known to occur in some noticeable degree in the work of another. The linguistic pattern that has been adduced for a dramatist on the basis of his unaided work will, of course, remain constant. However, the value of the evidence to be attached to the presence or absence of such linguistic forms as contribute to the distinctive nature of this over-all pattern will obviously shift in relation to the prevalence of those same forms within such other linguistic patterns as may be present with it in a single play. Or, stated in another way: if a given linguistic form is known to occur with approximately the same frequency in the work of dramatists A, B and C, but does not occur at all in the work of dramatist D, then while that particular form will have no value as evidence for distinguishing the work of A, B and C, it will have considerable value for distinguishing the work of any one of these from dramatist D. The use of the verb form hath in the plays of Massinger and Field will not serve to distinguish these dramatists from each other, but it

may serve to distinguish both from Fletcher. And while the absence of ye from the plays of Massinger will have very little value in distinguishing his work from that of Beaumont, who seems to have employed the form at least as sparingly as Massinger himself, the fact that Massinger almost never uses ye will serve to distinguish his work not only from Fletcher's, but from that of Field as well.

Clearly, no linguistic form can be regarded as the exclusive property of a single writer. Just as clearly, however, writers can, and often do, demonstrate a preference for certain colloquial and contracted linguistic forms (a fact that is strikingly evidenced in the case of Fletcher and Massinger) and such preferences can often serve to set apart the work of one author from that of another. In a study such as this, the problem must be to distinguish what are, indeed, an author's preferential forms, and then to determine which of these can serve to differentiate his work from that of his associates. For such a purpose, the very best linguistic evidence will always consist in those forms which a given writer can be shown to have used with conspicuous frequency, but which those with whom he collaborated can be shown to have used ever so sparingly or not at all.

3

The language forms which constitute the greater part of my evidence for authorship consist, as will have been observed, of linguistic preferences which—in a great number of cases—are made manifest in only the most minute typographical features of a printed text. In dealing with such forms, and especially when one is preparing to attach any great importance to the frequency of their occurrence, the question is naturally raised as to the extent to which an author's choice of contractions is preserved in the transmission of his text. It is well known that certain seventeenth-century compositors possessed clearly defined spelling preferences which were imposed upon whatever text they might be setting, and one wonders just how far such compositorial preferences were carried. Would a compositor, for instance, venture to impose his own preferences among colloquial and contracted forms upon a text as well? If so, then any study such as the present one is the sheerest kind of folly, for the linguistic forms by means of which one is seeking to identify a given dramatist's share in a collaborated play might have been introduced into the text by any number of unknown compositors.

There is no reason, however, to believe that compositors took undue liberties with the contracted forms in the manuscript before them; there is, on the contrary, good reason for believing that they reproduced such forms with considerable fidelity. Both W. E. Farnham and Paul Elmer

More have drawn attention to the extent to which the same contractions occur, with only slight variation, in the Beaumont and Fletcher quartos and folios. As Farnham has observed, it is clear from the verse that such contractions were intended by the author, and honoured by the printer, because they are a necessary part of the metrical structure. And equally to the point is his further observation that differences in the use of contractions in the parts of a collaborated play are "too orderly to be ascribed to the vagaries of a printer" (Farnham, op. cit., p. 332). No one can seriously consider the two linguistic patterns present in such a play as The Spanish Curate, coinciding as they do with the beginning of acts and scenes, to represent the language habits of two compositors. If such linguistic patterns did in fact represent the language preferences of two compositors, their occurrence would be found to accord with the bibliographical units of the printed text, and would not in any way be related to the act and scene divisions of the play itself. Finally, the manner in which the same linguistic preferences can be shown to persist throughout the unaided plays of a given dramatist, though the extant texts of these are the work of several different printers, affords the ultimate proof that language forms of the sort which can furnish evidence for authorship originated with the author himself, and are sufficiently preserved in a printed text. Fletcher's strong preference for the pronominal form ye is just as evident in the 1639 quarto text of Monsieur Thomas, printed by Thomas Harper, or in the 1640 quarto of Rule a Wife, printed by Leonard Lichfield, as in the remaining twelve plays of his unaided authorship, printed for the first time by Humphrey Moseley in the 1647 folio. The unaided Massinger canon presents what is perhaps an even stronger argument for this contention, for it is the product of even more diverse compositorial hands. Of Massinger's fifteen unaided plays, thirteen were published, and these represent the work of eleven printers. Yet the linguistic preferences which emerge from these are completely consistent within themselves, and what is equally striking, they are preferences which in no way contradict what we know of Massinger's language from the manuscript—in his autograph—of one of his unpublished plays. A study of the occurrence, in some one hundred plays, of the linguistic forms that are here employed as authorial evidence, convinces me that, in the greater number of cases, the use of such forms—either in the unaided plays of a given dramatist or in plays of divided authorship—is far too systematic to admit the possibility that their presence has been affected, in any truly significant degree, by compositorial intervention.

