University of Virginia Library

Establishing Shakespeare's Text: Notes on Short Lines and the Problem of Verse Division
by
Fredson Bowers

Because of its abrupt break with regularity, the Shakespearean short line within a pentameter speech is immediately noticed by the ear and, in a reading edition, by the eye. Some problems are associated with these short internal lines, principally whether they are authoritative or the result of compositorial or scribal mislineation, or whether they may be part of a general mislineation that requires the isolation of some adjacent line of the text as the internal short line intended. Such internal lines, usually inserted for special rhetorical effects, are relatively few in comparison with Shakespeare's use of short lines in other and more varied circumstances. Thus they form no part of the present inquiry.

Shakespeare uses short lines most frequently to divide a pentameter between two speakers. Next in order they may be used to end a speech but with no metrical link with the following speech, which begins with a full pentameter or with prose. Alone or in a series, single short lines without metrical relationship either to each other or to what precedes or follows may punctuate otherwise regular pentameter dialogue. Single short lines of what is presumably verse (but often subject only to arbitrary assignment as verse or as prose) may appear between prose speeches or between a verse and a prose speech. Least commonly, an unlinked metrically short line may begin a regular pentameter speech.

Not one of these categories but presents a difficulty to an editor that in some manner may be connected with the basic problem of the linked, or divided, line in which a pentameter is split between two or more speakers. It is idle to comment that on a practical level many although not all problems would disappear if typographically the dialogue were edited as it appears in Renaissance printed editions, which in this respect follow the conventions of the dramatic manuscripts of the period. That is, the two or more parts of a verse divided between speakers are

each printed flush with the left margin so that there is no visual linking of the parts to indicate the pentameter that they represent. Steevens was the first Shakespearean editor to indent the completing sections of divided lines so that the metrical structure was established and line reference was possible to the single verse. Despite the problems raised, the advantages of this indention are obvious. Without it, no line count for a play could be completed except on such a mechanical basis as in the new through-line-numbering system (TLN) employed by C. J. K. Hinman in his Norton facsimile of the First Folio. This arbitrary count of lines as printed in the Folio is invaluable for standard reference but it is useless for critical data since the vagaries of the Folio (and Quarto) compositors' treatment of the text offer no statistically valid count of the verse lines in a play insofar as these may be identified as units designed by Shakespeare.[1] Thus the 1793 system of indention has enabled editors to number verse lines in a critical instead of a mechanical manner, for with it one can distinguish independent separate short verses from those short lines that link to form an authentic pentameter. Critical line reference to metrical units can also be made exact by uniform editorial decision which specific short lines should link when alternative arrangements exist. Not to be ignored is the pleasure given a reader by the visual indication of the flow of the verse.[2] Finally, and sometimes a mixed blessing, on the basis of assumed short-line linkage between the end of one speech and the beginning of the next an editor may succumb to the temptation to reline the verse in order to secure what he takes to be the authorially intended arrangement that had been corrupted in the printing shop or in an antecedent scribal manuscript.

Modern editors have devoted more attention to purifying the language of the plays by emendation (or de-emendation) than to a critical examination, with a view to mending its faults, of the actual workings of Steevens' metrical system. Comparison of the most esteemed twentieth-century editions of Shakespeare indicates that—transmitted through the

basic Globe edition—Steevens' metrical distinctions, as marked by his choice of linked lines and his relining of the verse to secure certain linkages, have been generally accepted without the scrutiny that would have exposed inconsistencies and serious anomalies that demand editorial attention. Hence the modern effort to 'establish' Shakespeare's text has touched very little on the deficiencies of some parts of the traditional metrical arrangement. (Indeed, even the New Variorum may not record editorial variation in the linking of lines unless relineation is involved.) The failure of editorial curiosity to revise anomalies as well as positive errors is seemingly due to a lack of awareness that a problem exists, this intensified by the lack of any handy analysis of the metrical system that concerns short lines, whether or not part of a divided verse. It is a fact that there has been singularly little discussion of Shakespeare's metrics in recent years save at an aesthetic level that is of no practical assistance to a puzzled editor anxious to reproduce Shakespeare's intentions. Yet until the system of divided lines is fully investigated as a source of textual emendation[3] and the information applied to the necessary revision of standard texts, scholarship will be in no position to make a true start on a new edition that in any sense can begin to approach definitiveness.[4]

It is essential to learn whether Shakespeare had certain favorite patterns, a recognition of which might solve at least the more obvious problems. The purpose of the present study is to make a start on the whole complex matter by distinguishing certain patterns and offering a few limited answers, limited at least as applied to the solution of the most acute problems. That there are problems may be illustrated by the following examples from Measure for Measure:

The above quotations, drawn from the Folio, have been given the Globe's indention and system of lining, one that has been followed by all the recent editors I have looked into from the Globe to the Riverside. In the first, Claudio's short line has been linked with the last line of Isabella's speech instead of with the first line of her next speech. But in the same scene the procedure is reversed in the second and third quotations. Here the Globe and other editors have instead linked the short speech with the opening line of the next speech. Yet there is more inconsistency. For example, not many lines earlier the Globe in the fourth illustration indents Claudio's short speech to link it with the preceding speech, in the process creating a semi-run-on beginning line for Isabella. No discernible principle seems to have dictated why in two cases editors have linked the short separate speech with the preceding speech and in the other two with the beginning short line of the next when either way is metrically equivalent. We shall see that in fact only in the second and third quotations is the verse correctly divided by modern editors.

Before one can understand the conditions under which Shakespeare would link certain short lines but not others, one must trace the patterns of unlinked verse lines of less than pentameter length. The three major categories are (1) unlinked part lines ending a speech, (2) unlinked part lines beginning a speech, and (3) unlinked independent short lines that are themselves a complete brief speech.

Non-linking short lines concluding a speech occur in two major situations depending upon whether it is or is not theoretically possible for them to be completed by the short line of the next speech. Unquestioned short lines that cannot be completed are most frequently found ending a speech before a speech that begins with a full pentameter:

They also appear ending a verse speech before prose: Shakespeare may utilize the special effect of a short line to lend naturalness to a scene ending:

The incidence of these short concluding lines rises with the years, although erratically. The Two Gentlemen of Verona has about seven,[5] of which one is an exit line; the Dream (Shakespeare's most regular play) has only one; Romeo and Juliet has thirteen, of which five are exit lines and one before prose; The Merchant of Venice ten; As You Like It fourteen; Much Ado eleven; Twelfth Night fourteen (five before prose); Julius Cœsar thirty-one (three as exit lines and two before prose); Measure for Measure about forty-two (seven before prose); All's Well twenty-two; Antony and Cleopatra about sixty; Coriolanus about forty-seven; Winter's Tale about twenty-five; Cymbeline about twenty-eight; Tempest about twenty-nine.

Unlinked short lines beginning a pentameter speech are much rarer. (Ben Jonson, incidentally, declined to write them.) In situations where no possibility exists that they could complete a preceding line, they appear after the full pentameter line of a preceding speech:

They may also occur after a prose speech: They may also be used for their special effect as an entrance line, especially at the start of a scene: The following figures are only approximate because of the difficulties in distinction, but again the incidence of beginning unlinked short lines has a general tendency to rise. The Two Gentlemen of Verona has about eight, curiously enough; the Dream none; Romeo and Juliet three; Merchant of Venice about three; As You Like It perhaps two; Much Ado probably none; Twelfth Night about three; Julius Cœsar about five; Measure for Measure about seven; All's Well about seven; Antony and Cleopatra about twelve; Coriolanus about thirty-seven; Winter's Tale about twelve; Cymbeline about thirteen; Tempest perhaps five or six.

