Jump to ContentJump to Main Navigation
Shakespeare, Co-Author$

Brian Vickers

Print publication date: 2004

Print ISBN-13: 9780199269167

Published to Oxford Scholarship Online: October 2011

DOI: 10.1093/acprof:oso/9780199269167.001.0001

Timon of Athens with Thomas Middleton

Chapter:
(p.244) 4 Timon of Athens with Thomas Middleton
Source:
Shakespeare, Co-Author
Author(s):

Brian Vickers

Publisher:
Oxford University Press
DOI:10.1093/acprof:oso/9780199269167.003.0004

Abstract and Keywords

This chapter reports the history of Timon of Athens and its authorship. Since the play's represented action was not as violent as Titus Andronicus, it offended fewer critics. Middleton also wrote The Revenger's Tragedy. The division of authorship between Shakespeare and his co-author was solidly established by metrical and linguistic studies. With Timon of Athens, the impressive feature of authorship studies from the 1840s to the 1990s is that widely differing methodologies have converged, supporting sound attributions and discrediting others. All these methods agree in assigning to Middleton a substantial part of Timon, and Shakespearians who continue to deny this point risk forfeiting their scholarly credibility.

Keywords:   Thomas Middleton, Timon of Athens, Shakespeare, co-author, authorship

As we have seen, authorship studies of Elizabethan drama over the last century have passed through three stages. The first involved the identification of passages in anonymous or co-authored plays that closely echoed known work by one or more dramatists. This method considers longer verbal collocations, not single words or short phrases, and works best when it can show parallels of thought and attitude, in addition to verbal parallels. The second stage was the realization that different writers have different preferences within frequently recurring linguistic features: the use of contractions (ʼem, ʼee, ’tis), choice of alternative spellings (while or whiles), variant verb forms (has or hath, does or doth), or favourite exclamations (pish, phew). These usages could be located in a text, and systematically investigated, sometimes revealing clear differences. Variations in verse form could also be identified and tabulated. The third stage grew out of this, locating linguistic features— such as ‘function words’ (to, of, the), or the words beginning and/or ending sentences or speeches—and submitting them, to detailed statistical analysis. All three approaches are independently valid, but the most satisfying results are obtained when their results support each other.

The history of Timon of Athens and its authorship has followed this pattern. Since the play's represented action was not as violent as Titus Andronicus it offended fewer critics. Throughout the period 1660 to 1800, Timon was praised as a ‘Moral and Instructive’ tragedy (James Drake, 1699), full of ‘useful satire’ (Charles Gildon, 1710), a ‘warning against ostentatious liberality’ (Dr Johnson, 1765), or the effects of ‘inconsiderate profusion’ (William Richardson, 1783).1 The play could easily be absorbed into neoclassical theories of literature's moral function, despite the obvious defects of the text as printed in the First Folio. It was not until the 1840s that Charles Knight made a reasoned case for co-authorship, in the prefaces to his popularizing Shakespeare editions, collected as Studies of Shakspere.2 Knight began his essay with a scathing account of the co-called ‘regulation’ of the play's verse made by George Steevens, anxious to reduce (p.245) lines ‘to the exact dimensions of his ten-syllable measuring-tape’ (1849, p. 48). The consequence of Steevens's textual alterations, Knight pointed out, was that ‘some very important characteristics’ of the play's language ‘have been utterly destroyed in the modern copies—the record has been obliterated’ (p. 68). One significant detail destroyed by editorial modernizing is the fact that the Folio ‘presents to us in particular scenes a very considerable number of short lines, occurring in the most rapid succession. We have no parallel example in Shakspere of the frequency of their use. The hemistich is introduced with great effect in some of the finest passages in Lear. But in Timon of Athens its perpetual recurrence in some scenes is certainly not always a beauty’ (p. 68). Knight was the first to notice this feature of the play's style, and he rightly deplored the effect of Steevens's regularizings of the metre, which

have given to the Timon of Athens something of the semblance of uniformity in the structure of the verse; although in reality the successive scenes, even in the modern text, present the most startling contrarieties to the ear which is accustomed to the versification of Shakspere. (p. 70)

Knight's highly sensitive ear, attuned to Shakespeare's authentic versification, allowed him, to reject firmly both theories explaining the play's stylistic unevenness—the English, by which ‘the ancient text is corrupt’, and the German, which described Timon as a late, and probably unfinished play. Knight presented his own conclusions, ‘which have been hitherto entirely overlooked’, namely ‘that the differences of style, as well as the more important differences in the cast of thought, which prevail in the successive scenes of this drama, are so remarkable as to justify the conclusion that it is not wholly the work of Shakspere’ (ibid.). The play contains ‘parts not only out of harmony with the drama as a whole, in action, in sentiment, in versification, but altogether different from anything he had himself produced in his early, his mature, or his later years’ (p. 76).

Knight argued that Shakespeare had taken over some existing play, ‘which probably retained possession of the stage for some time in its first form’, and that Timon ‘has come down to us not only rewritten, but so far remodelled that entire scenes of Shakspere have been substituted for entire scenes of the old play’ (pp. 70–1). Whether or not we accept this theory, the importance of Knight's essay lay in the criteria he used in identifying two different hands. These criteria involved both questions of style, to be considered here, and of dramaturgy, postponed to Chapter 7 below. Knight began his detailed analysis with an observation repeated by many subsequent readers:

The contrast of style which is to be traced throughout this drama is sufficiently striking in the two opening scenes which now constitute the first act. Nothing can be more free and flowing than the dialogue between the Poet and the Painter [1.1.1–94]. It has all the equable graces of Shakspere's facility, with occasional examples of that (p.246) condensation of poetical images which so distinguishes him from all other writers. For instance:

  • All those which were his fellows but of late,
  • (Some better than his value,) on the moment
  • Follow his strides, his lobbies fill with tendance,
  • Rain sacrificial whisperings in his ear,
  • Make sacred even his stirrup, and through him
  • Drink the free air.
  • (1.1.78–83)

(p. 71)

Knight was sure that Timon, in this opening scene, was ‘Shakspere's own conception’, and although he found Apemantus' repartee not distinctly Shakespearian, ‘no one can doubt to whom these lines belong’:
  • So, so; there!—
  • Aches contract and starve your supple joints!—
  • That there should be small love ‘mongst these sweet knaves,
  • And all this court'sy! The strain of man's bred out
  • Into baboon and monkey.
  • (1.1.247–51)

(ibid.)

When Knight reached the second scene of Act 1, however, he felt that Shakespeare's hand had suddenly vanished, for

we find ourselves at once amidst a different structure of verse from the foregoing. We encounter this difference remarkably in the first speech of Timon:

  • I gave it freely ever; and there's none
  • Can truly say he gives, if he receives:
  • If our betters play at that game, we must not dare
  • To imitate them; faults that are rich are fair.
  • (1.2.10–13)
In the first scene we do not find a single rhyming couplet; in the second scene their recurrence is more frequent than in any of Shakspere's plays, even the earliest. This scene alone give us sixteen examples of this form of verse; which, in combination with prose or blank verse, had been almost entirely rejected by the mature Shakspere, except to render emphatic the close of a scene, (p. 72)

Knight was perfectly correct to point out the unusually high frequency of rhyme in this scene, running quite against Shakespeare's stylistic development. Even more perceptively, he observed both the awkward nature of the couplets and a striking difference in the imaginative element of poetry: ‘In the instance before us, we find the couplet introduced in the most arbitrary and inartificial manner—in itself neither impressive nor harmonious. But the (p.247) contrast between the second scene and the first is equally remarkable in the poverty of the thought, and the absence of poetical imagery’ (ibid.).

One of the recurring experiences of studying co-authored plays is to discover that the same character seems to be quite different from, one scene to the next. Charles Knight showed that in the parts of both Apemantus and Lucius (the Steward), quite different styles could be detected. In 1.2, a scene in which he could find no trace of Shakespeare, Apemantus makes this superficial moralizing commentary, complete with rhyming couplets and half-lines:

  • Hey day,
  • What a sweep of vanity comes this way!
  • They dance! they are mad women:
  • like madness is the glory of this life,
  • As this pomp shows to a little oil and root.
  • We make ourselves fools to disport ourselves;
  • And spend our flatteries, to drink those men,
  • Upon whose age we void it up again,
  • With poisonous spite and envy.
  • Who lives that's not depraved, or depraves?
  • Who dies, that bears not one spurn to their graves
  • Of their friends' gift?
  • I should fear, those that dance before me now,
  • Would one day stamp upon me: It has been done:
  • Men shut their doors against a setting sun.
  • (1.2.131–45)
In 4.3, by contrast, the great scene where the cynic confronts Timon the misanthrope, Apemantus speaks with remarkable energy, fusing metaphors together to contrast Timon's previous and his present state:
  • Thou hast cast away thyself, being like thyself;
  • A madman so long, now a fool: What, think'st
  • That the bleak air, thy boisterous chamberlain,
  • Will put thy shirt on warm? Will these moist trees,
  • That have outliv'd the eagle, page thy heels,
  • And skip when thou point'st out? Will the cold brook,
  • Candied with ice, caudle thy morning tast,
  • To cure thy oʼer-night's surfeit? call the creatures,—
  • Whose naked natures live in all the spite
  • Of wreakful heaven; whose bare unhoused trunks,
  • To the conflicting elements expos'd,
  • Answer mere nature,—bid them flatter thee;
  • O! thou shalt find—
  • (4.3.220–32)
Every reader, I imagine, will agree with Knight on ‘the impossiblity of [this] character having been wholly minted from the same die’ in the two (p.248) contrasting scenes (p. 72). Knight performed the same test with the Steward, quoting his long aside in this scene (1.2), in which we note again the irregular length of the verse lines and the unexpected use of couplets.

Flav.

  • What will this come to?
  • He commands us to provide, and give great gifts,
  • And all out of an empty coffer.—
  • Nor will he know his purse; or yield me this,
  • To show him what a beggar his heart is,
  • Being of no power to make his wishes good;
  • His promises fly so beyond his state,
  • That what he speaks is all in debt, he owes for every word;
  • He is so kind, that he now pays interest for't;
  • His lands put to their books. Well, ‘would I were
  • Gently put out of office, before I were forc'd out!
  • Happier is he that hath no friend to feed,
  • Than such that do even enemies exceed.
  • I bleed inwardly for my lord.
  • (1.2.191–205)
  • Knight commented: ‘We print the speech of the first act as we find it in the original. With the exception of the two rhyming couplets, it is difficult to say whether it is prose or verse. It has been “regulated” into verse, but no change can make it metrical;—the feebleness of the thought is the same under every disguise’ (p. 72). Knight contrasted this speech with one in Act 2, where Timon accuses the Steward of not having warned him of his impending bankruptcy, and of having cheated the estate, to which Flavius offers this self-defence:

    Flav.

  • If you suspect my husbandry, or falsehood,
  • Call me before the exactest auditors,
  • And set me on the proof. So the gods bless me,
  • When all our offices have been oppress'd
  • With riotous feeders; when our vaults have wept
  • With drunken spilth of wine; when every room
  • Hath blaz'd with lights, and bray'd with minstrelsy;
  • I have retir'd me to a wasteful cock,
  • And set mine eyes at flow.
  • Tim.

  • Prithee, no more.
  • Flav.

  • Heaven, have I said, the bounty of this lord!
  • How many prodigal bits have slaves, and peasants,
  • This night englutted! Who is not Timon's?
  • What heart, head, sword, force, means, but is lord Timon's?
  • Great Timon; noble, worthy, royal Timon!
  • Ah! When the means are gone that buy this praise,
  • The breath is gone whereof this praise is made:
  • Feast-won, fast-lost; one cloud of winter showers.
  • (p.249) These flies are couchʼcl.
  • (2.2.155–72)
  • Knight judged that ‘the harmony, the vigour, the poetical elevation of the second passage, like the greater part of the fourth and fifth acts, effectually prevent all substitution and transposition’ by a regularizing editor (p. 73). As for the substance of this scene, Knight detected in it ‘unquestionably … the master-hand of our poet. The character of Timon as his ruin is approaching him is beautifully developed. His reproach of his Steward, slightly unjust as it is, is in a tone perfectly in accordance with the kindness of his nature; and his rising anger is forgotten in a moment in his complete conviction of the integrity of that honest servant’ (p. 73).

    Charles Knight showed himself well able to distinguish Shakespeare's hand from that of his co-author—indeed, almost all. of his judgements have been confirmed by many careful studies in the century and a half since he wrote. In Act 3, he judged,

    very little … is Shakspere's. The ingratitude of Lucullus in the first scene, and of Lucius in the second, is amusingly displayed; but there is little power in the development of character—little discrimination. The passionate invective of Flaminius is forcible; but the force is not exactly that of Shakspere. The dialogue between the Strangers, at the end of the second scene, is unmetrical enough in the original; Steevens has made it hobble still worse. The third scene has the same incurable defects. It seems to us perfectly impossible that Shakspere could have produced thoughts so commonplace, and verse so unmusical, as we find in the speech of Sempronius. The fourth scene, again, has little peculiarity. It might be Shakspere's, or it might be the work of an inferior writer, (pp. 73–4)

    In Knight's eyes this whole sequence, in which Timon's servants try without success to borrow money from the friends whom he had so lavishly enriched, formed a unit in which he could find no trace of Shakespeare. As for the scene (3.5) where Alcibiades petitions the Athenian Senate to pardon a friend of his who has been found guilty of murder, Knight objected to it both on grounds of its un-Shakespearian dramaturgy (see Chapter 7 below), and ‘the internal evidence of thought and style’. As he was the first to observe,

    The scene between Alcibiades and the Senate consists of about a hundred and twenty lines. Of these lines twenty-six form, rhyming couplets. This of itself is enough to make us look suspiciously upon the scene, when presented as the work of Shakspere. Could the poet have proposed any object to himself, by this extraordinary departure from his usual principle of versification, presenting even in this play an especial contrast to the mighty rush and sustained grandeur of the blank verse in the speeches Timon in the fourth and fifth acts? (p. 74)

    What Knight pungently described as ‘the perpetual and offensive recurrence of the couplet’ was only one of several un-Shakespearian elements he detected in this scene:

    (p.250) The whole of the senate scene in Timon is singularly unmetrical; but, wherever the verse becomes regular, it is certainly not the metre of Shakspere. Mark the pause, for example, that occurs at the end of every line of the first speech of Alcibiades. ‘The linked sweetness long drawn out’ is utterly wanting … From the first line of this scene to the last, the speeches, though cast into the form of verse, are in reality nothing but measured prose, (ibid.)