If, however, the evidence available would tend to absolve compositors from the charge of tampering with the contractions in the manuscript which they were set to reproduce, the same cannot, apparently, be said

for certain scribes in their preparation of transcripts for the use of the theatre, the printer, or a private patron. The three scribal transcripts which exist for Fletcher's unaided plays demonstrate, on the one hand, a reasonable accuracy in reproducing the linguistic preferences of the author on the part of such a scribe as Ralph Crane and, on the other, the far more erratic practice of such a scribe as Edward Knight, with the practice of the unidentified scribe of The Woman's Prize falling somewhere between the two.

Crane prepared a private transcript of Fletcher's The Humourous Lieutenant (titled in his manuscript Demetrius and Enanthe). Since his text contains some seventy-five lines not present in the text of the first folio, the supposition is that Crane's transcript derives from Fletcher's original manuscript, whereas the folio text represents a prompt-book containing theatrical abridgements. In his transcript, Crane introduces some thirty-four ye's not present in the text of the folio, while he omits some fourteen ye's which the folio text exhibits, but the difference of approximately twenty ye's in the total occurrence of the form in the two texts is not great. It speaks, in fact, well for the care with which Crane reproduced his copy when it is compared with the wide divergence in the occurrence of ye in the two extant texts of another of Fletcher's unaided plays. Bonduca (For a careful study of Crane's characteristics as a transcriber, see R. C. Bald, Bibliographical Studies in the Beaumont and Fletcher Folio of 1647, p. 95, but more especially his edition of A Game at Chesse by Thomas Middleton, pp. 171-173.)

The text of Bonduca is extant in a scribal transcript, prepared by Edward Knight, the book-keeper of the King's Company, from Fletcher's foul papers, and in the text of the 1647 folio, printed from the prompt-book. In the folio text, the pronoun ye is used 352 times; in Knight's transcript, the occurrence of the form has been reduced by more than half, to 147 times. The variation in the two texts in this respect is of significance because, on the basis of the first folio, the percentage of ye's to you's is the highest to be found in any play of Fletcher's unaided authorship. If, however, Bonduca survived only in Knight's manuscript, the play would present the lowest percentage of ye's to you's in all Fletcher, with the occurrence of the form falling markedly below its normal frequency in his unaided plays.

There is evidence of scribal intervention affecting the use of ye in another Fletcher play, The Woman's Prize, and there is good reason to suppose that the scribe responsible for the reduction in the occurrence of the form is once again Knight. Like Bonduca, The Woman's Prize is extant in two texts: an undated private transcript, prepared by an unidentified scribe, and the text of the 1647 folio. In the first folio text, ye occurs but 84

times, a number far below the usual occurrence of the form in Fletcher's unaided work. In the manuscript, ye is found 133 times, and while this still represents the lowest occurrence of the form in Fletcher, the increase of 49 ye's makes for a rather more satisfactory basis for regarding the play as Fletcher's own.

There is external evidence which almost certainly has some bearing on the first folio text of the play and the linguistic forms which it exhibits. On 18 October 1633 the Master of the Revels, Sir Henry Herbert, suppressed a performance of The Woman's Prize (he refers to the play by its alternate title, The Tamer Tamed), which the King's Company had scheduled for that afternoon. On the following morning the prompt-book was brought to him, whereupon he proceeded to purge it of "oaths, prophaness, and ribaldrye" (Herbert, p. 20). The play, Herbert explains, was an old one, evidently licensed during the Mastership of one of his predecessors, which the King's Company had sought to revive, under a different title, without applying for a new license. Herbert was thereby deprived of his licensing fee, a matter about which he felt strongly, as he indicates in the entry in his Office Book, though he advances another and more public-spirited reason why old plays should not be restaged without the allowance of the Master of the Revels: "they may be full of offensive things against church and state; the rather that in former times the poetts tooke greater liberty than is allowed them by mee" (p. 22).