These statistics showing the general prevalence of independent short lines ending speeches but the relative rarity of unlinked short beginning lines carry over, significantly, into another situation—that in which an independent part-line separate speech either completes a pentameter begun by the short ending of the preceding speech, or else begins a pentameter completed by the short opening of the next speech.

Even in these circumstances of linked speeches, the preponderance of short ending versus short beginning lines is marked. For instance, in The Tempest there are fifty-two such ending versus seven such beginning linked lines; in Cymbeline as many as seventy-three short ending lines completed by an independent part-line speech as against only eight short linked beginnings; in Coriolanus the figures are eighty-three as against sixteen. Once more the statistics are five or more to one that when linking part lines are present Shakespeare will end a speech with a short line than that he will begin a speech with less than a full pentameter.

Indeed, the only frequent use of part beginning lines is this linkage (more and more often as time passes) with short endings to induce a flexible and varied flow of pentameter dialogue verse. The division of a pentameter between two multi-line speeches grows on Shakespeare remarkably. An exact count for any play is impossible to make in the present stage of our knowledge, since it may often be a matter of opinion whether two short lines are independent or were intended to join to form a pentameter. A rough count shows something like this (the figures in parentheses are the through-line-number counts, less epilogues, as found in Hinman): Dream (2152) about 35 or fewer; Merchant (2738) about 80 or fewer; Romeo and Juliet (3185) about 80; Julius Cœsar (2730) about 125; Much Ado (2684) about 35; As You Like It (2775) about 35; Twelfth Night (2558) about 42; All's Well (3072) about 135; Measure for Mesaure (2938) about 150; Antony and Cleopatra (3636) somewhere between 440 and 460; Coriolanus (3838) about 390; Winter's Tale (3369) about 335; Cymbeline (3819) about 400; Tempest (2319) about 220.

Before Edward II with its 29 divided pentameters, Marlowe's mighty line admitted practically no divided verses in texts that can be authenticated.[6] Similarly, as Shakespeare's verse increased in flexibility the number of divided verses expanded. For example, in the 83 verse lines of Cymbeline I.i—an expository dialogue between two gentlemen—there are 10 divided pentameters, 2 independent short lines, and 71 regular verses. More to the point, the 10 divided verses appear in 19 speeches, which means that except for a separate short-line speech at I.i.62 (Globe), TLN 72, each speech ends with a part line completed by the opening part line of the next speech. (The second separate short line is at the exeunt.) The second scene contains 178 lines and 39 speeches, linked by some 27 divided pentameters. Only six speeches in this scene are not

linked: of these, five (including the exeunt speech) end with a full pentameter and one with a trimeter. No independent short lines appear. These first two scenes, therefore, totalling 261 verses, contain as many divided lines as occur in the entire play in The Two Gentlemen of Verona, or Much Ado, or As You Like It. Since the same distinction of early and late holds in the tragedies, it seems clear that starting about the time of the so-called problem plays this feature of Shakespeare's dramatic blank verse began to expand rapidly. By the last plays it had doubled in frequency from the problem plays and was close to ten times the frequency of the early comedies.

As one would expect, accompanying the rise in this incidence of division is a rise in the number of irregular divided lines either in the sense that they are deficient or hypermetrical (and thus perhaps intended not as linked but as separate lines), or else in the sense that choices exist as to which two short lines were intended for linking and which third was to remain independent as in the illustrations from Measure for Measure above. The Dream has not a single such problem within its complete metrical regularity. The Two Gentlemen has about five possible metrical linking irregularities or questions of association in division, Twelfth Night about eight, but Measure for Measure close to thirty, Coriolanus about twenty. Cymbeline is unusual in posing no more than about fifteen possible difficulties, but Antony and Cleopatra (the metrically most irregular of Shakespeare's plays) is equally unusual in offering more than forty-five.

Outside of a few specific exceptions, no reason exists why in most circumstances the statistics about the incidence of short lines in the two different positions should not suggest the principle by which the usual problems may be resolved. This is, whenever a choice exists as to which two short lines should form a pentameter and which third should stand by itself, the odds strongly favor the principle that in ordinary circumstances a short line beginning a speech should be linked in preference to the short ending line of a speech. The general pattern of Shakespeare's versification indicates that unlinked final short lines are common whereas unlinked short beginning lines are rare and usually recognizable as inserted for special purposes. This general pattern observable in situations where no choice exists can be applied with confidence to the selection of the short lines to link when a metrical option is present. If so, then instead of the traditional arrangement of the first example from Measure for Measure given above as

the correct division should be Correspondingly, instead of the correct division should be Editors are also at fault in another option: whereas the division should be:

Because of the great predominance of short closing over short opening lines in dialogue where no linkage is in question, the burden of proof may be said to fall on an editor who associates an intervening short line with the preceding part line closing a speech instead of the following part line beginning a speech, as in the illustration above from Othello. On the other hand, Shakespeare does use a small number of short opening lines in unlinked dialogue and it is worth surveying these to see if they have any special identifying characteristics as against the purely random (except for occasional exit speeches) use of short closing lines. Analysis discloses that the majority indeed shows a pattern indicating a structural or rhetorical purpose. For example, the break may be syntactical as in

Intervening stage-business may justify a short line:

In the next illustration a change of address (from command to soliloquy) fortifies the call to Lucius:

Common are changes of address from one character to another:

Whatever the situation, almost invariably some distinct syntactical pause, some break, with or without stage-business or change of address, is made after the opening short line:

The urgency of the special effect that Shakespeare sought from the opening short line of a speech leads him, on rare occasions, into the unusual metrical pattern whereby normally unlinkable short lines end and begin two successive speeches:

The first comes under the head of change of address, which frequently calls for a short line, but the second is purely rhetorical.

In contrast, the short unlinked beginning of a speech without a special purpose is so rare that many plays have none, and perhaps no more than one or two can be found legitimately in any single play among the rest.

These patternless examples of run-on lines may be due as much to carelessness as to anything; no principle is involved and they are so extremely uncommon as to constitute sports that can have no evidential value in attacking the problem of divided short beginnings except to discourage any attempt by an editor to imitate them in his arrangements.[9] On the other hand, the special uses that Shakespeare makes of independent opening part lines undoubtedly have an intimate connection with problems of linking, and their evidence must be heeded.[10]

Unfortunately, Shakespeare does not restrict these special effects to short lines alone; indeed, the number of full pentameters containing in their first half commands, exclamations, changes of address, stage-business, and the like vastly outnumber the times that Shakespeare chooses to emphasize the same material by an isolated part line. One may open almost any play at random and come on such examples as these:

Of even greater pertinence, the rather uncommon cases of abrupt short beginning lines emphasized by being set off as unitary single lines

are also notably outnumbered by the frequency with which Shakespeare links just such lines with the ending of the preceding speech; or, just as commonly, and substantially the same thing, links them with a preceding short single speech:

These special-purpose short beginning lines, then, are so frequently linked as by no means always (or even usually) to require treatment as a separate verse. Yet even so, the independent ones offer the best evidence we have to assist an editor when problems of arrangement arise. For instance, if in two multi-verse speeches Shakespeare for a special purpose may (infrequently) end one speech with a short line and begin the next speech with an unlinked short line too,[12] then it stands to reason that on a few of the occasions when such lines could be metrically linked it is theoretically possible that Shakespeare might have intended them to stand as separate verses even though beginning a speech. Two examples may be cited where the meter encourages, if not enforces, one particular linking as against another possibility. The first is correctly lined by all editors:

Inconsistently, in the next example, the Globe and other editors insist on an arrangement that creates a hexameter: The passage should be rearranged as: The abrupt rhetorical switch between Bianca's ironic response to Cassio's reluctance and her new proposal for him to accompany her justifies the short line beginning a speech, especially when it releases the verse from an unnecessary hexameter.