    Knight's strictures applied not only to the verba but to the res of this scene, its thought, imagery, and feeling:

    But in addition to the structure of the verse, the character of the thought is essentially different from that of the true Shakspearean drama. Where is our poet's imagery?… The action of this scene admitted either of passion or reflection; and we know how Shakspere puts forth either power whenever the occasion demands it. The passion of Alcibiades is of the most vapid character:

    • Now the gods keep you old enough; that you may live
    • Only in bone that none may look on you!
    • (3.5.03–4)
    et us contrast for a moment the Shakspearean Coriolanus, under somewhat similar circumstances:
    • You common cry of curs, whose breath I hate,
    • As reek o'thʼrotten fens: whose loves I prize,
    • As the dead carcasses of unburied men,
    • That do corrupt my air: I banish you.
    • (3.3.120–3)

    Some of the stylistic differences that Knight indicated—the unusual number of couplets in the scene, the sequence of irregular lines in Alcibiades' opening speech (3.5.7–23)—can be quantified, and show a distinct numerical discrepancy compared to Shakespeare's normal verse styles at this period. Other perceptible differences cannot be quantified, but a cogent analysis can elicit agreement. As Knight argued,

    In this scene between Alcibiades and the senate, the usually profound reflection of Shakspere, which plunges us into the depths of our own hearts, and the most unfathomable mysteries of the world around us and beyond us, is exchanged for such slight axioms as the following:

    • For pity is the virtue of the law,
    • And none but tyrants use it cruelly.
    • (3.5.8–9)
    • To revenge is no valour, but to bear.
    • (39)
    • To be in anger is impiety,
    • But who is man that is not angry?
    • (56–7)
    (p.251) The form of expression in these scenes with Alcibiades appears to us as remarkably unShaksperean as the character of the thought. By nothing is our poet more distinguished than by his conciseness, the quality that makes him so often apparently obscure, (pp. 73–4)

    But such speeches axe never obscure, only too obvious. This moralizing style, with its love of sententious couplets, became one of the stylistic markers which allowed later scholars to identify Shakespeare's co-author as Middleton.

    Although Knight assigned most of Acts 4 and 5 to Shakespeare, he indicated some passages contributed by the collaborator: ‘The second scene of the fourth act, between the Steward and his servants, has some touches undoubtedly of the master's hand; the Steward's speech, after the servants have left, again presents us the rhyming couplets, and the unmetrical blank verse’ (p. 76). Knight referred to this un-Shakespearian speech, which clearly resembles the Steward's style in 1.2 and 2.2:

    • O, the fierce wretchedness that glory brings us!
    • Who would not wish to be from wealth exempt,
    • Since riches point to misery and contempt?
    • Who would be so mock'd with glory, or to live
    • But in a dream of friendship …?
    • ………
    • Strange, unusual blood,
    • When roan's worst sin is, he does too much good!
    • (4.2.30–9)
    Knight's judgement lapsed only once, it seems to me, when discussing ‘the scene between the Poet and the Painter at the commencement of the fifth act’, which ‘is so unmetrical, that it has been printed as prose by all modern editors’. Knight was tempted to describe ‘this hobbling approach to metre’ as un-Shakespearian; he believed that ‘the Poet and the Painter of this scene are as unlike’ those characters in the first act ‘in the tone of their dialogues, as can be imagined’; and he thought that ‘Timon, in the lines which he speaks aside, has caught this infection of unmetrical blank verse which reads like prose, and jingling couplets which want the spirit of poetry’ (p. 76). Later scholars did not share these judgements. Although the Folio text is certainly uneven, they found the verse in this scene recognizably Shakespearian, and quite unlike the co-author; the two artists are just as verbose and insincere here as in the opening scene; and Timon's three asides only contain one couplet, which is quite unlike the co-author's flaccid moralizing:
    • To thee be worship, and thy saints for aye
    • Be crown'd with plagues, that thee alone obey!
    • (5–1–52–3)
    (p.252) Knight was not alone, however, in having difficulty with this closing Act. But he fairly summed up his case for the presence of two different authors in the play by appealing to more general impressions:

    It is not by looking apart at the scenes and passages which we have endeavoured to separate from the undoubted scenes and passages of Shakspere in this play, that we can rightly judge of their inferiority. They must be contrasted with the great scenes of the fourth act, and with Timon's portion of the fifth,—the essentially tragic portions of this extraordinary drama. In power those scenes are almost unequalled. They are not pleasing—they are sometimes positively repulsive in the images which they present to us: but in the tremendous strength of passionate invective we know not what can be compared to them. In Lear the deep pity for the father is an ever-present feeling, mingling with the terror which he produces by his denunciations of his daughters. But in Timon the poet has not once sought to move our pity: by throwing him into an attitude of undiscriminating hostility to the human race, he scarcely claims any human sympathy. Properly to understand the scenes of the fourth and fifth acts, we must endeavour to form a general estimate of the character which Shakspere has here created, (pp. 76–7)

    We will return to questions of character and dramaturgy in a later chapter, in which Knight's qualities as a sensitive reader will again be apparent.

    Charles Knight's essay on the two styles he could detect in Timon of Athens is a remarkable performance, which has never received the recognition it deserves. English Shakespearians ignored this problem for a generation, and the next scholar to take it up was Nikolaus Delius,3 who had produced the first critical edition of Shakespeare's works in Germany (7 vols., 1854–61). Delius prefaced his essay with an excerpt from an earlier work of his, a commentary4 on Ludwig Tieck's Shakespeare criticism (1846), in which he had drawn attention to the ‘striking contrast in style, versification, and characterization that occurs throughout the whole drama, not only between separate scenes but even within the same scene’ (Delius 1867, P- 335)· Delius rejected both of the standard explanations of this contrast, textual corruption or the play's unfinished state, dismissing the latter by observing that ‘the weaker passages’, or ‘those which stand in glaring contradiction to the supposedly complete parts’, are not limited to one part of the play but alternate throughout, making it incomprehensible that Shakespeare could have put down on paper the finished and the unfinished parts not just next to each other but intertwined (p. 336). (This comment should be remembered when the ‘unfinished play’ theory crops up again in the 1930s.) In this 1846 essay Delius cited (p.253) with respect Charles Knight's attempt (originally published in his ‘Pictorial Shakespeare’) to solve this embarrassment by postulating that Shakespeare had extensively rewritten an earlier play.

    Although he praised ‘the astuteness and fine aesthetic feeling’ with which Knight had pointed out the variations in style running through the whole play, at this point Delius had rejected his theory on two grounds. First, he argued, there is no example of Shakespeare ever having treated an earlier play in this manner, ‘leaving one half of the original untouched in all its metrical and stylistic deficiencies, while transforming the other half to a Shakespearian property’. When he transformed the source-plays for King John and The Taming of the Shrew Shakespeare did so completely, not leaving them half-complete, half-incomplete (pp. 336–7). Secondly, Delius judged in 1846, those parts of the play which Knight had allocated to the co-author, ‘with all. their faults, still bore the stamp of Shakespeare's individuality’, so he concluded that the play was a reworking by the dramatist of some youthful work of his own (p. 337).

    But Delius soon came to reject his own theory, recognizing that in Shakespeare's early period a play's whole design may be simpler, but it is still ‘logically clear and coherent’, and ‘nowhere suffers from, the internal contradictions, the unmotivated transitions and insertions that confront us in Timon’ (p. 338). Characterization in the early plays may be neither deep nor precise, but it is always consistent, and gives us the impression of ‘dealing with figures of real flesh and blood, not merely with the silhouettes or pure genre figures which appear here and there in some scenes of Timon’. The style or diction of Shakespeare's early plays, also, in contrast to ‘the partially compressed and intellectual difficulty’ of his later work, shows a clarity and ease of understanding quite unlike the ‘deliberate obscurity and unpleasing strangeness’ of the non-Shakespearian parts of Timon. These inauthentic scenes also differ totally in versification from the early plays, in which Shakespeare took over the regular, monotonous blank verse of his predecessors, gradually transforming it to ‘the most lively and varied medium, perfectly adapted to dramatic purposes’. Compared to these strictly observed metrical conventions, Delius judged that, in the un-Shakespearian scenes of Timon,

    complete metrical anarchy reigns, defeating all editorial attempts to regularize the verse. And in a completely un-Shakespearian manner, rhymed couplets are plentifully inserted into the limping and stumbling iambs, which look like prose, as well as into prose passages, rhyming verse whose otiose tinkle is doubly objectionable in such a context.… (p. 338)

    In Shakespeare's early and late work, by contrast, rhyme is never introduced without some inner justification and motivation, usually at the end of either a scene or a long speech (p. 339).

    (p.254) For all these reasons, then, Delius had rejected his own theory of Timon being Shakespeare's revision of an earlier play, and in his edition of the text (1855) had signalled his complete acceptance of Charles Knight's identification of two different hands at work throughout (ibid.). Now, in 1867, Delius produced a new analysis, independent of Knight's, based on four general assumptions. First, that Shakespeare's revision dates from a late period in his career (pp. 339–40); secondly, that the play to which he contributed new additions must have existed complete, and may have already been a favourite with the public, so that Shakespeare could insert whole scenes without having to alter the plan of the whole (p. 340)—a point taken from Knight; thirdly, that the play which he worked over was not that old, since its verse style suggested that the co-author was one of Shakespeare's younger contemporary dramatists (p. 341); and fourthly, that since the extant play had already digested the classical sources, Shakespeare had no need to exert himself in that respect (pp. 341–2). I summarize these arguments merely to show the difficulties which meet even so.rn.eone as knowledgeable as Delius, once scholars begin to speculate about the putative old plays which Shakespeare may have known, attempting to re-create them ex nihilo in order to legitimize their own indeterminable theories. Here Delius made ‘den. Vorgänger’ responsible for the overall design of the play and for its handling of the sources, so exonerating Shakespeare from any blame for whatever structural deficiencies it might have. But at least Delius is to be praised for recognizing that the co-author was younger than Shakespeare, writing a newer form of verse.

    Although Delius's theories of revision need not concern us, his commentary on. the co-author's language and style remain, highly perceptive. Having assigned Act 1 scene 1 mostly to Shakespeare, Delius observed of 1.2 that nowhere in his work can we find ‘such verse and such prose, such sudden, transitions from one to the other and back again’. A strikingly un-Shakespearian detail is that Apemantus ‘decks out his lame iambics with rhymes completely lacking the epigrammatic balance and antithetical point that Shakespeare gives to his couplets’ (p. 345). As an. example of this stylistic feature I cite Apemantus' speeches at the first banquet, which move unpredictably from prose to blank verse, and to rhymed verse having eight, ten, or even, twelve syllables:

    Apem.

  • I scorn thy meat, ’twould choke me, for I should neʼer flatter thee. O you gods! what a number of men eats Timon, and he sees 'em not! It grieves me to see so many dip their meat in one man's blood, and all the madness is, he cheers them up too.
  • I wonder men dare trust themselves with men.
  • Methinks they should invite them without knives:
  • Good for their meat, and safer for their lives.
  • There's much example for't: the fellow that sits next him, now parts bread with
  • him, pledges the breath of him in a divided draught, is the readiest man to kill
  • (p.255) him; 't ʼas been prov'd. If I were a huge man, I should fear to drink at meals,
  • Lest they should spy my windpipe's dangerous notes:
  • (treat men should drink with harness on their throats.
  • Tim.

  • My lord, in heart: and let the health go round.
  • 2. Lord.

  • Let it flow this way, my good lord.
  • Apem,

  • 1 low this way? A brave fellow! he keeps his tides well. Those healths will
  • make thee and thy state look ill, Timon.
  • Here's that which is too weak to be a sinner,
  • Honest water, which neʼer left man i' th' mire.
  • This and my food are equals, there's no odds;
  • Feasts are too proud to give thanks to the gods.
  • Apemantus' grace.
  • Immortal gods, I crave no pelf,
  • I pray for no man but myself.
  • Grant I may never prove so fond,
  • To trust man on his oath or bond
  • (1.2.38–65)
  • In 3.3 Delius again pointed to the ‘wildness and waywardness’ of the verse, and the ‘trivial and tasteless rhyming couplets’ which the other dramatist constantly introduced into his blank verse, ending this scene, like its predecessor, with banal moralizing couplets:

    • And this is all a liberal course allows;
    • Who cannot keep his wealth must keep his house.
    • (3.3.40–1)
    and
    • Men must learn now with pity to dispense:
    • For policy sits above conscience.
    • (3.2.87–8)
    ‘How can anyone have ever taken these inept attempts at rhyme for Shakespeare's?’, Delius asked (p. 349). In 3.5, Alcibiades' scene before the Senate, he judged, the blank verse only appears to be such, mixed with incomplete lines and sententious couplets, the couplets often being elliptical and obscure (pp. 350–1). The soliloquy of Flavius, ending 4.2, also bears all the marks of Shakespeare's co-author, ‘the moralizing commonplaces whose triviality is made more obvious by the rhymes strewn among the unrhymed, jerky and incomplete verses’, the Oblique and obscure' language, the inconsequential reasoning, and the curious alternation of an ‘otiose jumble of sententiae with factual utterances’ relating to the plot and its development (p. 353). In sharp contrast to this clumsy and incoherent use of poetic forms is Shakespeare's verse, which in 2.2 ‘shows all the living dramatic variety, the strength, zest and fullness characteristic of his (p.256) middle period’ (p. 347). Delius believed that Shakespeare's interest in. the play reached its liveliest and most energetic point in the scenes of Timon's misanthropy {4.1 onwards), for which he reserved his most powerful verse (p. 352).