The upshot of the whole affair was that two days later, on 21 October, Herbert returned the prompt copy, properly expurgated, to the players, accompanied by a note to Edward Knight enjoining him to "purge [the actors'] parts, as I have the booke." The players' capitulation to Herbert's demands was complete; two of their chief members apologized for "their ill manners" and asked his pardon, and the following month Fletcher's The Loyal Subject, which had been licensed by Sir George Buc in 1618, was submitted to Herbert for re-licensing.

Mr. R. C. Bald, in a most valuable discussion of the two texts of The Woman's Prize in his Bibliographical Studies in the Beaumont and Fletcher Folio of 1647 (p. 60), points out that, while "the manuscript omits two whole scenes (II.i and IV.i), two passages of fourteen and seven lines respectively, and eight of three lines or less" that are included in the folio, the manuscript exhibits, on the other hand, "eleven passages . . . varying in length from half a line to nine lines," which the folio omits. It is Mr. Bald's opinion that the manuscript gives the play, which was originally performed in 1610 or 1611, "as cut for acting before Herbert's time," while "the folio gives a fuller version of the play, but observes the cuts that were made by Herbert in 1633." To observe the cuts that Herbert demanded, it does not seem unreasonable to suppose that a new prompt-book was drawn

up, and if a new prompt-book was prepared, it seems clear enough from Herbert's note of 21 October that the task would be performed by the book-keeper Knight. From Knight's transcript of Bonduca we know how the Fletcherian ye diminished under his hand (see Bald, pp. 99-100), and I can only account for the small number of ye's in the folio text of The Woman's Prize by supposing the manuscript from which that text derives to have been prepared by him. With regard to the scribal transcript, the supposition would be that the scribe responsible for it has been somewhat more faithful in reproducing the language forms that must have stood in the original. Since the manuscript text reflects more clearly than the folio the quality of the Fletcherian original, I have used it as the basis for the statistics set forth for The Woman's Prize in the linguistic tables at the end of the present study.

The possibility of scribal intervention should perhaps be considered in relation to two other of the plays which can be regarded as Fletcher's unaided work, Rule a Wife and A Wife for a Month. These, apparently Fletcher's last plays, exhibit after The Woman's Prize the least number of ye's of all the fourteen plays that I consider to be his. The first folio text of A Wife for a Month gives clear indication of author's foul papers, but it is not impossible that the text has derived from a not too careful transcript of these. Two speeches are printed in alternately abridged and expanded versions, and there is a bad tangle in the second scene of the fourth act which clearly would have had to be set to rights before the manuscript in back of the first folio text could have been used as a prompt book. But if Knight's transcript of Bonduca is any indication of his work for a private patron, he would not have been above letting such difficulties stand in a text which he prepared, if it were not to serve as a theatrical prompt copy. And if the total number of ye's still present in the text of A Wife for a Month (176) does indeed represent a reduction from the original number, Fletcher's favourite pronoun has here been given much the same treatment as Knight accorded it in his Bonduca manuscript.

The substantive text of Rule a Wife, that of the 1640 quarto, probably derives, as Prof. Jump has suggested, "either from a prompt-book or from a manuscript directly descended from a prompt-book."[7] The play was licensed for acting by Sir Henry Herbert on 19 October 1624, and four months later, on 8 February 1625, Herbert re-licensed The Honest Man's Fortune, for which Knight had prepared a new prompt book that is extant in his autograph. It would seem likely, then, since he was actively employed by the King's Company at this time, that Knight prepared the prompt-book for Rule a Wife as well. There is evidence of a sort in the quarto of Rule a Wife that might be considered to link it with his

work. The chief feature which the quarto and the Bonduca manuscript have in common is a frequent occurrence of the contraction 'um (for 'em). Since Knight employs 'em throughout his manuscript of The Honest Man's Fortune, 'um is not likely to represent his own linguistic preference. And since the form is 'em throughout the 1640 quarto of Rollo, Duke of Normandy, printed in the same house and in the same year as Rule a Wife, it seems improbable that the 'um spelling is compositorial. I regard it rather as a Fletcherian form which Knight has reproduced forty-six times in his transcript of Bonduca, and—perhaps—thirty-two times in the manuscript behind the quarto of Rule a Wife. Seventeen times in the Bonduca manuscript, Knight uses the spelling hir for her. The hir spelling occurs twenty-nine times in the quarto of Rule a Wife, and it is the prevalent spelling throughout the manuscript of The Honest Man's Fortune. The evidence is admittedly not great, but combined with the fact that Knight was the probable person to have prepared a prompt-book for the King's Company at this period, it seems at least possible that the diminished number of ye's (213) in the quarto of Rule a Wife may be traced to his intervention in the transmission of the text.