However, situations do arise where metrically a free option exists to link a separate short speech either with the preceding or the following

short line. The easiest cases to identify and to avoid are those that would produce a short beginning line without any special purpose. The failure to apply this test of significant purpose leads the Globe and its trail of captive editors into the frequent violation of Shakespeare's clearest-cut patterns. This must be wrong, for 'I muse my mother' is a natural linking line not a rhetorically motivated independent one. This forms a pentameter with the suppression of a stressed syllable at the caesura, a form of line occasionally met with in Shakespeare's regular verse (see below). On the contrary, editors almost wilfully have adopted an arrangement that has two defects: first, in order to link 'Be thus to them' with 'You do the nobler', they must suppress a complete iamb at the caesura, a possibility not certainly allowable, and definitely not one to venture on if a suitable alternative is present. Second, after this difficulty the run-on line 'I muse my mother' is then assigned as a complete verse, although this is contrary to Shakespeare's normal patterned use of such lines, as we have seen. Correspondingly, in the next example, only Alexander arranges the passage correctly, involving the caesural suppression of an accented syllable, whereas the Globe, Kittredge, New Arden, Pelican, and Riverside create a short beginning with a run-on line of a sort which must be avoided:

A more extended example of the Globe and other editors' mistaken arrangement in Antony and Cleopatra comes in Act II, scene vii.

In the Folio Menas' speech is clearly mislined, but we are now able to recognize this for what it is, not a mistaken extension of a short line but instead Compositor B's frequent habit of breaking in two a pentameter that was too long for his measure. The two part-lines in the Folio may be joined to form an acceptable verse, whereas the Globe's relining to create a short opening verse 'For this' has little justification from Shakespeare's customary use of such short openers for special effects. All one need do is to pronounce 'never' as nere and one has a pentameter composed of an iamb, an anapest, and three iambs. The Globe's decision that 'Be a child o' the time' should complete Cæsar's speech makes Cæsar's 'I could well forbear 't' a short beginning line. There is something of a rhetorical break after 'I'll make answer' but not a sufficiently sharp one to enforce an independent short line to mark the shift. The Globe's division 'fouler. | Be a child' gives a mechanically regular line of iambs with a probable concluding anapest; but the lyric caesura forces an unnatural stress on 'Be' as the second syllable of an iamb, whereas if, instead, a divided verse begins with Antony's advice, then 'Be a child' may be given the more natural anapestic stress. With complete certainty one can insist that 'Come, let's all take hands' has no justification as a separate verse[13] and with some confidence arrange it to complete a pentameter with Pompey's 'Let's ha't, good soldier.'

In contrast, even though the metrical options permit either arrangement so that a short beginning line is not obligatory, some situations might be thought to call for a venturesome editor to suppress the odds that so strongly favor the completion of a divided verse by the opening part-line of a speech. Required (not imaginary) stage-business involving the use of properties that calls for a real pause while action develops may lead Shakespeare to begin a speech with a short line, as in the following example, where there is no option to the arrangement:

In the next example, somewhat later in the same play, either the first and second or the second and third short lines may link with equal metrical regularity, and the conservative position would of course be that the divided verse should read 'Heere 'tis: . . . | How? let me see.' But the stage-business of handing over the ring and Proteus' examination of it could be thought to justify the emphasis of his short line as a complete verse in the traditional arrangement:

When such technical matters affect the arrangement, an editor might feel free to adjust the indention, in case of metrical ambiguity, to what are demonstrably Shakespearean patterns. With just the same situation, except that it involves the less compelling pattern of a change of address, the Globe and editors opt for the conservative position in As You Like It, 2313-17:

Editors are doubtless correct in this arrangement although a lining is possible that takes account of the change of address:

Traditional arrangements (which produce an excessive number of unpatterned short opening lines) are likely to show serious inconsistencies when the separate short opening verse might be created more for rhetorical reasons than to mark a structural pause such as the issue of commands at exits, a marked change of address, or the intervention of stage-business. Tradition may sentimentally support the emphatic short line in Antony's farewell speech, and there is a possible change of address for a moment to encourage such arrangement despite the metrical option that would conservatively suggest that 'I am dying' should be linked with 'a heavy sight.'

In the next quotation the Folio lining is by no means impeccable, but there would seem to be no reason for the Globe and other editors to ignore what hints it may offer and for them to reline the Folio to produce a short ending rather than a short beginning line, when the beginning is tolerably emphatic: Of course, if the Folio had been lined like the Globe, it need not have been changed to produce an independent short line. Similarly, in the next example there is even less reason to reline and in the process to create a divided hexameter when Menenius' exclamation makes a more than acceptable short line by itself, even at the cost of ending and beginning adjacent speeches with unlinked short lines. Examples have been provided above (as in Oth. 373-375 and Cor. 2985-88) of clearcut Shakespearean occurrences of this unusual pattern. For a rhetorical situation like this, the Folio's two short lines are by no means so anomalous as to require indention to form a hexameter. In the next illustration, from Cymbeline, Iachimo brings Posthumus letters from Britain: This represents unexceptionable division up to the last, but the price of a linked tetrameter, or of a doubtful pentameter with the caesura substituting for a whole foot, is too high to pay merely to force 'All is well yet' into a divided verse. This 'All is well yet' has no relation to the dialogue but refers to something in the letter that Posthumus has been reading as he inquires about the court with half his mind; hence it is appropriate for it to stand alone, unlinked, the divided pentameters resuming when he dismisses the business in the letter and starts his dialogue with Iachimo about their wager of the ring on Imogen's constancy. Given the metrical problem and what amounts to a change of address, it may seem that editors have no more right to form a divided tetrameter here than a divided hexameter in Coriolanus V.i.28, above. Except that the change of address is wanting (although there is a rhetorical shift), the same may be true of Granting all the independent emphasis of 'That I beheld', it may be an open question whether it is enough to justify a linked tetrameter (or a whole missing foot at the caesura in a linked pentameter) as traditionally arranged, or whether the thrust of 'That I beheld' is not accentuated if a short line follows a short line and the pentameters start again with 'Mine eyes did sicken'.

When very special reasons for a separate short opening line are wanting, however, an editor will play the odds by following the common pattern of a short ending line followed by a divided pentameter composed in its first half of a brief separate speech and in its second half either of another brief separate speech or else the opening part line of a longer speech. The following illustration, from Cymbeline, where no options are present, shows that rhetorical pauses alone are far from enough to enforce short emphatic separate lines.

One should note that such rhetorical examples as 'You have abus'd me', and 'I, I said so Sir', and 'Your Mother too', but especially 'Ile be reveng'd', could be thought to raise the question of separate short beginning lines if found in a situation where an option existed in the arrangement. But Shakespeare links them here, and we may take the suggestion that rhetorical breaks alone should be viewed with caution as offering ground for an arrangement of divided verse that would isolate short beginning lines as single units.