    Delius gave both a general account of the differences between the two dramatists and a detailed division of their authorship (pp. 344–61), which can. be summarized as follows:

    Shakespeare:

  • 1.1; 2.2.123–233; 3.4.79–117; 3.6.86–105; 4.1; 4.2.1–29;
  • 3.1–299, 381–457; 5.1.47–228; 5.2; 5.4.1–64
  • Co-author:

  • 1.2; 2.1; 2.2.1–122; 3.1; 3.2; 3.3; 3.4.1–78; 3.5; 3.6.1–85,
  • 106–120; 4.2.30–50; 4.3.300–80, 458–536; 5.1.1–4.6; 5.3; 5.4.65–85
  • Modern scholarship, which has developed a much wider range of approaches, would endorse about three-quarters of Delius's assignations, giving Shakespeare all. of 2.1, and only the first forty-five lines of 2.2. In the second banquet scene, where Timon feasts his guests on stones and warm water, scholars today attribute to Shakespeare not only the verse diatribe (85–105) but the bitterly ironic prose prayer which precedes it (65–84). Modern authorities would agree with Delius's division of 4.2, but would assign, to Shakespeare the whole of 4.3—the scene where various groups visit Timon as he digs up roots and gold–up to the concluding sequence between Flavius and Timon. Here Delius judged the Steward's soliloquy on Timon's misery (458–71) to be identical in content and attitude to his soliloquy closing 4.2, both un-Shakespearian (again echoing Knight). In the dialogue following Delius pointed to the unmistakable signs of the co-author's hand in the blank verse, with its ‘faulty construction and scansion’, and attempt at rhyme:
    • and believe it,
    • My most honour'd lord,
    • For any benefit that points to me,
    • Either in hope or present, I'd exchange
    • For this one wish, that you had power and wealth,
    • To requite me by making rich yourself.
    • (517–22)
    Delius was certainly right to give this closing sequence to the co-author. But where he assigned several passages in. Act 5 to ‘the predecessor’, mostly on the grounds of awkward dramaturgy, modern scholars tend to think it all Shakespeare's, explaining the inconsistencies and loose ends in terms of the rough and untidy ‘foul papers’ (pre-theatrical manuscript) from which it derived. These exceptions apart, Delius made a judicious division of responsibilities between Shakespeare and the other dramatist, which agreed on many details with that made by Charles Knight.

    (p.257) F G. Fleay, that pioneer of scientific verse tests, discussed Timon in two essays produced (or so he claimed) in 1869 and r 874.5 Fleay disagreed with Delius's theory of Shakespeare having revised a predecessor's work, arguing that ‘the un-Shakespearian parts were certainly the latest written’ (Fleay 1874c, p. 139). Using a number of different criteria, Fleay presented his results in tabular form, asserting that his table had been ‘grounded on an examination of every line of the play, one by one, as regards the metre; on a specific analysis of the plot with regard to the bearing of each scene or portion of a scene on every other; and on a minute examination of the Folio of 1623, with regard to the printing and spelling of proper names, stage directions, etc., which have been altered by modern editors …’ (p. 140). Fleay is to be commended for the range of approaches he applied and indeed, for his conclusions, which agree surprisingly well with more recent scholarship—I see no grounds for denying Fleay the title ‘modern’. Fleay divided the play as follows:

    Shakespeare:

  • 1.1; 2.1; 2.2.1–193, 205–42; 3.6; 4.1; 4.2.1–29; 4, 3.1–291, 363–97, 414–52; 5.1.50–231; 5.2; 5.4
  • Unknown:

  • 1.2; 3.1; 3.2; 3.3; 3.4; 3.5; 4.2.30–50; 4.3.292–362, 398–413, 453–543; 5.1.1–50;5.3
  • More recent scholars would reject Fleay's division of Act 4 scene 3 into so many small parts, on the grounds that co-authors normally took responsibility for whole scenes; but they would agree that the final part of that scene (lines 458–536 in the Riverside edition) is not by Shakespeare.

    This unnecessarily fragmented treatment of 4.3 shows Fleay's weak spot, his treatment of Shakespeare's prose. Whereas his ear for authentic verse was often unerring, he found Apemantus' prose, in scenes which are undisputedly Shakespearian, to be ‘bald, cut up’ (p. 130) or ‘chopt-up’ (p. 135), and he impatiently ascribed these passages to the other hand. When it came to Shakespeare's verse, Fleay believed that he could detect ‘the true ring’ of Shakespeare in 2.1, and in 2.2 he judged the verse to be ‘pure Shakespeare. No one else could have written it. The “drunken spilth of wine”, the “one cloud of Winter showres, | These flyes are coucht”, the “halfe-caps and cold moving nods, I They froze me into silence”, bear the lawful stamp of his mintage’ (p. 133). But Fleay's judgements, based on long and wide reading, were not merely impressionistic. He supplied a metrical table with precise computations, which I reproduce in Table 4.1. Rearranging Fleay's data gives the information supplied in Table 4.2. Those figures clearly show that Shakespeare used far less rhyme than his co-author, and much less prose. More important than the statistics, however, was Fleay's penetrating analysis (p.258)

    Table 4.1. Metrical table for Timon of Athens

    Shakespeare

    Unknown

    A Prose

    B Blank

    C Irregular

    D Rhymes

    A Prose

    B Blank

    C Irregular

    D Rhymes

    I.I

    58

    208

    25

    2

    1.2

    64

    126

    21

    36

    2.1

    31

    2

    2

    2.2

    11

    2.2

    85

    133

    8

    6

    3.1

    49

    11

    3

    2

    3.6

    III

    12

    2

    6

    3.2

    58

    30

    2

    4

    4.1

    33

    2

    6

    3.3

    8

    19

    9

    6

    4.3

    339

    28

    2

    3.4

    18

    78

    12

    8

    5.1

    162

    14

    6

    3.5

    73

    14

    30

    5.2

    14

    I

    2

    4.2

    36

    4

    10

    5.4

    77

    4

    4

    4.3

    85

    53

    9

    18

    5.1

    46

    4

    5.3

    5

    1

    4

    Totals

    254

    1,009

    86

    36

    Totals

    339

    441

    75

    122

    Source: Fleay 1874c, p. 141.

    Table 4.2. Rhyme and blank verse in Timon of Athens

    Division

    Length

    Percentage of rhyme to blank verse (cols. B + C)

    Percentage of prose to all verse (cols. B + C + D)

    Shakespeare's scenes

    1,385 lines

    3.28

    18.3

    Unknown's scenes

    977 lines

    23.64

    34.7

    of the stylistic confusion in the non-Shakespearian part of the play, which echoed those made by Knight and Delius. The ‘style of these [scenes], and especially the metre, is utterly unlike anything in the other plays of Shakspere. It is marked by great irregularity, many passages refusing to be orthodox, even under torture; it abounds in rhymes, in emphatic and unemphatic passages alike; the rhymes are often preceded by incomplete lines; one of the rhyming lines is frequently imperfect or Alexandrine’ (p. 144). More particularly,

    The rhythm of the two portions of the play differs in every respect. The Shakspere parts are in his third style (like Lear), with great freedom in the rhythm, some 4 and 6 syllable lines, some Alexandrines with proper caesuras, and rhymes where the emphasis is great, at the end of scenes, and occasionally of speeches in other places. The other (p.259) parts have irregularities, both in defect and excess, of every possible kind. There are lines of 8 and 9 syllables, Alexandrines without caesura, imperfect lines in rhyming couplets, broken lines preceding rhymes, and other peculiarities, not one of all which is admitted in Shakspere's rhythmical system. [1.2.257–9] is one of many instances of intolerably bad rhythm:

    • I'll lock thy heaven from thee.
    • O that men's ears should be
    • To counsel deaf, but not to flattery!

    Fleay added that ‘one point in the metre may appear clearer if expressed statistically’, namely that ‘there are proportionally 8 times as many rhymes’ in the portion by ‘Unknown’ as in Shakespeare's part (p. 148). He also noted that in the scenes where the creditors dun Timon (3.1–3.4) ‘there is not a spark of Shakspere's poetry, not a vestige of his style’, nor of his ability to differentiate characters through their language. The ‘creditors' servants, who can rhyme much more easily than the best-educated personages in the Shakspere part of the play’, he complained (echoing both Knight and Delius), are not individualized: ‘they speak the same dialect, and use the same rhetoric, as all the characters of the second author; any speech of any one might be spoken by any other, so far as the language and form of expression are concerned’ (p. 133).

    Fleay used other, non-stylistic arguments to differentiate the two hands involved. He observed that one character ‘is called Ventidius in the Shakspere part of the play’, such as 1.1, but ‘Ventigius’ in the scenes by ‘Unknown’, such as 1.2 and 3.3 (pp. 131, 146). Fleay pointed out (p. 147) that the name ‘Apemantus’ is also spelled ‘Apermantus’ in some scenes, and he showed that in the non-Shakespearian 1.2, ‘and here only’, the steward is called Flavins, while in 2.2 ‘Flavius is given by Shakspere as the name of one of Timon's servants who is not the steward’ (pp. 131, 146–7). Fleay also observed that the value of the Greek ‘talent’ varies considerably throughout the play, a loan of 5 talents being a considerable sum in 1.1, and 3 talents forming an acceptable dowry in 2.1. In 2.2, 3.1, and 3.4, however, sums of 50 talents are mentioned, while at 2.2.205 Timon even asks for ‘A thousand talents’. In 3.2 we find both the completely vague ‘so many talents’ (12, 24, 37) and ‘fifty-five hundred talents’ (39). Fleay tried without success to correlate these striking discrepancies with his authorship division (pp. 144–5), and was forced to start emending the text (for ‘thousand talents … I would read 1000 pieces’), a dangerous step in authorship studies.

    Fleay had devoted a lot of thought to Timon, and made several intelligent observations. He deduced from the Folio pagination that Troilus and Cressida had originally been designed to fill this space, and that when that play was moved to its place between the Histories and Tragedies (for copyright reasons, as we now think), ‘this space, then, of pp. 80–108 … being left unfilled, it became necessary to fill it’ (p. 137). Credit for this discovery is (p.260)

    Table 4.3. Verse table for The Revenger's Tragedy

    Total number of lines

    over 2,400

    Number of rhyming lines

    exactly 460

    Double endings

    exactly 443

    Alexandrines

    exactly 2

    Deficient and short lines

    about 125

    Source: Fleay 1874c, p. 138.

    usually given to J. Q. Adams in 1908,6 but Fleay must be awarded some share, even though he used it to support his completely improbable theory of a second playwright being engaged by I leminge and Condell in 1622–3 to Pad out the play in order to fill the vacant space—which he signally failed to do, since the Folio pagination still has a gap between pp. 98 and 109,

    I have left till last Fleay's one suggestion as to the identity of the second hand:

    Having then, laid down as certain the division of the play, and the assignment of the nucleus to Shakspere … we come to the more difficult question still—who was the second author? The ratio of rhyme to blank verse, the irregularities of length (lines with four accents and initial monosyllabic feet), number of double endings & c, agree with only one play of all that I have analyzed (over 200), viz., The Revenger's Tragedy. But I am doubtful as to pressing this argument very strongly, unless we give up (as I am quite ready to do) the notion of the play being finished in 1623, as The Revenger's Tragedy was written in 1607. The evidence of general style, however, appears to me strongly to confirm, the conjecture that Cyril Tourneur was the second author, (pp. 137–8)

    Now that we know that Middleton, not Tourneur, wrote The Revenger's Tragedy, Fleay's suggestion seems uncommonly prescient. He supported it (see Table 4.3) with ‘numerical data … as near as I can count them in such a badly printed edition, as we yet have’. Although Fleay did not provide exactly the same range of statistics for Timon, if we accept his figures here, the percentage of rhymed verse (19.1) is close to that of the ‘unknown’ dramatist in Timon (23.6), and far above Shakespeare's rate in that play (3.3). My own calculation for The Revenger's Tragedy, based on Foakes's edition, with a total of 2,192 lines of verse, 243 lines of prose, and 380 rhymes, would be 17.3 per cent rhyme, which is not dissimilar to the figure for the non-Shakespearian scenes in Timon. Fleay then quoted (from, a corrupt text) several passages of verse from The Revenger's Tragedy which seemed to (p.261) him to have ‘exactly … the metre’ of the co-author in Timon, including part of Vindice's concluding speech. I print the whole speech from a modern scholarly edition:7
    • May not we set as well the duke's son?
    • Thou hast no conscience; are we not reveng'd?
    • Is there one enemy left alive amongst those?
    • ’Tis time to die, when we are ourselves our foes.
    • When murdʼrers shut deeds close, this curse does sealʼem:
    • If none disclose 'em, they themselves reveal 'em.
    • This murder might have slept in tongueless brass,
    • But for ourselves, and the world died an ass.
    • Now I remember, too, here was Piato brought forth a knavish sentence once: no
    • doubt, said he, but time will make the murderer bring forth himself. ’Tis well he
    • died, he was a witch.
    • And now, my lord, since we are in for ever,
    • This work was ours, which else might have been slipp'd,
    • And, if we list, we could have nobles clipp'd,
    • And go for less than beggars; but we hate
    • To bleed so cowardly. We have enough, iʼfaith;
    • We're well, our mother turn'd, our sister true;
    • We die after a nest of dukes. Adieu.
    • (5.3.1.1.7–25)
    As we saw in Chapter 1, this jumbling of blank verse, rhyme, and prose is typical of Middleton, and if the reader will compare this speech with the same stylistic melange in Apemantus' speeches in Timon, 1.2 (quoted above), the affinity will be instantly visible. It was not until 1928 that E. H. C. Oliphant suggested Middleton's authorship of The. Revenger's Tragedy, an identification subsequently confirmed by the work of R. PL Barker, David Lake, and MacDonald Jackson, so it is hardly surprising that Fleay's identification has been either ignored or ridiculed.