4

The following tables set forth the rate of occurrence, in the unaided plays of Fletcher and Massinger, of those linguistic forms which are of value in distinguishing the respective shares of the two dramatists in plays of divided authorship. I have omitted The Faithful Shepherdess from the number of Fletcher's unaided plays, for although it is undoubtedly Fletcher's own, linguistically at least it has nothing in common with any other of his unaided works. Its language is that of pastoral poetry, uncol-loquial and somewhat archaic. It abounds in linguistic forms (most notably the third person auxiliary forms hath and doth) which Fletcher seldom or never uses in his other unaided plays, while all the most distinguishing of his colloquial forms are either completely absent, or present in only a negligible degree. Nothing could be more misleading than to regard the language of The Faithful Shepherdess as typically Fletcherian.

Of the linguistic forms cited in the tables below, ye is much the most important for purposes of authorial evidence. Since Fletcher employs the form as both subject and object, direct or indirect, in either singular or plural number, the rate of its occurrence in his unaided plays is very high. In the fifteen unaided plays of Massinger, the form occurs but twice. Contractions in y' (y'are, y'ave and the like) are much less frequent in Fletcher, and are of no value in distinguishing Fletcher's work from Massinger's. The two occurrences of y'are in Fletcher's Monsieur Thomas, Rule a Wife, Bonduca, and The Pilgrim, for example, are matched by the

two instances of the form in Massinger's The Bondman. The single instances of y'ave and y'have in, respectively, Fletcher's The Chances and Bonduca are paralleled by single appearances of the same forms in, respectively, Massinger's A New Way to Pay Old Debts and The Guardian. There is nothing to distinguish Massinger's use of contractions in y' from Fletcher's, and I have not included them among the forms cited in the following tables. Regarding the verb form hath, there is a distinct difference in the Fletcher-Massinger usage. In Fletcher, the form never occurs more than 6 times in a single play, and in two plays it occurs not at all. In Massinger, on the other hand, hath never occurs less than 8 times in any one play, and generally it is found a good deal more often than this—as often as 46 times in a single play. Doth comes in only one of the fourteen Fletcherian plays listed below, but since it appears but 5 times in Massinger, the distinction in the practice of the two dramatists on this point is not great. The contraction 'em appears in all of Fletcher's unaided plays, from 23 times in Women Pleased to 130 times in The Loyal Subject. In certain of Massinger's plays, it will be noted, 'em is to be found occurring as frequently as it does in certain of Fletcher's. But it seems significant that all of these (e.g., The Picture, The Guardian, The City Madam) are late plays, licensed for acting after Fletcher's death in 1625.[8] In Massinger's early plays, which would presumably reflect his language practices at the time of his collaboration with Fletcher, 'em is used a good deal more sparingly than in the unaided plays of Fletcher or in the later work of Massinger himself: 7 times, for example, in The Parliament of Love, 9 times in The Renegado, 12 times in The Duke of Milan. I tabulate the occurrence of the form for whatever value it may have as a piece of corroborating evidence for distinguishing the work of the two dramatists.

The evidence to be derived from the contraction i'th' is, on the whole, good. Despite the fact that the 7 occurrences of the form in Fletcher's The Island Princess are equalled in Massinger's The Guardian, the form is found at least 4 times in all of Fletcher's plays, where it may appear as many as 28 times, while it is found in but 5 plays of Massinger's, and in none of these more than 7 times. It may be worth noting that the five plays in which the form occurs are late ones, and that i'th' appears in no

play of Massinger's written before Fletcher's death. A form, however, which Massinger tends to employ occasionally, but which occurs only a single time in Fletcher, is the contraction i'the. The contraction o'th' affords evidence of a sufficiently clear-cut sort: the form occurs at least once in all fourteen of Fletcher's unaided plays; it occurs not at all in Massinger. The colloquial form a (for he) is found in six of Fletcher's plays, but appears in none of Massinger's. Of a similar nature is the contraction 'is (for he is), present in five of Fletcher's unaided plays, but not present in Massinger. H'as (for he has) is found at least twice in each of the fourteen unaided plays of Fletcher, but it occurs only a single time in Massinger. The contraction t' (for to, before a following vowel or h) affords evidence of a sort for Massinger; it occurs at least once in ten of his fifteen unaided plays, but is found only a single time in Fletcher. Contractions involving 's for his occur chiefly in Fletcher following the prepositions in and on. There are single instances in Fletcher of enclitic 's for his with four other prepositions (at, for, to, up); with an adverb (than); with a verb (strike). In Massinger, 's for his occurs but three times: twice in the contraction in's, once in the contraction of's. Only the uses of 's for his with in and on have seemed worth recording in the tables that follow.