When editors attempt to decide between such options on impressionistic grounds, inconsistency is inevitable. For example, in the following illustration from Othello

most editors arrange as above with the short beginning of Brabantio's speech properly completing a pentameter; but in contrast the New Penguin divides the two short lines 'Not I . . .' and 'My name is Rodorigo', thus attempting a false rhetorical emphasis by isolating Brabantio's 'The worsser welcome' as a short beginning. In the next two illustrations from Othello all editors agree with the Globe in inconsistent arrangements: It is difficult to distinguish any substantive rhetorical difference between 'Good Sir, be a man' and 'O Sir content you' that in the one case calls for linking but in the second has superior claims to separation as a single line against all the statistics for Shakespeare's pattern and the examples quoted above from Cymbeline. It would seem clear that the normal and preferred lining should be: just as in As You Like It the Globe and other editors agree that, despite the vocative, 'O unhappie youth' should be linked:

A problem in division may arise when two part-lines that seem as if they should be linked do not produce a regular pentameter when joined, either because of a deficiency or else a surplus of syllables or of stresses. Deficiency is the more common pattern: it occurs frequently enough, in fact, to suggest with some force that Shakespeare's intention in certain specific patterns was nevertheless to link the two part-lines and not

to have them stand as an independent pair.[17] In the most common pattern two part-lines would form a regular pentameter provided the caesural pause—coming after a stressed syllable—were given an unstressed syllabic value forming the first part of the post-caesural iambic foot, what Jacob Schipper calls "The suppression of the anacrusis."[18] The effect in the second half of the line is identical with that in the first half of a full pentameter line that begins without the unstressed syllable of its first iamb, an acceptable metrical device. The syntactical emphasis that such a metrical break provides is perhaps more obvious in divided than in undivided verse, but a special rhetorical effect can be produced in a full pentameter, as in On the example of such lines as these, it may readily be taken that Shakespeare intended two part-lines to be linked when the caesural pause between them could substitute for a missing unstressed syllable. Or possibly:

A second kind of metrically deficient line is more troublesome. The normal complement of five stresses is not found in this variety: instead, the verse contains only four stresses although arranged in a significant syllabic pattern that distinguishes the line from a normal tetrameter. In this form of divided line, the caesura follows a weak or unstressed syllable. One can think of this as a line with a lyric caesura in which the caesura substitutes for the immediately following stressed syllable. The pattern exhibits what Schipper calls the suppression of the arsis.

Although it has been suggested that 'monst(e)rous' should be the pronunciation to make the line regular, it is more probable that we have below the special sort of line under discussion:

Despite examples such as these from Shakespeare's full pentameter verse, an editor might still feel uneasy about attempting to link two

separate short lines that joined would create a verse with this pattern. But an interesting repetition of the pattern in divided verse in The Merchant of Venice demonstrates that Shakespeare deliberately introduced it for variety: The symmetry here is worth examining since it can only be designed. The first two occurrences of divided lines with 'In such a night' have regular caesuras and ten-syllable verse (TLN 2410-11, 2414-15); the next two have epic caesuras, i.e. the first half with a feminine ending (TLN 2418-19, 2421-22); and the last two exhibit the line in which the caesura must substitute for a stressed syllable (TLN 2425-26, 2429-30). That this passage legitimizes linkage in such a pattern of the suppression of the arsis can scarcely be doubted, particularly when it is considered that 'In such a night' has no qualifications to be segregated as an independent short line beginning a speech.

The caesural elision of an unstressed or even a stressed syllable is of course not confined to dividing the closing and opening lines of two speeches but also affects the linking of short single lines that might otherwise be viewed as independent. An example of each may be drawn from Cymbeline:

The legitimacy of this division cannot be challenged, for under no circumstances could 'Wherefore then' or 'Mine eyes' be acceptable as an independent short line beginning a speech. Whether two short lines should be independent or metrically linked can be decided as much on the basis of this elided syllable pattern at the caesura as if they formed a regular ten-syllable pentameter. But since Shakespeare could link some short lines in a series and others not,[20] it is a helpful determinant in the above that the preceding and following speeches for some lines have been full pentameters.

Nevertheless, restraint is needed when an editor chooses to reline acceptable verse in order to produce these special lines. For example, in Two Gentlemen the Folio prints TLN 1891-97 as follows:

after which pentameters resume. Lines 1892-93 doubtless form one divided verse; but even so 1895-97 are acceptable, particularly in this play, as an exchange in trimeter and tetrameter verse. It seems a distortion for the Globe to reline: If 'Alas' be accepted as a separate line, there is no reason why the others may not be independent short lines as well, particularly in view of such a passage as this, which occurs shortly afterward: The Globe here retains the short unlinked lines, and justly, and on this analogy should not have relined 1895-97. Curiously, the New Arden, which prints 1891-97 unlinked, reverses the Globe here and links 1932-33 and then 1934-35 although limping verse is produced. Correctly, neither attempts to print 1935-36 as a divided verse: the stage-business connected with the bringing of the picture after Silvia's command combined with the change of address justifies the beginning of her pentameter speech by an independent short verse.

The Globe, usually followed by modern editors, has something of a tendency to distrust the common short ending lines to speeches and to reline in order to link these, particularly if thereby a short independent line can be brought within the system of divided verse. For example, the Folio correctly prints Coriolanus 2930-34 as

whereas the Globe, and modern editors, reline as Although the use of the caesura after 'assistance' to substitute for a stressed syllable is less common than its substitution for an unstressed, the divided line is metrically acceptable as accentuating the thrust of Menenius' 'I think not so.' But since this abrupt denial was already given superior emphasis as a separate short line placed between two pentameters, it is difficult to see what is gained by relining, especially since the short ending 'Self-loving,—' is very far from being anomalous. Very likely this is one of the examples of how Shakespeare's versification has been sophisticated. A more curious example of the Globe relining, always followed, comes in Antony and Cleopatra 723-726: The Globe relines: Since TLN 725 is a completely acceptable pentameter, it need scarcely be broken up to mend the meter; yet all the Globe has done is to interfere with the Folio lining only to trade off one short ending line for another.[22] The virtue of such tinkering remains obscure. Something of a parallel may be found in which—like 'My being in Egypt . . . '—the Globe and editors chop up as Once more there has been a simple trade-off of one short line for another; but here the Globe seems to have been relining in order to end the scene with a short line, just as it relined the Folio to form the part-lines Shakespeare sometimes—but by no means in a majority of instances—may close a scene with a short line; but the pattern is not so invariable or so powerful as to require an editor to reline satisfactory pentameters in order to oblige what is by no means even a majority convention.

Two particular problems next occur which, since they may prove to be insoluble, require mention but not extended discussion. The first comes in a situation where, say, a speech ends with a short line and is followed by two separate short-line speeches but where the meter would permit the first of the two short speeches to be linked either with the closing line of the preceding speech or with the second short line. In either case a short line remains. The problem is to decide if any pattern may be discerned as to whether Shakespeare would ordinarily have had any preference for the second short speech to be the closing half of a divided verse or else for it to be a separate short-line speech between pentameters, the preceding pentameter acting as the divided verse.

Pending a more complete analysis one may suggest that consistently useful patterns do not seem to exist. One possibility would be to inquire whether in a divided line Shakespeare more often preferred three stresses than two stresses before the break. What too few examples for statistical certainty reveal is something of a stand-off. If in Julius Cœsar I.iii, for instance, we count the position of the caesura in undivided full pentameters (corresponding to the break in a divided line), we find about 58 lines with a three-stress first half as against about 87 with a two-stress first half. In Measure for Measure II.iv the proportion is relatively balanced at about 75 each; in Cymbeline I.i there are about 22 three-stress versus 29 two-stress first halves before the caesura (although in this same scene in the divided verses the proportion is 6 and 3 with one case ambiguous). If, instead, we count the number of unlinked short ending and beginning lines as bearing more closely on the problem, we find that in Julius Cœsar as a whole we have about 20 three-stress unlinked closing lines as against only 6 two-stress; in Measure for Measure about 19 three-stress versus about 10 two-stress. For short unlinked beginning lines (which might be taken to have a general correspondence with the second half of a verse) Julius Cœsar shows perhaps only 1 of three stresses versus 4 of two stresses; Measure for Measure about 8 (some doubtful) three-stress beginning lines but only 2 two-stress.