    In 1910 an American scholar, E. H. Wright, published his dissertation on the play's authorship,8 following a by now firmly established tradition in. identifying two authors on the grounds of wide variations in style, characterization, and dramaturgy. The co-author, to whom Wright attributed about one-third of the play, strove hard to be a poet by seeking ‘rhetorical effect’, but Wright found his work ‘too thin in substance, too halting in expression, too tame and trite in imagery, too clumsy in characterization, too lacking in dramatic fitness, in a word too uninspired, to pass unsuspected’ (Wright 1910, p. 27). Wright selected three passages from the play as illustrating the co-author's style, first that sequence in the opening banquet scene (1.2.38–65), previously singled out by Delius and quoted above, drawing (p.262) attention to the ‘quick and aimless shifts from prose to verse’ and back— found in Apemantus' speech and elsewhere—‘usually not even due, as in that excerpt, to a desire to rhyme’ (p. 31). In this banquet scene, Wright pointed out, the language ‘staggers aimlessly from prose to verse, from verse back to prose, sixteen times in 257 lines’ (p. 36). Following Knight, Delius, and Fleay, Wright noted that the co-author is also very fond of rhyme, indeed ‘twenty per cent of his verses rhyme. The ratio is practically constant with him; and the rhymes are scattered indiscriminately throughout his scenes. Shakespeare has only four per cent of rhymes and almost all of these at ends of scenes.’ If one omits such final couplets, the difference is even more striking, the proportion of rhymes being less than I per cent for Shakespeare, 18 per cent for his co-author (p. 30). Another characteristic of the co-author is his fondness for irregular verse lines, which Wright computed to amount to 18 per cent of his verse, while for Shakespeare he counted only 4 per cent (p. 31). Taking the three characteristics together, Wright judged the co-author's technique to be ‘irreconcilable with Shakespeare's … Frequent and useless shifts from prose to verse were never part of Shakespeare's practice; irregular verse was always scarce with him; and by the period of Timon he had all but discarded rhyme’ (p. 30).

    Although coming to similar conclusions as Knight, Delius, and Fleay had done about the two different verse styles in the play, Wright (correctly) criticized Fleay's treatment of prose. ‘Bent on giving all the prose to the inferior author’ (p. 35), Fleay had to evolve complicated theories to account for prose passages in scenes that otherwise seem Shakespearian. Thus in Act I scene I, undoubtedly by Shakespeate, Fleay wanted to ascribe the ‘bald’ and ‘cut-up’ style of Apemantus' prose speeches to the other hand. But, Wright objected, ‘the change [at 1, 1.180] from verse to prose—definitive, and by no means like the uncertain vacillations of the inferior author into and out of meter—may … prove nothing but that prose is the natural, almost the only suitable, medium for Apemantus' brand of sarcasm. The balder this is, the better it suits Apemantus; its “cut-up” nature is its merit’ (p. 34). Wright also showed the pernicious effect of Fleay's blind spot concerning prose in two later scenes, 4.3. and 5.1. In the first of these Timon the misanthrope is visited in his cave by various groups of people, including some bandits, a passage which Fleay had assigned to the co-author, his ‘sole reason for thinking … the bandits' dialogue … spurious [being] that it is prose. When we find that Mr. Fleay has given the inferior writer every word of prose, without exception, in the play, we begin to doubt his judgment where it has no further basis …’, especially when we see ‘that Shakespeare meant the bandits to hold some dialogue before addressing Timon' (p. 50). Here, as in the scene following, where the Poet and the Painter approach Timon's cave, ‘the prose is Shakespearean enough’ (p. 52) to leave no case for assigning it to the other hand.

    Having well characterized the style of the co-author (who would ‘habitually (p.263) hash his prose and verse’: p. 49), Wright described Shakespeare's style in terms that we can still accept. In 1.1, for instance, he noted ‘the rare quality of the copious and involved imagery’, but at the same time ‘the orderly technic of the verse’ (p. 33). In 2.1, ‘the solidity and dignity, yet delicacy, of the thought and verse alike, unmarred by metrical deformity, the unerring stroke of every sentence, driven home with telling phrases and with striking images, leaves no doubt’ as to Shakespeare's authorship (p. 36). ‘He alone was capable’ of Timon's blistering address to the ‘knot of mouth-friends’ gulled and assaulted at the mock banquet (3.6.98), and in Timon's curse on Athens (4.1.iff.) ‘the majesty of the style’ declares his hand (p. 46). Wright judged the parting of the servants (4.2.1–30) to be ‘the tenderest scene in Timon; in it is concentrated more pure poetry, perhaps, than is found in any scene of equal length in the play. All the critics note the break between it and the twenty-line soliloquy the Steward stays to speak after the servants go; which is little more than prose run mad’, completely in the co-author's manner (p. 46). The most recent scholarship would support all these ascriptions.

    Wright's cogent analyses of discrepancies in the plotting, and his description of the two different styles in the play, show the strengths of traditional aesthetic criticism. But its weaknesses are also apparent. Convinced as he was that Shakespeare would never have left a play in such a mess, and that the other dramatist made ‘interpolations’ with little concern for unity (pp. 39 ff., 47 ff., 58 ff.), Wright evolved a complex theory which involved postulating insertions into the text which the modern scholar is able to identify without difficulty, and which, once removed, restore the coherence of Shakespeare's pure invention before the interpolator tampered with it. Thus in 2.2, where the loyal Steward tells Timon how parlous his finances are, Timon dispatches three of his servants to his friends to request a loan, in a prose passage (2.2.186–94 m the Riverside text). Wright argued that ‘this bit of prose, intrinsically trivial, [forms] a kind of keystone in the entire structure of the first three acts’ (p. 61), for it was originally written by Shakespeare in verse, but had to be set as prose once the interpolator had inserted the words ‘—I hunted with his honour today; you to Sernpronius’: as Wright put it, ‘one line out, the passage settles into blank verse. Even the pieces of the broken lines fit. So, we may be sure, Shakspere wrote it. Such a reconstruction is not possible by accident’ (p. 66).

    The self-serving nature of such speculations, all too common in authorship studies, needs no comment. But they drove Wright on to deny Fleay's argument that the three dunning scenes that open Act 3 were written by the co-author, claiming instead that ‘a purely esthetic judgment as to the first two of the three scenes … indicates that Shakespeare wrote them; but an esthetic and technical judgment as to the third scene, which is mainly in bad verse, points to the other author’ (pp. 39–40). But aesthetic reasons are not enough: authorship studies must pursue differing approaches independently. If the (p.264) methods are valid they will converge, giving the same result, mutually supporting each other. What was needed at this point was a close examination of the language of these scenes, and this was soon provided. The studies by William Wells in 1920 and H. D. Sykes in 1921 showed the linguistic unity of 3.1, 3.2, 3.3, and 3.4, giving them all to the co-author. Wright's stylistic analyses, and his apportionment of responsibility to the two dramatists, displayed critical acu.rn.en but were vitiated by his theory that Shakespeare wrote the original draft, which the co-author revised. All the inconsistent and uncoordinated plot elements that he pointed to can be explained by a much simpler hypothesis, namely that Shakespeare and his co-author drafted the play together, each writing their allotted scenes, but failed to complete the final stage, which might—or should—have involved checking the joint work for internal coherence and rewriting where necessary. Of course, we will never be able to reconstruct the compositional process of this, or any other Elizabethan collaborative play, but experience of the severe distortion and special pleading produced in the work of Delius and Wright by either of the two arguments involving a temporal dislocation between the two writers' contributions—Shakespeare wrote first; Shakespeare wrote last—suggests that we should keep open, the third possibility, which does at least correspond to what we know about co-authorship in this period, that both writers worked on the play simultaneously.

    The first four serious students of the authorship problem in. Timon of Athens, Knight, Delius, Fleay, and Wright, had all given abundant evidence of the existence of two discrepant verse styles in the play, and of a whole series of differences in. characterization and dramaturgy. The only attempt to identify the co-author had been Fleay's indication that several stylistic features were shared by The Revenger's Tragedy, now known to be by Middleton. The next two scholars to study the play's authorship each made an explicit case for Middleton as co-author of Timon. William Wells cited a whole series of verbal parallels between. Timon and Middleton,9 particularly convincing for the sequence from 2.2 to 3.3, ‘the abortive attempts of Timon to borrow money’, which forms a frequently recurring situation in Middleton's plays. The idiosyncrasy of Middleton's language in these situations consists in his frequent use of ‘the verbs “to supply”, “to furnish”, and “to pleasure”, in. situations where other terms could be employed just as correctly’ (Wells 1920, p. 267). 1 select a few of the many parallels cited by Wells. Particularly striking are those from. Michaelmas Term,10 the scene where Shortyard tries in vain to raise money:

    1. (p.265) (a) ‘… commend me to their loves; and I am. proud, say, that my occasions have found time to use 'em towards a supply of money’ (Tim., 2.2. 197 ff.). ‘Let them both rest till another occasion; you shall not need to run so farre at this time, take one near hand to Master Quasimodo the Draper, and will him to furnish me instantly’ (Michaelmas Term, 2.1.93–6).

    2. (b) ‘… I come to entreat your honour to supply; who having great and instant occasion to use fifty talents, hath sent to our lordship to furnish them’ (Tim., 3.2.17 ff.).

      ‘… Run presently to Master Goome the Mercer, and will him to tell out two or three hundred pound for mee, or more according as he is furnished’ (M.T., 2.1.81–3).

    3. (c) ‘My lord … has only sent his present occasion now … requesting your lordship to supply his instant use with so many talents’ (Tim., 3.2.3 5 ff.).

      ‘Run to Master Goome, or Master Profite, and carrie my present occasion of money to 'em’ (M.T., 2.3.156–7).

    4. (d) ‘I count it one of my greatest afflictions, say, that I cannot pleasure such an honourable gentleman’ (Tim., 3.2.56 ff.).

      ‘It is my greatest affliction at this instant, I am not able to furnish you’ (M.T, 2.3. 117–18).

    5. (e)What a wicked beast was I to disfurnish myself against such a good time …’ (Tim., 3.2.45 ff.).

      What a beast was I to put out my money tʼother day’ (A Mad World, My Masters, 2.4.26–7).

    These, and other parallels cited by Wells, formed conclusive evidence for Middleton's authorship of the satirical dunning scenes in Act 3.

    The whole justification for authorship study is that it helps us understand the nature of a collaborative play, and the differences it displays between the writers involved. When I worked on Shakespeare's prose I was struck by several oddities in the scene (1.2) where Timon feasts his friends. There was the great stylistic variation in Apemantus' speeches, which I could describe —with ‘their curious mixture of prose and couplets of ten, twelve, and eight feet’—but could not account for, and there was the strange oscillation between verse and prose in Timon's part, culminating in his prose speech in what seemed to me ‘an artificial style, with its inflated language … and its expanded images’.11

    O you gods, think I, what need we have any friends, if we should neʼer have need of 'em? They were the most needless creatures living, should we neʼer have use for 'em; and would most resemble sweet instruments hung up in cases, that keeps their sounds to themselves. (1.2.95–9; my italics)

    (p.266) This concluding image still seems affected, and is no doubt a deliberate effect, but it was not until I (belatedly) read William Wells's essay that I could identify its real author:12
    • I commend
    • The virtues highly, as I do an instrument
    • When the case hangs by the wall.
    • (Middleton, More Dissemblers
    • besides Women, 1.3.22–4)
    Wells also identified words and phrases characteristic of Middleton, such as ‘apperie’ (1.2.32) and ‘rioter’ (3.5.68), neither of which is found in Shakespeare. In 3.6 we find the phrase ‘draw near’, which occurs four times in Timon and often in Middleton, together with Middleton's favourite expletive ‘push’. Wells also pointed to a characteristic grammatical usage in Middleton, who ‘habitually dropped the personal pronoun in the nominative case’. Among the instances Wells cited are ‘Has only sent his present occasion now’ (3.2.34); ‘Must I take the cure on me? | Has much disgraced me in't’ (3.3.13); ‘How fairly this lord strives to appear fine! Takes virtuous copies to be wicked’ (3.3.30–1)—all Middleton scenes. The point is valid, but unfortunately Wells worked from a modernized edition. It was left to H. D. Sykes to consult the Folio text and to provide a more precise account.

    Of many other significant parallels cited by Wells, I quote one from the subplot, the scene where Alcibiades pleads with the Senators to issue a pardon for his friend, who has committed murder:

    Alcib.

  • Must it be so? it must not be. My lords.
  • I do beseech you know me.
  • Sec. Sen.

  • How?
  • Alcib.

  • Call me to your remembrances.
  • Third Sen.

  • What?
  • Alcib.

  • I cannot think but your age has forgot me;
  • It could not else be I should prove so base
  • To sue and be denied such common grace:
  • My wounds ache at you.
  • (3.5.90–7)
  • Both italicized phrases recur within a short passage in A Chaste Maid in Cheapside:

    (p.267) Sir Wal.