As for contractions in 's for us, these occur most commonly in Fletcher with the imperative verb form let. I find only two occasions in which Fletcher has used enclitic 's for us after other notional verbs (put and make); elsewhere, he uses the form only after the preposition on (5 times). In Massinger, 's for us is used only in the contraction let's, and even this quite normal form Massinger uses very sparingly. It is the only contraction in 's for us that I have recorded below. The enclitic use of 't for it with both prepositions and verbs (in contractions such as in't, on't, for't, to't, is't) is standard in the work of Elizabethan and Jacobean dramatists, and contractions of this sort are of no worth in distinguishing the work of Fletcher and Massinger, for their rate of occurrence in the work of each is virtually identical. In the following tables I have recorded only one form in 't for it, the contraction of't, and this only because the form does not appear in Fletcher, while it occurs from one to nine times in thirteen of the fifteen unaided plays of Massinger.

To summarize the chief features of the linguistic patterns of Fletcher and Massinger: the Fletcherian pattern is one which is marked above all by the constant use of ye; one which exhibits a strong preference for the contraction 'em to the expanded form them; one which regularly employs such other contractions as i'th', o'th', h'as, and 's for his, and which makes sparing use of the third person singular verb forms hath and doth. Stated

Linguistic Tables for the Unaided Plays of Fletcher and Massinger[*]

                                                                 

												's(his)	's(us)
	ye	hath	doth	'em	i'th'	i'the	o'th'	a	'is	h'as	t'	in's	on's	let's
Fletcher
M. Thom.	343	6		27	9		6			2		3		10
R. W.	213	2		35[†]	20		12		5	7		3	1	12
Bon.	352	1		95	14		10	4		10		4	1	27
Chan.	290	2		44	12		4	4		10		2		20
I. P.	258			64	7		8	1	8	5		1		14
H. L.	367	5		80	28		11	2		11		3	3	11
L. S.	424	3		130	13		10			4		2	1	10
M. L.	308	6		25	16	1	4	15	4	7		3		17
Pilg.	400	3		62	15		9			9		7		18
Valen.	412	4		71	12		8	2		4		2		16
W. M.	176			41	4		1		1	2		5	1	17
W. P.	288	3		23	15		16			6		4	2	3
W. G. C.	543	1		61	8		6		1	3	1		1	15
W. Pr.	133	4	3	58	14		21[‡]			3		3		10
Massinger
D. M.		46		12		1					8			1
Bond.		8	1	15							1			2
P. L.	1	21		7										1
R. A.		28		14		2					3			5
Pict.		35		52	5	3					3			1
Ren.		21		9										3
Bel.		36		26		1
E. E.		31	1	26							2
M. H.		25		31		5					2
N. W.		16		36	2	1				1	2			1
G. D. F.		26		15
U. C.		23		16										1
B. L.	1	41	3	21	3	2					5	2		1
Guard.		26		47	7	4					4			3
C. M.		19		46	1	1					1

numerically, it is a pattern in which the average rate of occurrence for the forms in question is as follows:                  

Contraction	Average occurrence per play
ye	322
hath	3
'em	59
them	8
i'th'	14
o'th'	9
h'as	6
's (his)	5

The full significance of these figures can best be realized when they are compared with the average rate of occurrence for the same forms in the unaided work of Massinger. There ye occurs twice in fifteen plays. Hath occurs at an average rate of 27 times. In the seven plays of Massinger's sole authorship written before Fletcher's death, and so reflecting most nearly the author's linguistic preferences during the period of his collaboration with Fletcher, 'em is used an average of 12 times per play, them an average of 23 times. The contraction i'th' is found 18 times in five of Massinger's unaided plays, all of which date after the death of Fletcher. O'th' does not appear in any of Massinger's unaided plays; h'as is found but once (in a post-Fletcher play); 's for his occurs twice (both times in a play written after Fletcher's death). In the linguistic pattern which emerges from the unaided plays of Massinger written during Fletcher's lifetime, it can fairly be said then that the Fletcherian ye has no parallel; that Massinger's average use of hath is nine times greater than Fletcher's; that the Fletcherian preference for 'em to them is precisely reversed in Massinger; and that the contractions i'th', o'th', h'as, and 's for his are completely absent from his work at this period. The linguistic patterns of the two are as nearly opposite as they could well be.

Notes