On the other hand, it may seem that statistics for divided pentameters where no option exists should be the most trustworthy. In this situation in Julius Cœsar one may find about 45 divided verses with a three-stress first half but 60 with a two-stress. On the contrary, Measure for Measure shows about 80 linked lines with three stresses in the first half as against about 40 with two stresses in the first half. Smaller samplings as in All's Well III.iii reveal perhaps 4 linked lines with three stresses in the first half versus 5 with two stresses; Coriolanus I.ii has 5 three-stress first halves and 2 two-stress, III.i has approximately 32 opening

three-stress halves and 13 two-stress; Cymbeline I.ii has about 16 three-stress but only 7 two-stress.

What these partial figures seem to show is that no exact correlation is necessarily found between the point of division of a full pentameter line and the two halves of a linked one. Julius Cœsar, it is true, in I.iii has a preponderance of full verses with a two-stress first half and also a preponderance in the play as a whole of two-stress first halves in linked lines. (In I.iii there are 2 linked verses with a three-stress first part as against 1 with a two-stress.) Also, the statistics for short unlinked closing lines agree in a majority of two-stress verses. If II.iv is typical of Measure for Measure, its relatively balanced count for three- and two-stress openings of full pentameters does not agree with its strong predilection for three-stress short separate ending lines and for three-stress first halves of a divided verse. On the whole, then, the more pertinent results are about what one would expect: the larger proportion of three-stress short closing lines to speeches agrees in general with the proportion of three-stress closing lines completed by two-stress opening lines of another speech in divided verse. The figures for the length of unlinked opening short lines correlate less exactly with the figures for two-stress second halves of divided verses, although this occasional difference may be due to the smaller number of short beginning lines upsetting the statistical picture. There is at least a possibility, then, that in the later plays Shakespeare increased the number of three-stress first halves in a divided verse over his habit in earlier plays; but fuller statistics are needed.

It follows that if in three short lines an absolute metrical option existed whether to link the first and second or the second and third to form a pentameter by indention, beginning at least as early as Shakespeare's middle period (but with due regard for the characteristics of the play concerned) an editor would be playing the odds, all things being equal, if he opted as a consistent principle to divide ambiguous choices by selecting a three-stress short line for the first half.

The question naturally follows, of course, whether all things are usually equal in such a situation. For instance, occasions arise in which the choice of a three-stress opening half of an ambiguous division could clash with what might be another possibility, a special factor that could have influenced Shakespeare in certain circumstances. Some indication exists that Steevens, followed by the Globe and most editors, was occasionally influenced in his choice of which short lines to link by a belief that some interchange between the characters, whether as intimate as a question-answer or merely a matter of direct address, was a valid justification for linkage as a metrical unit. A difficult question of content thus might enter the picture.

Here not only the sense but also the anomaly of 'From Fiffe, great King' as a short beginning line dictate the inevitable arrangement. But when the question of address and the problem of an anomalous short opening line clash, the Globe's preference for speech and return produces debatable verse-division: Here there is insufficient reason to break the really significant rule that run-on lines without change of address, stage-business, or strong syntactical shifts should not begin a speech as an unlinked unit. The correct division is better as which clashes with the odds (though not in this play) that the beginning line of a linked speech will more often be three than two feet. Something of an interchange is still present, of course, but shifted from the Globe's division between Cassius and Antony (the more direct perhaps) to one between Antony and Brutus. A clearer case of disjunction comes in All's Well that Ends Well in which the Folio should be divided as

whereas the Globe incorrectly links the King's comment to Lafeu's first address to Helen, a direct relationship, but at the excessive price of creating an anomalous short beginning line:

The same situation is found in Antony and Cleopatra which should be lined:

In both cases, presumably to emphasize direct address, the Globe faultily arranges as

Although the two slight pieces of evidence of three-stress first halves of divided lines and direct interchange are present in the Globe arrangement of the above, these weak reeds must bend before the superior force of the evidence against ordinary beginning short lines of speeches, which the Globe so frequently violates. That this question of interchange as a pattern for linking lines is suspect is shown not only by the above but by any number of examples where the meter enforces a linking contrary to such a pattern, as in Macbeth:

Under such conditions, Flatter (p. 88) is perhaps wrong in allowing a fancied pause after the end of Polonius' speech and then a rapid interchange between Claudius and Gertrude to sway him into arguing for a division of

instead of the conventional

The meter here is perfectly regular iambic, which in this play is to be preferred to the anapest required if the King's question begins a verse. In the next passage options are present but the simplicity of linking all part-lines, as is customary here, overweighs any theoretical attempt to divide question and answer:

Utterly ambiguous situations where no metrical or patterned hint suggests one or other division of short lines are not very common, a fact in itself perhaps of some significance. In such cases an editor may flip a coin or try to be consistent according to some principle or else inconsistent but according to some suggestion in the scene. An interesting example occurs in Macbeth:

Here editors have differed, as is to be expected. Curiously, the Globe arranges as in the Folio without any linkage. Kittredge and the Pelican divide as

whereas Alexander, the New Arden, New Penguin, Riverside prefer the question-answer (and the three-stress first half):

In an example from Measure for Measure

modern editors follow the Globe in linking Mariana's and Isabella's short lines although the New Arden instead, divides a verse, between 'That's bitter to sweet end' and Mariana.

In the nature of the case we lack all evidence as to Shakespeare's intentions, and indeed we cannot be certain that he would always follow the same pattern, if in fact he had one in mind.[24] If consistency is a virtue, an editor has a choice of adhering regularly to one of three options for dealing with these troublesome but fortunately rare problems: (1) he can prefer always to link the short closing line of the preceding speech with the first short line, or, contrariwise, always to link the two independent short lines; (2) he can link variably according to his estimate of the directness or not of an interchange; (3) he can play the odds that in most plays an initial three-stress part-line is more likely to link with a completing two-stress than a two with a three, but he can vary according to the characteristics of the specific play and the evidence of its non-optional linked lines. In the present state of our information either the first or the third seems the best bet: little confidence can be placed in the interchange or question-and-answer assumption;[25] and if such an interchange is not present, logically the first separate short line would then need to be linked willy-nilly with the short closing, a good option also in case the 3-2 division does not work.

As remarked, however, all things may not be equal; and though it seems advisable for an editor to be guided chiefly by one of these three (or two) options, yet some calculated inconsistency may be introduced on what could appear to be evidence of another nature (even though

fragile) about Shakespeare's intentions. A small example comes in Cymbeline: Modern editors follow the Globe in arranging a solution with which there can be little quarrel. This scene has begun with five passages of dialogue between the two gentlemen, each speech of the first ending with a part-line completed by a short question from the second. The first four have linked a three-foot closing line to a two-foot following question, but the fifth has reversed the stresses in the two halves. If this pattern were to be followed, then three-stress 71 should link with closing two-stress 70, leaving 72 as the only short unlinked line in the scene except for the exit closing dimeter. One may note, also, that the two speeches immediately following this crux both link the second gentleman's three-stress part-line question with the closing two-stress short line of the first's preceding speech.