  • Touch me not, villain! my wound aches at thee,
  • Thou poison to my heart!
  • Allwit.

  • He raves already:
  • His senses are quite gone, he knows me not.
  • Look up, an't like your worship; heave those eyes,
  • Call me to mind! is your remembrance left?
  • Look in my face.
  • (5.1.15–20)
  • As Wells showed, ‘the same association of an aching wound with anger is seen in Your Five Gallants; ‘Forgive me, dear boy; my wound ached and I grew angry’ (3.3).

    This pioneer essay by William Wells, identifying Middleton as the co-author of Timon, inspired H. Dugdale Sykes to a more detailed analysis. Citing Fleay's theory that Shakespeare was the original author, into whose work an interpolator placed his own ‘inferior’ additions—a theory which had apparently convinced K. Deighton, editor of the first Arden edition (1905)·— Sykes set out to argue the opposite case, already urged by Wright and Delius, that ‘Shakespeare worked over an existing play, or draft of a play’.13 I shall not summarize Sykes's thesis, nor his attempt to prove the additional presence of John Day's hand (Sykes 1924, pp. 12–20), since neither argument carried conviction. The valuable parts of Sykes's essay were his general observations on the characteristics of Middleton's style found in Timon, and his detailed commentary on the individual scenes. Developing the observation made by Knight, Fleay, and E. H. Wright, that the play contains much more rhyme than any of Shakespeare's other plays of this date, a proportion rising to 20 per cent in some scenes (pp. 2–3), Sykes also echoed earlier scholars by observing that Timon had many ‘irregular unscannable verse lines’, and ‘frequent and aimless shifts from verse to prose’, all features ‘characteristic of Middleton at the time this play was written’. Sykes instanced The Phoenix (1604) and Michaelmas Term (1606) as being ‘most closely akin to Timon … In these plays we find the same high proportion of rhyme, the same irregularities of metre, the same habit of jumbling together verse and prose, rhymed verse and unrhymed’ (p. 22). The passages that Sykes quoted for comparison are well worth a modern reader's attention:

    • (a) Poor honest lord, brought low by his own heart,
    • Undone by goodness! Strange unusual blood
    • When man's worst sin is, lie does too much good!
    • Who then dares to be half so kind agen?
    • For bounty, that makes gods, does still mar men.
    • My dearest lord, blest to be most accurst,
    • (p.268) Rich only to be wretched, thy great fortunes
    • Are made thy chief afflictions. Alas, kind lord,
    • He's flung in rage from this ungrateful seat
    • Of monstrous friends;
    • Nor has he with him to supply his life,
    • Or that which can command it.
    • I'll follow and inquire him out:
    • I'll ever serve his mind with my best will;
    • Whilst I have gold, I'll be his steward still,
    • (Tim., 4.1.37–50)
    • (b) None can except against him; the man's mad,
    • And privileged by the moon, if he say true:
    • Less madness ’tis to speak sin than to do.
    • This wretch that lov'd before his food his strife,
    • This punishment falls even with his life.
    • His pleasure was vexation, all his bliss
    • The torment of another;
    • Their hurt his health, their starved hopes his store;
    • Who so loves law dies either mad or poor;
    • (The Phoenix, 4.1.138–47)
    • (c) The happiest good that ever Shortyard felt!
    • I want to be express'd, my mirth is such.
    • To be struck now eʼen when his joys were high!
    • Men only kiss their knaveries, and so die;
    • I've often mark'd it.
    • He was a famous cozener while he liv'd
    • And now his son shall reap't; I'll haʼ the lands,
    • Let him study law after; ’tis no labour
    • To undo him for ever; but, for Easy,
    • Only good confidence did make him foolish
    • And not the lack of sense; that was not it:
    • ’tis worldly craft beats down a scholar's wit.
    • (Michaelmas Term, 4.2.10–19)
    Looking at these three excerpts, with their incomplete verse lines and unpredictable use of couplets, we may well agree that ‘it would be difficult to find passages more alike in style’ (p. 23). Sykes pointed out that Apemantus' speeches at the first banquet {1.2.38–72) exhibit ‘the same features (irregular metre, and jumbling together of rhymed and unrhymed lines and prose) as are found in Middleton's plays at this time’ (pp. 3 2–3). This confusing mixture of styles is also found in the Steward's soliloquy (4.3.458–71), which has several echoes of earlier scenes (3.5, 4.2), proving that it, too, was written by Middleton. Sykes identified another characteristic of Middleton's style, his liking for ‘antithetical couplets’, shown three times in the first banquet scene (p.269) of Timon,14 and he noted that The Phoenix ‘contains above a score’ of such couplets (p. 34).

    Sykes praised Wells for pointing to the dunning scenes in Act 3 as providing ‘the clearest evidence of Middleton's authorship’ (p. 23), and cited even more parallels between them, and the acknowledged work of Middleton, who ‘… dwells much upon duns and debtors’ (pp. 24–9, 37). He also showed that these scenes contain several Middletonian expressions which are either never or seldom, used by Shakespeare. The word ‘free-hearted’ (3.1.10) ‘occurs nowhere else in the Shakespeare Folio’; the expression ‘give thee thy due’ (35), occurs only once in the rest of Shakespeare (I Henry IV, 1.2.59), but frequently in Middleton; and the word ‘endeared’ (3.2.31) regularly has in. Middleton, but nowhere in Shakespeare, ‘the sense of “obliged” or “bound” to a person’. Correcting a point made by Wells, Sykes pointed to other linguistic details indicating Middleton's hand, such as ‘the contraction, of “he has” to “hʼas”‘, which occurs six times in the folio text of Timon, always in Middleton's scenes.15 Another characteristic Middleton contraction, according to Sykes, is ‘'tas’ for ‘it has’, which occurs twice in another scene identified as Middleton's (1.2.49, 144). Both contractions ‘are so frequent in Middleton as to attract the attention of the least observant reader’, the two together occurring six times in Michaelmas Term and eight times in Your Five Gallants (p. 20). Returning to an observation of Fleay's, Sykes added a ‘purely bibliographical’ argument for the presence of two authors, namely the occurrence of the spelling ‘Apemantus’ in Shakespeare's 1.1 (10 times in. full, 33 times abbreviated as ‘Ape.’), as against ‘Apermantus’ in Middleton's 1.2 (4 times in full, 14 times abbreviated as ‘Aper.’). In 2.2 and 4.3, both by Shakespeare, the spellings are ‘Apemantus’ (9 times) and ‘Ape.’ (38 times, once as ‘Apem.’). Sykes recognized some exceptions—in 1.1 the stage direction is ‘Enter Apermantus’, and Timon once addresses him in that form; ‘Apermantus’ occurs once in. 2.2.76—but argued that here ‘for once the compositor failed to follow his copy’ (pp. 21–2).

    Between, them, developing earlier demonstrations by Knight, Delius, Fleay, and Wright that two clearly distinguishable authors had collaborated on Timon, William Wells and H. Dugdale Sykes provided challenging evidence that the second hand was Middleton's, But despite these six independent studies, which reinforced each other's findings, the Shakespeare world once again closed ranks against the entry of an outsider. The most common all-purpose validation of the case for Shakespeare's sole authorship was to revive the favourite theory of early nineteenth-century German, critics, and pronounce Timon an unfinished play. E. K. Chambers, a considerable scholar, (p.270) but seldom willing to conceive of Shakespeare having collaborated with anyone, recognized that the play contains several scenes ‘in which the verse at least cannot be the complete and jointed work of Shakespeare’, It is significant that most of the scenes in which Chambers pronounced the verse to be untypical of Shakespeare are those already identified as being by the second author, probably Middleton: 1.2; 3, 1; 3, 2; 3.3; 3, 4.1–79; 3.5; 4.2.30–50; 4.3.1–47, 464–543; 5.1.1–118. Indeed, the account Chambers gave of the un-Shakespearian quality of this verse closely echoed the description of Middleton's characteristic confusion of styles given by Wells and Sykes. These un-Shakespearian qualities are

    particularly noticeable in the longer speeches. These contain Shakespearean ideas, sometimes inchoate, and scattered Shakespearean phrases. But they are not constructed as articulated paragraphs at all. They consist of juxtaposed sentences, now in blank verse, now in rhyme, now in wording which can most easily be read as prose. There are many short lines, occasionally successive, for which no rhythmic or dramatic justification is apparent. There are unmetrical long lines. It must be added that the structure of Timon as a whole is incoherent. There are many small confusions and inconsistencies.16

    Chambers reported that these features had given rise to theories which agree ‘in the assumption of a second hand’ at work—the hostile term ‘assumption’ being evidently chosen to deny such theories the status of rational argument. Making easy capital out of the disagreement between scholars as to who the second author was (the ‘multiplicity [of theories] suggests that their exponents are on the wrong tack’), Chambers avoided discussing any of the voluminous evidence already assembled, and curtly dismissed the issue:‘I do not doubt that [Timon] was left unfinished by Shakespeare, and … that it is unfinished still. The passages of chaotic verse, in particular, look very much like rough notes, hastily jotted down to be worked up later’ (Chambers 1930, i. 482). But Chambers hastened to deny that Shakespeare habitually worked in such a chaotic manner: ‘I do not suggest that Timon throws much light upon Shakespeare's normal methods of working. It is, perhaps, a subjective view that he dealt with it under conditions of mental, and perhaps physical stress, which led to a breakdown’. So, after this life-crisis, Shakespeare ‘seems to have abandoned [the play), and never to have taken it up again’ (p. 483).

    To anyone who has read the detailed case for a co-author's contribution already made by six careful scholars, it must seem that Chambers simply ignored all the evidence not suited to his biographical construction of a Shakespeare writing ‘chaotic verse’ on the verge of a nervous breakdown, and never returning to the residuum of that unhappy period. It is hardly good scholarship, needless to say, to take refuge in biographical speculations simply (p.271) to avoid an awkward issue. It is particularly disappointing that Chambers himself should have described the chaos of styles in some scenes of Timon in terms that exactly echoed those made by two Middleton scholars identifying this characteristic feature of his early plays, yet refused to acknowledge the strength of this evidence of a second hand at work. It is also regrettable that Chambers did not refer to Delius's 1867 essay, which cogently refuted the theory of Timon being an unfinished play, the ‘completed’ and ‘uncompleted’ parts being so thoroughly intertwined. Predictably enough, many Shakespearians approved of Chambers reviving this older theory of an unfinished play, for it provided an easy answer to anyone tempted to propose a co-author. Una Ellis-Fermor developed the Chambers position in 1942, in an often-cited essay,17 arguing that the large number of ‘broken and irregular lines, the patchwork effect’ of many speeches are the normal signs of a writer ‘roughing out a scene’ (Ellis-Fermor 1961, p. 163). In order to show ‘the characteristics of these broken speeches’, Ellis-Fermor quoted from Alcibiades' speech to the Athenian senators (3.5.38–58), a passage which includes several irregular lines and four rhymed couplets (45–6, 52–7), arguing that this ‘preliminary rush of isolated fragments’ was a common experience for anyone who ‘has ever written blank verse’ (p. 165). This claim left Ellis-Fermor open to the objection that arrived, with some delay, from David Lake, who declared her theory ‘fantastic. In my own experience as a practitioner of verse (and I have written hundreds of lines of couplets and unrhymed verse), couplets are not a prelude to any tolerable kind of blank verse—indeed, they block the development of run-on blank verse of the kind used by Shakespeare about the presumed date of Timon (c.1607)’, and which are found in Timon ‘in the clearly Shakespearean sections such as 1.1.1–176’,18

    As that riposte shows, the theory of Timon being an unfinished play forces critics to ignore the many visible differences between some parts of the play and others, differences which Knight, Delius, Fleay, Wright, Wells, and Sykes had defined by a number of scholarly approaches—metre, the proportion of rhyme, unmotivated oscillations between verse and prose, verbal parallels, linguistic preferences, orthography, even bibliographical variants—and which two of them had shown to typify, respectively, Middleton and Shakespeare. But upholders of ‘Shakespeare the Non-collaborator’ have never had any difficulty in ignoring unwelcome evidence, as the editors of the two major scholarly editions in modern times showed, J. C. Maxwell in his ‘New Cambridge edition’ (1957), and H, J. Oliver in his volume for the ‘New Arden’ Shakespeare (1959).19 Maxwell devoted only three pages (pp. ix–xi) (p.272) to the arguments for co-authorship—or, as he called it, ‘disintegration’. He briefly alluded to the main lines of two opposed theories (Shakespeare rewriting an earlier play, or being in turn rewritten), but cited none of the detailed evidence that had accumulated from Knight in 1849, Delius in 1867, and on up to Sykes in 1924 concerning the strikingly discrepant treatment of prose and verse—clearly distinguished in Shakespeare's scenes, but in the scenes ascribed to Middleton mixed with each other, and with rhymes. In his dismissal of the authorship issue Maxwell never even mentioned the essays by Knight, Delius, Fleay, Wells, and Sykes, even though the last two scholars published in Notes and Queries, the journal of which Maxwell was co-editor from 1959 to 1976.20 Nor did he draw the reader's attention to the play's high proportion of rhyme, which is completely unlike any authentic play by Shakespeare at this point in his career. Maxwell did mention E. H. Wright's 1910 monograph, even including it in his bibliography (1957, pp. x, 100), but he never cited it, and certainly did not mention that Wright, like the five other scholars who had investigated the co-authorship issue, had distinguished two different authors in terms of style, characterization, and dramaturgy. Maxwell dismissed Wright's book in the manner of E. K. Chambers, trying to make it seem, self-defeating,21 before observing that in the authorship debate ‘the tide soon began to turn’ against acknowledging any ‘non-Shakespearian matter’ in the play (p. xi)—a metaphor favoured by the ‘fundamentalist’ camp. Maxwell declared that the Charnbers/Ellis-Fermor theory of the play being ‘a Shakespearian rough draft that had never been completed’ is now ‘generally accepted’.