Sufficient evidence of another sort is exhibited in Coriolanus 354-359, the Folio arranged thus by all editors following the Globe:

Here the key is Shakespeare's occasional use of a separate short line for an exit. It is true that this would still be preserved if lines 355-356 were linked, but the contrast with pentameters would then be lost. The next example depends upon a moot point of metrics and is thus more arguable: The Globe and all editors line the Folio as above. If 1871-72 were to be linked, a line of five iambs would throw a wrong stress on 'Manifest'; the option would be to scan the second half after the caesura as a dactyl and trochee. It could be suggested that such powerful emphasis conventionally appears in the beginning of a line instead of in its latter half and that a very strong verse is made by linking Brutus' and Sicinius' short lines. Two of these examples happen to link a closing line with a short separate, although the third example makes a complete verse of the short closing and links the next two short lines. In the present state of our knowledge it is impossible to say whether Shakespeare had any preference in such a situation. Two have the first half of the linked line in three stresses, the more popular pattern for many plays. The third example, with a two-stress first half, appears in a scene in which elsewhere four such linkages have been of the three and two pattern but three of the reverse. If the example is correct, the scene would be evenly balanced in this pattern.

The only point in presenting these examples is to suggest that some reason for selecting the two parts of a divided verse from among three equal components is better than none. Among the lack of reasons one may well place editorial tradition transmitted through the Globe, even though in the above three cases the tradition seems to be preferable to the alternatives. When no grounds may exist for a reasoned preference, then an editor may appeal to consistency either in linking the closing short with the first of the two short separate lines, or dividing the verse between the two separate shorts. Or he can split the ticket by surveying the non-optional divided lines in the play in question and then making his decision to correspond with the majority pattern they present, either three and two stresses, or two and three, according to the play.

The second problem concerns the kind and amount of metrical anomaly that an editor can accept in joining two short lines to form a divided verse according to his estimate of Shakespeare's intention. In brief, the problem boils down to the question whether an editor is justified in attempting to link two short lines that, joined, would create either a normal tetrameter or a hexameter. In the preceding discourse the general position has been taken that a divided verse could exhibit the same irregularities as an irregular pentameter within a pentameter speech, especially in the matter of the caesural suppression of either an unaccented or an accented syllable. The question then arises, whether the occasional tetrameters, and the rarer hexameters, that one sometimes finds within pentameter passages justify the linking of two or more short lines to form a corresponding irregularity in a divided verse.

In both of these examples, the Globe says 'yes,' and by indenting the short beginning line to complete the short ending line, and numbering the verse accordingly, it and various editors assert the principle that lines other than pentameters may be treated as divided verse.[26] Other irregular lines in the Folio that the Globe et al link as one divided verse may be quoted: Tetrameters are not expandable, but various hexameters may yield (not to slurring, which is infrequently required in full pentameter passages, but) to elision: where the normal elision of 'We are' to 'We're' takes care of the pentameter by an anapest. So with Coriolanus 3073-75. Here the New Arden and New Penguin are alone in reproducing the Folio lining, which in the New Arden arrangement depends upon Menenius contracting 'You have' to 'Y'have' to match Cominius' 'Y'are': The New Penguin arrangement without indention creates an anomalous and undesirable run-on short opening line for Menenius. The New Arden, thus, is superior to it and superior to the bibliographically unnecessary relining by the Globe and most editors with its final hexameter or else a double feminine ending.[27] This divided line is debatably a hexameter with a lyric caesura, not a forced pentameter formed with an epic caesura and 'Not so sicke' as an anapest. (But see below.)

It seems clear that enough evidence can be gathered to compel an editor to face up to the problem of division posed by anomalous extra-metrical and deficient lines. Without calling Shakespeare up from the grave to testify, one may confess that his intention can often be in considerable doubt. On the one hand, it is legitimate to take it that if he can write internal hexameters (or tetrameters) in pentameter context like

then he can also write them as divided verses. Yet the difficulty seems to arise that given the relative frequency of undivided to divided lines in any play, there are many more of these problem divided lines than should be statistically present; moreover, it is uncomfortable that the divided hexameters seem to be in the considerable majority whereas the opposite seems to be true in undivided verse. If all of the examples exhibited the characteristics common to the separate short beginning lines of speeches following a pentameter ending of the preceding speech—that is, commands, stage-business, change of address, marked rhetorical break—it would be easy to argue that these lines must not be linked, for by juxtaposing this attention-gathering device to the flexibility of a short line concluding a speech, Shakespeare had achieved the ultimate in natural ease of verse dialogue. Some observed examples, it is true, encourage this view, as in Enobarbus' 'That I beheld' or Isabella's 'Oh just, but severe Law', both linked by the Globe but able to stand alone. To these may be added two examples, already mentioned, which the Globe inconsistently treats as separate unlinked lines:

On the other hand, several examples (all linked by the Globe) may be observed that under ordinary circumstances should not permit a separate short beginning line. In this category comes Menenius' 'For 'tis a Sore upon us' and Prospero's 'Doe so: and after two daies'. To these can be added:

This last example, especially, poses a problem, since Kittredge, New Arden, and New Penguin, for example, reline to avoid the hexameter:

This is an attractive emendation, particularly since it not only rids the verse of an extreme irregularity in the opening short line, if unlinked, but also provides a short closing line to the speech before Cæsar's entrance. (The Globe, Alexander, and Riverside, on the contrary, retain the Folio lineation and link the two part-lines as a hexameter.) Relining is also possible in the Tempest 427-431 crux, as in

Ariel's exclamation followed by what is close to a change of address in his direct question to Prospero makes an acceptable short line to begin his speech, much better than 'Doe so: and after two daies', completing a divided hexameter.[28] Coriolanus 1958-61 might be subject to more drastic relining, though no more so, perhaps, than the frequently accepted relining of Julius Cœsar 273-276.

If the rhetorical tetrameter is objectionable, one could read:

In this case the divided line with Comenius would suppress the arsis.[29]

Nevertheless, too many of these divided hexameters cannot be removed by relining. In such cases it is a question whether the cure is worse than the disease; that is, does an editor take it that it is better to link two hypermetrical part-lines or to treat them as separate verses at the expense of violating one of Shakespeare's most important verse patterns, the special purpose for the short beginning line of a speech. Something may be said for the view that, at least in these exceptional circumstances, editors have traditionally but not invariably preferred to link two part lines to form a divided hexameter.

The same problem involving the nature of the opening line of a speech occurs when it is preceded not by the short closing line of a speech but by an independent short line which cannot be prose. In the first example the change of address in Cæsar's speech makes it quite ambiguous whether Antony's short line should be treated as an independent verse, and Cæsar's speech should begin with an independent short beginning line, or whether they should be linked at the expense of forming a hexameter.

Editors have followed the Globe in linking the part-lines, but the special nature of 'Bid them prepare within' suggests that they have been ill advised not to treat it as a separate verse as in the Folio. In the next example it could barely be an open question whether the beginning line qualifies as a special situation, for the rhetorical break is slight indeed: The Globe here chooses inconsistently to treat each as a separate verse.

On the whole, an editor is playing the odds in such situations when he declines to create hexameters and opts for a series of short lines.

The Globe and editors go through considerable contortions in arranging these lines, including the creation of an unnecessary hexameter: The hexameter in line 133 is no more justified than the two concluding lines of the scene, which the Globe correctly treats as independent short lines: If one accepts short separate lines, the passage is more simply arranged, close to the Folio: Under any circumstances the last line is going to be irregular.[30] Another case where two short lines may be preferable to the Globe's hexameter comes in Measure for Measure, 1111-15: As would be expected, the Globe and other editors link in a hexameter 'My body up to shame' with 'Then must your brother die.' It is interesting to observe that elsewhere in the dialogue of Isabella and Angelo between her entrance and his exit (1034-1184) Isabella speaks six independent short lines and Angelo four for a total of ten. Excluding the passage above, Isabella has four speeches that end in a short unlinked line as compared with eight for Angelo, for a total of twelve. Finally, there are elsewhere eight clear cases of linked short lines. The proportion of separate independent lines in this dialogue by no means discourages the suggested arrangement, then, whereas the high proportion of unlinked ending lines suggests that if an editor were to link any of the three short lines it should be 'Then must your brother die' with 'And 'twer the cheaper way', a course not to be recommended, however.

About the first of the next examples there can be only a trifling question, and it is properly linked by the Globe despite the hexameter that is created. The second is only slightly more doubtful, since relining is possible, but the odds favor its authenticity as a hexameter.[32]

The final situation involving the division of hypermetrical lines concerns the concluding short line of a speech followed by a single short line. Here there is a much weaker incentive to link the two short lines since the criteria applicable to the short line beginning a speech have no applicability to an independent or single short line. Ordinarily editors are inclined to link such lines if the result is metrically regular but to be somewhat wary (although not wary enough) when anomalies would

be created. Sometimes the pattern in the adjacent lines both before and after may provide a hint, but Shakespeare is by no means so regulated as to close himself in with rigid schema even in the more formal versification of the early plays. Editors follow the Globe in linking the above lines. Again, editors follow the Globe in linking the two short lines. Most editors follow the Globe in inconsistently treating these short lines as separate verses, but Kittredge arbitrarily divides 2175 so that 'Quick, dispatch' completes a pentameter with the Provost's exit line. Editors follow the Globe in treating these short lines as separate verses.

When one surveys the interrelationships between these three categories, certain tentative observations may be made. So far as observed, the separate short lines preceding a speech beginning with a short line offer no evidence in favor of divided tetrameters although some occasional evidence is found in favor of divided hexameters. The evidence of a short line ending a speech followed by another beginning the next speech, insofar as one can trust the four examples observed, is not in favor of editorial arrangement as linked verses. Coriolanus 1191-92 and 2079-80 show a change of address, perhaps the strongest evidence normally found in favor of the independence of a short line beginning a speech. Antony and Cleopatra 1994-95 and Measure for Measure 788-789 both have a marked rhetorical pause, reinforced in the example from Measure for Measure by what is very close to a change of address from Isabella's exclamation made largely to herself then shifting to a farewell to Angelo. Until equally pertinent evidence to the contrary is

turned up (and perhaps it does exist), an editor would be advised to heed the pattern of short opening lines and to decline any attempt to link short lines to form a tetrameter. Cautiousness in this respect may be reinforced by the example of editorial struggles to reline Coriolanus 2904-6, which reads in the Folio: The Globe and Riverside arrange, with an opening hexameter (or a normal pentameter if 'he is' is pronounced 'he's'), and emend as: Kittredge, Alexander, New Arden, and New Penguin prefer an arrangement that produces a divided tetrameter by completing a speech with a short separate line: This would be acceptable if 'Hail to you both' were not indented as part of a line but were given separate-verse status: it would appear to be anomalous to reject a normal short separate line and hitch it to an ending short line at the expense of creating an irregular tetrameter. Another possibility suggests itself, particularly when one sees how crowded is the Folio page by the compositor's difficulty with cast-off copy forcing him to set verse as prose and to tuck Menenius' line in the white space to the right of 'Haile Sir' after the full line of the opening ''Tis he' etc. Three Folio lines would have been required by an arrangement that might perhaps have existed in the manuscript:

If it is possible in the present state of our knowledge to be fairly forthright in barring divided tetrameters (for if one accepts tetrameters, why not divided trimeters and even dimeters?), some relaxation seems

required by the evidence in the matter of divided hexameters, a slightly easier concession when one can take account of possible slurring on some occasions and no doubt simple carelessness in others, with textual corruption also possible but not to be relied on. In the nature of the case, some few divided hexameters may be avoided by relining, particularly if a bibliographical explanation for the printed anomaly can be offered. Nevertheless, certain of these hexameters appear in the text on a damned-if-you-do and damned-if-you-don't basis: that is, it is a nasty choice to be forced to accept a divided hexameter or else the violation of the strong pattern of the comparative self-sufficiency of short beginning lines. This may be the situation with Coriolanus 1957-61 (although somewhat draconian relining is possible) but it can be observed more clearly in Julius Cœsar 1599-1602 and Cymbeline 3392-94.

What one comes up with is the view that although some must be accepted, many of the divided hexameters that editors customarily adopt and some that, inconsistently, they do not adopt, are better off as separate lines whenever the observed pattern justifying a short beginning line can reasonably be applied, this whether the preceding short line is the closing line of a speech or a separate single line. If it is logical for the Globe and later editors to treat 'My brother did love Juliet' (MM 1154) and ''Tis certaine then for Cyprus' (Oth. 374) as separate verse lines, it is even more logical to remove the editorial linkage of cases like 'Bid them prepare within' (JC 1117). Such procedures will solve a number of problems and leave in a minority, perhaps, the situations where the treatment of a short run-on or other non-special-purpose beginning line as a verse unit is inapplicable. In some cases plausible although not demonstrably correct relineation may remove the problem, as in The Tempest 427-430 and possibly in Julius Cœsar 273-276. Yet Cymbeline 3392-94 illustrates a second category in which an editor would be almost certainly wrong were he to reline, although the below is just faintly possible:

It is perhaps wisest to assume that a few problems like this will always arise where one must accept a divided hexameter instead of an anomalous run-on short opening line or else submit the passage to radical relineation. However, it would also be wiser to accept these as sports, and to print as few of them as possible. As an alternative to the 'naked' opening line such as AYL 1044, TN 980, Cor. 88, or Cymb. 2178 cited earlier as examples of probably authentic but anomalous short first lines not to be linked and generally not to be removed by relineation, a divided hexameter is probably preferable so long as such divisions are not considered to represent normality. As remarked, carelessness, the exigencies of revision, or textual corruption may account for them chiefly; nevertheless, it seems to be true that on some occasions Shakespeare deliberately wrote a loose verse scannable either as a hexameter if strictly accented but actually as an intended pentameter with more flexible accentuation.[33]

As for short single lines that would create a tetrameter or hexameter if linked to the short closing line of a preceding speech, or to another short single line, no possibility exists for determining their authenticity. It is therefore conservative to ignore anomalous linkings and to treat extra-metrical single lines as separate verses in arrangement and in line numbering.

The Globe correctly does not attempt to divide the above short lines, but it cannot resist the temptation in to link the short line despite the resulting hexameter.