    But although Maxwell totally rejected the case for co-authorship in the Introduction, his notes reveal some tensions. He several times cited specific Middleton parallels, as for the ellipsis in Timon's sententious observation that

    • ’Tis not enough to help the feeble up,
    • But to support him after
    • (1.1.110–11)
    —the ‘But’ implying ‘but it is necessary to’—where he quoted a passage in Middleton (More Dissemblers besides Women, 1.3.38–9):
    • ’Tis not enough for tapers to burn bright,
    • But to be seen.
    Maxwell scrupulously recorded that this parallel was cited by H. D. Sykes in Notes and Queries, 1 (1923): 167, ‘who attributes the Tim. passage to (p.273) Middleton’ (p. 107), On the one hand, since Maxwell gave his readers no information either as to the number of parallels cited by Sykes or to the full argument for Middleton's presence, this brief acknowledgement lessened Sykes's credibility. But on the other hand, the fact that Maxwell thought it worth recording, adding another Middleton instance, shows that it had aroused some agreement in him. Maxwell noted that Sykes attributed to the co-author 3.5, Alcibiades' fruitless appeal to the Senate (not observing that Knight, Delius, Fleay, and Wright had also done so), and he took over another Middleton parallel, between a Senator's complaint against Alcibiades,
    • Your words have took such pains as if they laboured
    • To bring manslaughter into form
    • (3.5.26–7)
    —that is, an acceptable legal form or procedure—and a passage that Sykes had cited from Middleton's Phoenix (4.1.6–9): ‘I'll strive to bring this act into such form | And credit among men&’. Maxwell added several other Middleton parallels on his own account,22 including a suggestive resemblance in one of the dunning scenes that we may confidently ascribe to Middleton, Sempronius' aggrieved remark:
    • I was the first man
    • That eʼer received gift from him
    • And does he think so backwardly of me now,
    • That I'll requite it last?
    • (3.3.16–19)
    which he compared to The Phoenix, I.I.47: ‘I did not think so unfashionably of you’. Finally, since Maxwell had at least read the work of Wright, who showed how the non-Shakespearian scenes jumble prose, verse, and rhyme indiscriminately, it is revealing that he should ignore that point and in his own edition strive to regularize the prose/verse distinction in scenes that we can attribute to Middleton, even following Alexander Pope's notorious attempts to regularize Shakespeare's verse.23 Maxwell's silence on this issue is a disappointing exception to his otherwise open-minded scholarship.

    The other major edition, two years later, by H. J. Oliver for the New Arden series, showed again how Shakespearians can be encouraged to close their mind on the question of co-authorship. Following the example of Chambers, Oliver lumped together all the candidates proposed as co-author, including an obviously dotty claim that the play was an allegory of Francis Bacon's treatment of Essex, and asserted that ‘the various theories identifying the (p.274) co-author also cancelled each other out, of course’ (Oliver 1959, pp. xxii–xxiii)—of course. Like Chambers, Oliver did not deign to quote any of the many parallels already provided between the Middleton scenes and Middleton's known work, and he erroneously reported that proponents of co-authorship assigned 1.1 and 2.2 to Middleton.24 In any case, Oliver decided, ‘such theories defeated their own object’, for they began by wanting ‘to explain the presence in. the play of “inferior” work and almost without exception ended up by assigning to Shakespeare's collaborator some of the best scenes in the play (e.g. the first three of Act 3) as well as parts that were crucial in the plot (such, as 2.2)’ (p. xxii). That remark was intended to make these eccentric theorists look ridiculous, but the fact that Oliver liked the dunning scenes in Act 3 does not mean that they must have been written by Shakespeare, and he breezily ignored the extraordinary number of parallels drawn by Wells and Sykes with Middleton's speciality, dunning scenes (after all, Middleton was a pioneer in Jacobean city comedy from whom Shakespeare could learn). Again, the fact that 2.2 is essential to the action only means that the co-authors had worked out their plot properly, and those scholars who claim part of it for Middleton (perhaps lines 1–45, 120–233) also detect Shakespeare's hand. Boosted by his own assertiveness, Oliver simply dismissed all the verbal parallels, claiming that ‘the Oxford English Dictionary has … shown most of this alleged evidence in a new light and has made it impossible to believe any longer that most of the words quoted were in fact characteristic of one dramatist’. This was a particularly devious statement, for of course Wells and Sykes had access to the OED, and when discussing individual words they were careful to document their presence in Middleton and their (frequent) absence in Shakespeare. But in any case, they mostly quoted much longer verbal collocations, even whole sentences, syntactical, or idiomatic constructions which went far beyond the scale of single words, and showed them to be indeed characteristic of Middleton. Sweeping away all unwelcome evidence for co-authorship, Oliver fell back on the Chambers case for the play being unfinished, only disagreeing with. Sir Edmund's claim that this mode of composition was abnormal for Shakespeare, the product of a mind on the verge of a nervous breakdown. Rather, Oliver countered,

    Timon would suggest that thoughts often came to him in a kind of incomplete verse form, sometimes in prose and sometimes (interestingly) in rhyme, and that only on revision did the text evolve into, predominantly, blank verse, (p. xxviii)

    Once again we see that a ‘Shakespeare sole author’ case, while consciously denying the evidence that Middleton regularly wrote in a jumble of verse (p.275) forms, is often forced to make other speculations, here biographical. Oliver somehow knew that Shakespeare's ‘thoughts often came to him in a kind of incomplete verse form’, and he even added a further deduction, that the dramatist ‘wrote scenes as he felt in the mood for them, not bothering to complete one if at the minute he was more interested in another’. Theories of this generality, and generosity, can be used to account for any kind of anomaly in any text, and are fundamentally unscholarly.

    Another editor who could not reconcile the oddities he noticed in Timon of Athens with Shakespeare's normal compositional practices was G. R. Hibbard, in his New Penguin edition.25 Hibbard noted several respects in which the play differed from other Shakespeare tragedies: the hero is alive when we last see him, leaving the stage ‘on his own two feet’ (Hibbard 1970, p. 7); he is a socially isolated character, having neither ‘close personal relationships’ nor friends (pp. 8–9); we know nothing of Timon's past, indeed he is ‘the most generalized of all the tragic heroes’ (pp. 9–10); the play ‘comes much closer to fable and parable than does any other of the tragedies’ (p. 11). These are all justified observations on the peculiarities of the play's plot, which (I suggest) proved an inauspicious choice for dramatic treatment. Hibbard also noted deficiencies in the play's dramaturgy. The way in which Alcibiades ‘is incorporated into the general design is far from satisfactory’, he judged (p. 13). In his first appearance (1.1) he is a minor character, but in 3.5 ‘he suddenly and quite unexpectedly assumes a major role as he confronts the Senate and pleads passionately for the life of his friend. Viewed in isolation, the scene is highly dramatic, quite the most exciting thing in the play so far, but it loses some of the impact that it ought to make, because it has not been prepared for’ (p. 13). The ‘whole incident remains tantalizing and obscure when it ought to be explicit, for it is of considerable structural importance’ in suggesting that the ingratitude shown to Timon by the rulers of Athens is a recurrent feature of that city (p. 14). Similar uncertainty, Hibbard complained, surrounded other key elements in Timon's relationship to Athens, the fact that he had performed the state some service (2.2.201–4, 4.3.93–6), giving it both financial and military help. ‘Had we known’ about these details from the beginning, Hibbard argued, we would have appreciated more clearly the city's ingratitude towards Timon, and the ironic reversal by which two of its Senators come to ask for his help (5.1.1 32–66), a scene as equally unprepared for as Alcibiadesʼ petition to the Senate.

    Something odd is happening when in a play by Shakespeare, who in the rest of his work is such a master in the art of plotting, it is possible to point to inadequacies in this respect that badly blunt the impact of what should have been two of the most significant scenes in the play. (pp. 15–16)

    (p.276) Hibbard listed other elements in the play's dramatic structure that he found ‘otiose and irrelevant’.

    Turning from the play's dramaturgy to its style, Hibbard complained that its ‘writing … is uneven’, in many places ‘comparatively thin’ (p. 16). Like so many scholars in the century or more since Knight and Delius wrote (although he cited none of them), Hibbard instanced several passages ‘in which regular blank verse, what appears to be highly irregular blank verse, and what can most easily be read as prose, are found mixed with each other in an erratic and unpredictable manner’ (p. 17). Hibbard pointed out that the scene between Alcibiades and the Senate,

    though written in verse throughout, also has some unusual qualities. A high proportion of the lines are short lines, and the transitions from one idea to another are strangely abrupt. The firm yet flexible control of the complex verse paragraph which is one of the most impressive characteristics of Shakespeare's mature writing is absent from this scene. (pp. 17–18)

    Hibbard was an experienced scholar, sensitive to Renaissance language, and he documented some striking anomalies in the play's dramaturgy and style (although, oddly, he did not notice the high proportion, of rhyme in. this scene). All these anomalies could have been accounted for on the supposition that Shakespeare had a co-author. Indeed, as we shall see in Chapter 7, it was precisely these uncharacteristic dramaturgical weaknesses that provoked nineteenth-century scholars to look for evidence of two authors at work. But Hibbard merely concluded that ‘Timon of Athens is not … a completely finished play’, and that ‘some scenes … still needed to be fitted in when Shakespeare laid the draft, by’ (p. 17). In his ‘Account of the Text’ Hibbard discussed the several striking discrepancies in spelling (Apemantus/Apermantus; Ventidius/Ventigius; Flavius/Steward), ‘which must go back to the manuscript’ (p. 259), and briefly considered the ‘possibility that Shakespeare had a collaborator’, only to dismiss it: ‘this seems unlikely in view of the consistency of the play's imagery’ (p. 260). Instead, Hibbard fell back on the comforting idea that Timon ‘was composed in an irregular manner’, due to ‘Shakespeare going off to Stratford, for example, and leaving the completed portions of the play behind him in London’ (p. 261). Although he included a section on ‘Further Reading’ (pp. 45–9), which was even updated in April 1980, Hibbard at no point mentioned the studies identifying Middleton as the co-author. Hibbard's edition, excellent in many ways, is still in print, and one may well hope that Penguin, or its general editor (Stanley Wells) will soon add a note pointing out that, on the issue of the play's joint authorship, Hibbard, like so many of his predecessors, simply avoided the issue.

    The orthodoxy of ‘Shakespeare the Non-collaborator’ undoubtedly proved influential on many generations of readers—indeed, I confess that the (p.277) combined authority of the New Cambridge and Arden editions of Timon blocked me for many years from looking at the issue afresh. Yet, as we see time and again when anatomizing its defensive posture, it has shown itself to be not only conservative but evasive, ignoring genuine scholarship while clinging to a mantle of authority. Fortunately, the tradition of scholarly enquiry that has animated authorship studies proved its vitality once again in the 1970s, when David Lake and MacDonald Jackson, working independently on the Middleton canon, converged on Timon of Athens with a new set of methods. The effect of their work, as in several other instances we have been following, was not to overturn the authorship divisions made by earlier scholars, but to restate them with new force and precision.

    David Lake found that the identifications of Middleton's share carried out by William Wells and H. D. Sykes made the case for Middleton's participation ‘strong enough to warrant a careful check by means of fresh evidence’, drawn from the Spevack concordances. Lake compared both writersʼ preferences in colloquialisms and contractions, showing that

    through all his work Shakespeare prefers them to ʼem and hath to has; he makes very little use of I'm (only 5 authentic instances outside Timon) or of ʼHas for he has (14 authentic instances outside Timon). Middleton, of course, shows the opposite preferences and makes considerable use of I'm and ʼHas. (Lake 1975, p. 281)

    The total figures for Timon show that the relative frequencies differ greatly from those of other late plays. ‘The ʼem: them ratio of Timon is higher than that of any unassisted Shakespeare late play except The Tempest; the has: hath ratio is higher than any (has being uniquely more frequent than hath), and so are the relative frequencies of I'm and’Has (p. 282). Even more significant is the distribution of these features throughout the play. Accepting the Wells-Sykes ascription of Middleton scenes (1.2; 3.1; 3.2; 3.3; 3.4; 3.5; 3.6; 4.2.30–51; 4.3.458–536), Lake compared the distribution of these features in Timon, as presented in Table 4.4. Middleton's scenes have no instance of the Shakespearian forms doth and moe (for ‘more’), indeed Middleton

    Table 4.4. Colloquialisms and contractions in Timon of Athens

    Ascription

    Length

    'em

    them

    has

    hath

    does

    doth

    I'm

    I am

    ʼHas

    'tas

    moe

    Middleton scenes

    897 lines

    16

    16

    25

    8

    16

    0

    3

    13

    4

    2

    0

    Shakespeare scenes

    1,418 lines

    4

    50

    6

    21

    8

    9

    0

    27

    1

    0

    4

    Whole play lines

    2,315

    20

    66

    31

    29

    24

    9

    3

    40

    5

    2

    4

    Source: Lake 1975, p. 283.

    (p.278) never uses moe, and only uses doth once in each of his early comedies, and twice in The Revenger's Tragedy. Lake commented that ‘the ratios of ʼem to them and has to hath are about normal’ for Shakespeare's late plays (in Lear they are 9:45 and 13:49; in Ant. 3:52 and 22:43, respectively), and that in the Middleton scenes the ratios for these two features, and for I'm: I am are all within the dramatist's normal range (p. 284). Some of Middleton's linguistic characteristics, however, are underrepresented (the oath faith, the contractions I'thʼ, and theyʼre), and one (I've) does not occur (p. 285). But of course collaboration does affect the styles of both writers, and ‘accommodation theory’ predicts that each will suppress some aspects of his normal usage while adapting to the other.