Some erratic and inconsistent arrangements of part-lines appear in modern editors as a heritage of the Globe's dislike of independent short lines in pentameter context. That these short lines are legitimate and for a verse-speaker are to be considered as a separate line of verse is incontestable:

The above examples are scarcely to be tinkered with,[34] but in less obvious cases the Globe is often inconsistent in its treatment. Sometimes it goes to extremes in an attempt to link lines that by any rational standards should remain independent, as in Folio which is arranged by the Globe and modern editors as That 1812-13 should divide a pentameter does not require editors to force the same pattern on 1810-11 to produce a tetrameter or else a line impossible to scan though with five stresses, as by making 'evening' into three syllables and placing the final stress on the 'ing' of 'coming'. Given the numerous examples of short-line dialogue in this play, which the Globe does not try to link, like it may seem that editors are ill-advised not only in TLN 1810-11 but in the following, to force division at the expense of a hypermetrical line: arranged as an example that is inconsistent with (But see Ant. 1488.) However, the Folio continues: which editors reline as: an arrangement that is suspect.[35]

An equally common type of editorial interference comes in Coriolanus:

Here the Globe and editors, perhaps correctly, link the first two lines[36] but then take it that this division justifies relining two separate verses to secure a continuation of the linkage: The peculiar virtue of relining Brutus' speech, which is an excellent pentameter, in order to complete 1733 is obscure when the result is merely to substitute a short ending line for the normal independent line given to Cominius. But the Globe dislikes these independent short lines. An egregious example comes in All's Well: which the Globe relines as: Only some quite unwarranted doubt that the Countess' line 564 was not an acceptable pentameter could rationalize such interference, which modern editors have perpetuated. So in Antony and Cleopatra we have the Folio relined superfluously by the Globe as Or another example from All's Well: which the Globe treats as Usually when there is an opportunity, the Globe will split the concluding pentameter of a scene in order to produce a short ending line for the exit: The Globe arbitrarily breaks up Antony's and Lepidus' pentameters[37] to read: Although Shakespeare occasionally ends scenes with the short line of a concluding longer speech (or an independent short line), his pattern is by no means so marked as to justify editorial interference with the Folio lining.[38]

The case is altered, of course, when adjustment of division must be made in extremely anomalous Folio arrangement.

Here the relining to is required not at all to provide a link for the short ending line 444 but instead to make a divided verse of the Folio's highly irregular independent short opening run-on line 'But as 'tis' at 446, a consideration that could not have weighed with the Globe even though it relines correctly.

Other considerations, such as desirable double linkage, may dictate the division of a full Folio pentameter. A good example comes in Macbeth:

where the Globe and editors rearrange to This relining appears to be justified only in part because it links both halves of the divided line but chiefly because it occurs in a scene in which without exception all other short lines in the dialogue between Macbeth and the murderers are linked. Under these circumstances the two short lines 1116 and 1118 would seem to be anomalous and in need of emendation.[39] However, if only one half of the divided line were linked by rearrangement, the division of a full pentameter in this manner would not be advisable. The Globe is too often guilty of this latter unnecessary tinkering, as illustrated above by Cor. 1731-35, AWW 563-564 and 460-461, or Ant. 723-726.

The Globe's inconsistency in dividing verses is a serious fault.

Here, presumably in order to provide a link with the closing line 3080, the Globe, followed by Kittredge, the New Arden, and Riverside, arranges the short lines so that, in effect, this unlinked closing line is transferred awkwardly by splitting 3086, an arrangement for which no other need exists. Alexander, followed by the Pelican editor, recognized the ineptness of this solution but had no better expedient than to print all as unlinked short lines, an improbable arrangement. Actually, the division of the verses is quite simple once one accepts the conventional short ending line 3079:

On the other hand, in Measure for Measure the Globe allows a short ending line to stand that is eminently linkable, as

and instead combines 2368-69 into a single pentameter. This is an eccentric rearrangement and cannot be urged even by the present-day bibliographical knowledge that 2368-69 being too long for the measure could have been divided by the compositor. The Folio arrangement by which a pentameter single speech is divided to link its first half with the short closing line of the preceding speech although in its second half leaving an unlinked short line is so common that emendation to restore the undivided pentameter is unnecessary and even intrusive.[40]

A final problem is at present so complex and relatively unanalyzed as to prevent more than cursory mention. As would be expected, editors

have differed widely in their attempts to create divided verses out of short lines in a dialogue in which a prose speaker discourses with one or more verse speakers. Some scenes offer little difficulty: editors of Julius Cœsar have treated Casca, properly, as a consistent prose speaker in I.ii despite a tolerable opening pentameter, I.ii.215 (318-319), though set as prose, and a tolerable tetrameter in his second speech, line 218 (322); at line 233 (336) he could complete a pentameter with Cassius; he approaches his exit with what could be scanned as verse in lines 295-296 (396-397); and, in between, various of his short answers would lift no eyebrows were they present in verse dialogue. Nevertheless, it is probable that Casca is intended to be exclusively a prose speaker in this scene, although a verse speaker elsewhere.

Not all decisions are so relatively simple, as evidence the varied editorial treatment given Lucio's speeches in Measure for Measure, I.ii. Shakespeare's general tendency seems to be either to be consistent in treating a prose speaker, like Casca in JC I.ii, or else to set off in blocks the shift of a character from verse to prose (or the other way round) as with Oliver in As You Like It when, joined by Rosalind as Ganymede, he switches at the end of IV.iii.166ff. (2321). Or in Measure for Measure V.i when Lucio opens as a verse speaker (l. 75, TLN 2435), but shifts to prose (l. 179, TLN 2552) and, pace editors, remains a prose speaker for the rest of the scene. In this stretch of Lucio's prose, editors, nevertheless, have him divide a line with the Duke (l. 214, TLN 2587). He speaks what could be an acceptable pentameter in line 306 (2683). For the rest, prose prevails, and it is certain that in this block system the attempt to introduce a part-line of verse at V.i.214 is ill-advised and a distortion.

When one reverts to I.ii the main question arises whether Lucio is a prose speaker throughout or else a verse speaker at random. The evidence is slight for Lucio as a verse speaker, scarcely more than in V.i, in fact. He opens the scene in prose and so continues until his re-entrance at l.128 (TLN 215) when he addresses Claudio with a perfect pentameter: 'Why how now Claudio? whence comes this restraint.' Yet his next speech in response to Claudio's verse is in prose and then follow (228-231) four short question-and-answer lines between the two that do not link (except when forced to by the New Arden editor alone), until one comes to the Folio mislining at ii.146-148 (233-236):

This the Globe and most editors arrange as: No editor has been able to resist the temptation to link Lucio's 'With childe perhaps? (248) by completing a pentameter with Claudio's 'Unhappely, even so.' Finally, after two prose speeches at 265-267 and 280-284, most editors follow the Globe in assigning Lucio's 'Within two houres' (286) as verse by completing it with Claudio's exit line 'Come Officer, away' (287). The difficulty of deciding what is prose and what verse has led editors to a variety of arrangements. All print 215 as verse; Alexander and the Pelican arrange 235-236 as prose but the rest as verse; all divide 248-249 as verse; and all divide 286-287. Curiously, the Globe (and Riverside), which likes to rearrange so as to end a scene with a separate part line, links instead of the equally metrical alternative (with an epic caesura): which has whatever virtue may inhere to a three-foot first half and a direct exchange of words.

The evidence for Lucio as a mixed prose and verse speaker in this scene is scarcely better than that for Casca in Julius Cœsar I.ii, where there is also what could pass for an opening pentameter by Casca and some short lines that could link. Until the whole question of verse lining is resurveyed with different texts from our sophisticated modern editions, and the question of Shakespeare's use of characters as random prose and verse speakers instead of with ordered, or blocked variation, is settled, editors may be suspicious of counting as verse such possibly fortuitous hits as 'With child, perhaps?' and the like. At the most, it would seem, Lucio in this scene might be allowed an opening pentameter on his re-entrance but thereafter be confined to prose exclusively, including such fortuitous hits as 'With child, perhaps?' and the like. Whatever the results of further investigation, it may seem that conventional editorial arrangement in Shakespearean mixed scenes of verse and prose needs to be more finely tuned since it is probable that too many prose speakers have been assigned random part-lines of verse.

Notes