    MacDonald Jackson, the most inventive scholar in attribution studies over the last thirty years, while also endorsing the authorship divisions proposed by Wells and Sykes, extended them in several new directions. He pointed out further parallels with Middleton's usage, for instance, the fact that ‘shine as a noun, a well-known favourite of Middleton's which Shakespeare uses in no other play … appears at 3.5.101’.26 Like Knight, Delius, Fleay, and Wright, Jackson also diagnosed a ‘stylistic unevenness’ in the play which would be untypical of Shakespeare, for ‘at times the diction degenerates into an uncharacteristic medley of prose, blank verse of varying degrees of regularity, and lurching rhyme’ (p. 54), and he agreed with Wells and Sykes that this stylistic jumble was typical of Middleton's early comedies. Endorsing Sykes's observation about the presence of ‘antithetical couplets’ in the Middleton-ascribed scenes and in Middleton's known work, Jackson added (p. 61) that Timon also contains two instances of another Middleton trait, ‘gnomic couplets of the kind in which one line begins with who (meaning “he who”) plus verb’, as in:

    • And with their faint reply this answer join:
    • Who bates mine honour shall not know my coin.
    • (3.3.25–6)
    • And this is all a liberal course allows:
    • Who cannot keep his wealth must keep his house.
    • (3.3.40–1)
    with which he compared Middleton's acknowledged lines:
    • O, they must ever strive to be so good!
    • Who sells his vow is stamp'd the slave of blood.
    • (The Phoenix, 2.2.23–4)
    Like Sykes, Jackson cited a long passage of verse from Middleton (Phoenix, 1.4.197–227) to show that his style regularly produces what apologists for (p.279) Shakespeare's sole authorship regard as ‘Shakespearian jottings which remain to be worked up into fully coherent blank verse’ (p. 61).

    Jackson analysed the play's use of contractions and colloquialisms independently of Lake, but reached identical conclusions. He noted (pp. 56–7) that the ratios for hath and has, doth and does vary considerably between the Middleton and Shakespeare scenes, as Wells and Sykes had defined them, but in each case within the normal range for each writer.27 Jackson then showed that the distinction between these two groups of scenes is further strengthened once we observe the occurrence of ‘eight contractions, hʼas, hʼad, 'tas, haʼ, 'em, I'm, eʼen, and neʼer, [which] occur uniquely or with unusual frequency for Shakespeare in Timon, and … mostly in those scenes in which the treatment of hath, has, doth, and does was found to be markedly un-Shakespearian’ (pp. 57–8). When different approaches within authorship studies produce identical results, we can be sure that the methods being used are reliable. Jackson reported that all eight contractions are ‘known Middleton favourites’, but that even if one omits hʼas and hʼad, the figures for the remaining six still support the division between the two authors. Proportionally measured, none of the ‘one hundred non-Middleton plays’ that Jackson studied had a higher total for these contractions than the Middleton part. Jackson argued that no one could ‘satisfactorily explain this agreement between two different sorts of evidence except on the assumption that [several scenes in Timon] really are largely Middleton's work’ (p. 63).

    Jackson added several further tests, one involving the frequency of thirteen selected ‘function words’ in the play. Taking a thousand-word sample from each section of the play, Jackson (using a chi-square ‘goodness of fit’ test) compared it with three other plays by each dramatist. The Shakespeare sample turned out to correlate very closely with Othello, Macbeth, and King Lear, but not at all well with the Middleton samples. Conversely, the Middleton sample from Timon correlated very closely with Women Beware Women, The Phoenix, and Michaelmas Term (pp. 89–90). Moreover, ‘the “Shakespearian” and “Middletonian” portions are extremely divergent from each other’ (p. 90). Jackson achieved a similar result when applying his ‘rare word’ test, which showed that the Middletonian portion of Timon had ‘proportionally fewer “rare” words than the rest of the play’, and stood in a ‘less plausible and clear-cut chronological relationship (as determined by an analysis of the vocabulary) to Shakespeare's other plays’ (p. 1 55). In the ‘Shakespearian’ part of the play ‘rare’ words occurred once in every eight lines (‘about right’ for this period in his career) but in the Middletonian part once only in every thirty-eight lines, a much lower rate than for any Shakespeare play, and almost five times less frequently than in the Shakespearian scenes. Referring to the ‘vocabulary index’ that he had (p.280) constructed, Jackson found that the figures for the Shakespearian portion of Timon would date the play to 1604–5, a date supported by other evidence, whereas the Middletonian figures would place it in 1594–5, ‘a ludicrously inappropriate date …. I see no convincing way of reconciling this discrepancy between the “Shakespearian” and “Middletonian” portions of Timon with the assumption that Shakespeare was sole author of the play’ (p. 155).

    Jackson's final piece of evidence unsettling that assumption was bibliographical, drawing on Charlton Hinman's work on the Folio, which showed that Timon was printed by one craftsman, Compositor B, that the names Apemantus and Ventidius were ‘differently spelt and abbreviated in different places’, and that ‘the variant spellings fall into more or less well-defined groups’.28 Since only one printer was involved, Hinman reasoned that this variation must reflect peculiarities in the copy, for ‘such patterned alternations inevitably suggest different hands’, perhaps ‘scribal or (as for various reasons seems to me much more likely) authorial’. In other words, the compositor accurately reproduced the forms used in each author's manuscript. Jackson showed that the variants Apermantus and Aper. occur consistently in Middleton's 1.2, as do Ventigius and Ventig. (also in Middleton's 3.3), whereas the Shakespeare scenes have Apemantus and Ape., Ventiddius and Ventid. (p. 213). Thus Fleay's observation a hundred years earlier was validated, Jackson (following J. C. Maxwell's 1957 edition) also showed that the two conflicting valuations of the Greek talent corresponded to the authorship division, ‘the four unequivocal references to a “currently small” number’ occurring in Shakespeare's 1.2 and 2.2, while the four inappropriate or confused references occur in Middleton's 3.2. There are also four references in the play to an inappropriately large number, two of which occur in the Middletonian 3.1 and 3.4,

    the other two at the end of 2.2 within the dialogue between Timon and his Steward which prepares for Act 3. This short dialogue contains references to both small numbers (2.2.230, 233) and large numbers (2.2.197, 203). The large numbers are only six lines apart, and each has the contraction 'em in the line which precedes it; an example of the parenthetical ‘say’, which Wells pointed to as characteristic of Middleton, occurs in the same context. So there may be some of Middleton's writing in the latter part of 2.2, as both Sykes and Wells believed. (p. 214)

    Where earlier scholars believed that Shakespeare, in the process of writing Timon, came to learn the value of ancient Greek money,29 we can now agree with Jackson that ‘Middleton was uncertain of the value of a talent, and that he was responsible for the references to large and indefinite numbers’ (ibid.), while Shakespeare had a much more accurate sense of its worth. (p.281)

    Table 4.5. Exclamations in Timon of Athens

    Oh spellings

    O spellings

    Shakespeare scenes

    0

    14

    Middleton scenes

    13

    10

    Source: Jackson 1979, pp. 214–15.

    Jackson added to the varied and detailed arguments a previously unnoticed piece of bibliographical evidence which, as he put it, ‘comes near to clinching the case for recognizing Middleton's presence in Timon’—many readers will think that case already clinched. This concerns the spelling distinction between Oh and O, which most sixteenth- and seventeenth-century compositors reproduced as they found it in their copy. Middleton, to judge from his holograph of A Game at Chess, ‘strongly preferred the spelling Oh: in the Trinity College MS he uses Oh 36 times, O only once. Shakespeare, in contrast, seems to have preferred O’, as Jackson showed from an analysis of eight good Quartos, which have ‘an overwhelming preponderance of O spellings’, a total of 540 as against 31 of Oh (pp. 214–15). Since, as seems probable, the folio text of Timon was set from the authors' foul papers, it is likely to retain authorial spellings, as indeed Jackson's computation showed (see Table 4.5). Statistical tests show that the chances of this disparity being fortuitous are less than one in five hundred.

    The combined studies of Knight, Delius, Fleay, Wright, Wells, Sykes, Lake, and Jackson, drawing as they did on an extensive knowledge of Elizabethan and Jacobean drama, using a varied set of approaches, and synthesizing impeccably scholarly work by other writers, have established Middleton's co-authorship of Timon beyond any doubt. In the Oxford Shakespeare, gratifyingly enough, Stanley Wells, who had been unwilling to recognize Peek's presence in Titus Andronicus, endorsed the case for Middleton's co-authorship, reporting that ‘during the 1970s and 1980s strong linguistic and other evidence has been adduced’ on its behalf (p. 997). Of course, the case is much older than that, for Wells in effect was endorsing the allocation made by his namesake William Wells and H. Dugdale Sykes in the 1920s. In the Textual Companion to the Oxford edition Gary Taylor described the work of Wells, Sykes, Lake, and Jackson, together with the still unpublished dissertation by E. V. Holdsworth,30 as providing ‘extensive, independent, and compelling evidence that approximately a third of the play’ (38.7 per cent, if one (p.282) accepts David Lake's division), was written by Middleton. As Taylor summarized Holdsworth's case, based upon study of the whole Middleton canon,

    Middleton's presence in Timon is indicated by the distribution of (a) linguistic forms, (b) characteristic oaths and exclamations, (c) function words, (d) rare vocabulary, (e) characteristic stage directions, (f) verbal parallels, (g) spellings, (h) inconsistencies of plotting, (i) rhyme. Holdsworth's investigation of verbal parallels for the first time comprehensively compares every phrase in an entire play with the complete corpus of both candidates for authorship; although, as might be expected, each author occasionally uses phrases which occur in the other's works, the great bulk of the verbal parallels, and all of the most striking ones, fail into distinct patterns, corresponding to the division of authorship already established on other grounds. The consistency of all. these independent forms of evidence cannot be plausibly dismissed. (TxC, p. 128)

    Two of the other Oxford editors, John. Jowett and Stanley Wells, agreed that the case has been made ‘conclusively’, and gave some account of Holdsworth's authorship division (p. 501). Apparently, Holdsworth's researches endorsed the accepted attributions, giving Middleton 1.2, nearly all. of Act 3—which the Oxford edition divides into seven scenes, making 3.4.103–117 (Riverside text) into 3.5—and 4.3.459–536. In addition, they report, Holdsworth assigned to Middleton. 1.1.272–83, part of 2.2.118–233, and 3.7 (in other editions, 3.6), lines 1–36, 104–10. In fact, all of these passages had been singled out by earlier scholars as showing Middleton's hand. H. Dugdale Sykes pointed out that the brief dialogue at the end of 1.1 between two Lords who agree to go ‘in | And taste Lord Timon's bounty’ (1.1.273–4) links up with the scene following, by Middleton, where Cupid greets the host:
    • Hail to thee, worthy Timon, and to all
    • That of his bounties taste!
    • (1.2.122–3)
    And this in turn links up with the First Stranger's claim that ‘I never tasted Timon in my life’ (3.2.77–8), which may now be confidently ascribed to Middleton. Holdsworth's assignation to Shakespeare of 3.6.37–103 recognizes his hand in Timon's two speeches at the second banquet, first the mock prayer in prose, before the parasites are served dishes of warm water, followed by Timon's devastating denunciation, in verse (‘You knot of mouth-friends!’). The verse speech was ascribed to Shakespeare by Delius in 1867, Fleay in 1874, Wright in 1910, Wells in 1920, and Sykes in 1923. All five scholars assigned to Middleton the final sequence of 4.3, the dialogue between Timon and his Steward. Once again, more varied and detailed research at the end of the twentieth century has confirmed the findings of pioneers up to a hundred years earlier.

    Although Holdsworth's main case for Middleton's co-authorship of Timon has not yet been published, he recently itemized some eighteen close verbal parallels between Middleton's scenes and A Yorkshire Tragedy, whose ascription (p.283) to Middleton now seems certain. The most striking of these is between Alcibiades' sophistic attempt to exonerate his unnamed friend from a charge of murder:

    • in hot blood
    • ………
    • Seeing his reputation touch'd to death,
    • He did oppose his foe
    • (3.5.11, 19–20)
    This situation, of a man ‘goaded into fighting a duel to defend his honour’, actually occurs in A Yorkshire Tragedy, where the Gentleman prepares to fight the Husband to defend his reputation:
    • I am past my patient bloode: shall I stand idle
    • And see my reputation toucht to death?
    • (ii. 166–7)
    Apart from the resemblance between the two situations, Holdsworth could ‘find no evidence that “touched to death” was a common idiom’.31 Several of the parallels Holdsworth cited (pp. 23–4) are indeed striking (A Yorkshire Tragedy is quoted first):
    • he that has no coyne | Is damnd
    • (ii. 28–9)
    • moulten coine be thy damnation
    • (3.1.52)
    • himselfe withered with debts (iii 11)
    • Debts wither 'em
    • (4.3.531)
    • pollitick … subtiller then nine Devils
    • (iii. 54–5)
    • the divell … made man Politicke
    • (3.3.28)
    • man … in the mire
    • (iv. 76–7)
    • man i'thʼmire
    • (1.2.59)
    • [both are denunciations of drunkenness]
    • (p.284) I bleede for you
    • (ix. 14)
    • I bleed inwardly for my Lord
    • (1.2.205)
    • [in both bleed = grieve]
    In addition to these verbal parallels, the general situation in the two plays is very similar, both protagonists being spendthrifts. The Steward in Timon laments that his master is ‘so senseless of expense | That he will neither know how to maintain it, | Nor cease his flow of riot’ (2.2.1–3), while the Wife in A Yorkshire Tragedy complains that ‘My husband never ceases in expence’, and warns that ‘Ryotts child must needs be beggery’ (ii. 2–5). And in due course, as Holdsworth puts it, ‘having squandered their wealth, both are driven to vindictive misanthropy and despair by an obsessive consciousness of poverty and “despʼrate want” (Timon, 4.3.462)’ (p. 24). Holdsworth made the case for Middleton's co-authorship of Timon so convincingly that we can look forward with some impatience to him publishing all his evidence.

    At this point, chronologically speaking, one might be able to consult Marina Tarlinskaja's work on Shakespeare's metrics in the hope of finding an independent approach to the problem, which would either corroborate or extend the work of these seven scholars. Unfortunately, Tarlinskaja did not study this play with her normal care, and she seems to have been confused by the rival claims of Winifred Nowottny (in a vague and impressionistic article)32 that Act 5 was largely un-Shakespearian, and MacDonald Jackson's much more scholarly ascription to Middleton of 1.2, Act 3, and parts of Act 4 (Jackson 1979, p. 48). Regrettably, Tarlinskaja then conflated these heterogeneous analyses to conclude that ‘the whole of acts III and V are suspected of not belonging to Shakespeare’.33 She provided stress profiles for whole acts, but—unsurprisingly, given this confusion of methods—was unable to draw any firm conclusion (Table 3.1, p. 101; pp. 195–6). Her hesitation here may also have been due to the fact that while she provided comparative figures for other Elizabethan dramatists, including Norton and Sackville, Marlowe, Jonson, Beaumont, and Fletcher, she seems not to have analysed Middleton's metrical practices. It is unfortunate that a book so helpful in other respects should have failed to treat Timon of Athens in the light of (p.285) modern scholarship, especially since the earlier study of pause patterns in Renaissance drama by Ants Oras also failed to follow up the evidence for Middleton's authorship, and only provided statistics for the play as a whole.34 It would be a distinct service to Shakespeare studies if future researchers could apply these valuable techniques for analysing verse to the authorship division so clearly established by the seven scholars whose work we have examined.

    However, some impression of the metrical differences between the two authors can be achieved if we return to the statistics provided by E. H. Wright, and rework his data according to the division of authorial responsibility which has achieved consensus in recent years. The information will not be perfectly comparable, since Wright assigned a much shorter part of 4.3 to Middleton than recent scholars would (Wright 1910, p. 5 6), and it is not possible to recalculate his data within a scene. That reservation apart, the metrical figures presented in Table 4.6 show a clear distinction, indicating that Shakespeare was more prone to use feminine endings than Middleton, and almost twice as likely to favour run-on lines. Conversely, Middleton used rhyme nearly six times more often than Shakespeare (this figure, 19.6 per cent, compares well with ray calculation for The Revenger's Tragedy, 17.3 per cent), and produced irregular verse lines five times more often. All four tests clearly distinguish the two dramatists.

    The division of authorship between Shakespeare and his co-author, solidly established by metrical and linguistic studies, was recently confirmed by Jonathan Hope's important analysis of Shakespeare's grammatical preferences from the viewpoint of socio-historical linguistics.35 Hope first studied the use of the auxiliary do in Timon of Athens, that is, the degree to which Renaissance dramatists adopted the ‘regulated’ form, which is now standard (‘I went home’; ‘I did not go home’, etc.). Hope had shown earlier that this grammatical feature, expressed as a relative percentage of regulated sentences, could distinguish authorial preferences: the regulation rate for Shakespeare ‘never exceeds 84 per cent’, while ‘no other author shows a regulation rate below 85 per cent’ (p. 17). From the statistical appendices, where Hope displayed all his data, we can see that the two Shakespeare samples (six early, ten late plays) showed an average regulation rate of 81 and 82 per cent respectively, while the Middleton sample (five plays) gave a figure of 90 per cent (pp. 156–7), Many scenes in Timon are short, and in statistical terms ‘give too small a sample to be judged’, but Hope found enough instances in 1.1 and 1.2 to perform a viable comparison. He showed that the figure for 1.1, 82 per cent, was typical for Shakespeare, while 1.2, with a figure of 90 (p.286)

    Table 4.6. Verse tests for Timon of Athens

    Shakespeare scenes

    Verse lines

    Feminine endings

    Run-on lines

    Rhymes

    Irregular

    1.1

    210

    51

    57

    2

    4

    2.1

    35

    7

    12

    2

    3

    2.2a

    46

    11

    10

    2

    3

    2.2b

    109

    31

    30

    4

    7

    4.1

    40

    5

    10

    6

    1

    4.2a

    29

    8

    4

    2

    1

    4.3a

    390

    76

    104

    10

    15

    5.1

    188

    48

    44

    10

    4

    5.2

    17

    5

    6

    2

    0

    5.3

    10

    0

    1

    4

    1

    5.4

    85

    10

    35

    4

    1

    1,159

    252

    313

    48

    40

    Percentages

    21.7

    27.0

    4.1

    3.5

    Middleton scenes

    Verse lines

    Feminine endings

    Run-on lines

    Rhymes

    Irregular

    1.2

    156

    24

    19

    32

    24

    3.1

    16

    4

    4

    2

    2

    3.2

    32

    7

    6

    4

    4

    3.3

    34

    7

    5

    6

    9

    3.4

    72

    7

    7

    8

    12

    3.5

    117

    20

    18

    30

    20

    3.6

    18

    2

    4

    4

    1

    4.2b

    22

    2

    2

    8

    6

    4.3b

    13

    0

    1

    0

    2

    480

    73

    66

    94

    80

    Percentages

    15.2

    13.8

    19.6

    16.7

    percent, ‘conies out as strongly un-Shakespearian and within the Middleton comparison sample range’. As Hope commented, the fact ‘that two scenes of more or less equal length should vary by 8 per cent is strongly suggestive of there-being two hands in the play’ (p. 101). The figures for auxiliary do in the remaining scenes, Hope reported, ‘provide broad support’ tor David Lake's division of authorship, with his claimed Middleton scenes having a regulation rate of 87 per cent, those by Shakespeare 83 per cent (pp. 101, 165). Hope commented that ‘a 4 per cent difference is not clinching evidence’, although (p.287) of course the lower than usual figure for Middleton could reflect the well-known fact that authors collaborating on a text tend to be influenced by each other. Still, Hope added, the evidence from 1.2 ‘remains strongly suggestive of there being a non-Shakespearian presence in the play—and there is nothing in the auxiliary do evidence which would rule Middleton out from being that presence …’ (p. 101).

    Hope also studied the use of relative markers in the play, applying the categories he had described earlier (pp. 27–53). Present-day Standard English uses who or whom in connection with personal antecedents, which with non-personal antecedents; Early Modern English did not distinguish the two.36 In the older usage, that could be used in non-restrictive relative clauses, and the zero relative could be used in the subject position, both usages that younger, upwardly mobile authors in Shakespeare's age would avoid (p. 28). Hope's detailed analysis clearly differentiated the normal usages of Shakespeare and Fletcher (pp. 34–40), but he recorded that ‘Middleton's proportions of usage of relative markers do not provide a basis for distinguishing him from other early modern writers’ (p. 49). However, when he came to study Timon of Athens Hope found that the relative marker evidence ‘does support the theory of collaboration’, and that it endorsed Lake's division of authorship (pp. 102–3, with tables 5.10 and 5.11). The ‘key feature’ to emerge from these figures ‘is the fact that the who/which distinction is almost fully observed in the Middleton scenes, but not in the Shakespearian—which is what the comparison samples would lead us to expect in a Shakespeare-Middleton collaboration’. The figures for Middleton were higher in one respect than those for his unaided plays, lower in another, both cases involving ‘a shift towards Shakespearian usage, [which] either indicates that Middleton is accommodating’ himself to his senior, or that some of the scenes that Lake (and others) have ascribed to Middleton are really by Shakespeare. Hope judged the ‘collaborative nature of Timon of Athens’ to be firmly established, but suggested the need for ‘further work to improve the precise division of the play …’ (p. 104).

    It is highly satisfactory that Hope's approach, through socio-historical linguistics, should confirm the findings of nine earlier scholars, using a very different series of approaches, who showed the presence of two distinct hands in the play. Yet, however great the volume of research suggesting that Shakespeare shared composition with co-authors, the Shakespeare conservators still manage to avoid the issue. That witty and erudite scholar, A. D. Nuttall, in a ‘New Critical Introduction’ to the play,37 followed H. J. Oliver in describing the variations in the spelling of Ventidius as forming the main basis for claims of ‘multiple authorship’ (Nuttall 1989, p. 31)—when of course this is a relatively minor sign of collaboration. Further, Nuttall conveniently (p.288) ignored the parallel case of the spelling of Apemantus, not citing the demonstration by Jackson (whose work he knew, describing him in a back-handed compliment as ‘the best of the disintegrators’; p. 37) that the variants correlate with the authorship division. Nuttall then listed the names of some scholars in the authorship debate, siding with Chambers and Ellis-Fermor, whose ‘unfinished play’ theory ‘is widely regarded as having finally routed the disintegrators’ (pp. 32–3). Like every other defender of Shakespeare's sole authorship, Nuttall deigned neither to summarize the case made for Middleton nor to quote any of the parallel passages, merely stating his belief that ‘the severest disintegrator cannot confidently banish Shakespeare from a single line or phrase’ (p. 34). That is a challenging claim, but of course it is wrong, since Shakespeare can be confidently freed of responsibility for several hundred lines of this play. Nuttall concluded that ‘The arguments for regarding Timon of Athens as a collaboration are too strong to be ignored’ (p. 39), but his perfunctory treatment of them did just that.

    The orthodox view continues to flourish, apparently immune to rational demonstration. Lois Potter, writing in 1992, could seriously cite J. C. Maxwell's 1957 edition as having ‘succinctly disposed of claims for co-authorship. Maxwell's view, we are told, ‘which is also the consensus of the last thirty years, is that the play is an unfinished, but wholly Shakespearean, draft. However, the two-author theory recently resurfaced in the Oxford Complete Works (1968)’ (sic!).38 The question many readers will be asking, confronted with Nuttall's belief that Ellis-Fermor's 1942 essay is ‘widely regarded as having finally routed the disintegrators’, or Potter's belief that ‘the consensus of the last thirty years’ plumps for sole authorship, is, in what circles do these critics move? How is it possible to ignore so much scholarly work which challenges your belief in sole authorship?

    The persistence of the Shakespeare conservatorsʼ denial of a scholarly tradition now dating back a century and a half is strikingly illustrated in the recent edition of Timon of Athens for the New Cambridge Shakespeare by the late Karl Klein.39 This gives a brief and wholly inadequate survey (Klein 2001, pp. 62–3) of the scholarly tradition from the mid-nineteenth to the late twentieth century. Klein made no attempt to give a fair summary of the many linguistic features which have been repeatedly shown to be untypical for (p.289) Shakespeare but characteristic of Middleton. He ignored the detailed work of William Wells and H. D. Sykes showing the presence of many Middleton usages in the sequence (2.2 to 3.3) dealing with Timon's abortive attempts to borrow money. He described Fleay (1874) and Wright (1910) as taking ‘great pains to differentiate in detail between Shakespeare's authentic passages and those which they thought could not be his’ (p. 62), without indicating the empirical evidence for their conclusion, other than that they judged some ‘couplets as well as irregular and unscannable lines … unworthy of Shakespeare’ (my italics). But their point, rather—following observations made by Charles Knight and Nikolaus Delius—was that such verbal features occurred far more frequently here than in any other Shakespeare texts of this period, and that they were in themselves distinctly different from Shakespeare's usage. Klein paid a little more attention to the work of Lake and Jackson, granting them ‘more rational’ and ‘more methodical procedures’, but he still swept aside the remarkable amount of detail they assembled, since after all, however ‘cogent … these patterned resemblances may be in detail, they include … only a small percentage of the possible elements of the text’ (p. 63). But no, the stylistic markers identified by these two scholars cover many aspects of both Shakespeare's and Middleton's style in verse and prose, a large percentage of the significant stylistic characteristics.

    Klein reserved most of his scorn for the discussion by Gary Taylor in his essay on ‘The Canon and Chronology of Shakespeare's Plays’ in the Oxford Textual Companion, taking exception to his general remarks on authorship criteria (TxC, pp. 77–80). Oddly, Klein ignored two further discussions there (mentioned above), Taylor's specific discussion of the Timon authorship issue (pp. 127–8), and the remarks of Stanley Wells and John Jowett prefixed to the Textual Notes on the play (pp. 501–2), both of which add detail, and give more (albeit incomplete) information concerning the scholarly literature on the authorship question. Klein quoted Taylor's statement that ‘“unusual mixtures of rhyme, irregular verse, and prose in the suspect scenes of Timon” substantiate a claim regarding authorship, since they “resist rationalization”’, but he complained that ‘the Oxford editors give no clue as to why these metrical anomalies, if such they be, should point towards Middleton …’ (Klein 2001, p. 64). However, if Klein had actually read all the secondary literature he cited, he would have seen that critics have time and again observed that the ‘aimless’ shifting back and forward between prose and verse is highly uncharacteristic of Shakespeare at any period, who regularly establishes clear-cut conventions for the two media, and sticks to them, but that it is found frequently in Middleton, together with the scattering of moralistic couplets ad libitum throughout the text. These observations were made, with an impressive agreement about the scenes where such a mélange occurs, by a whole series of critics: Delius (1867), Fleay (1874), Wright (1910), Sykes (1924), Chambers (1930), and Jackson (1979). As we have seen, later scholars have (p.290) evolved many new tests, including verbal contractions, function words, exclamations, speech headings, and several grammatical features. It is particularly disappointing that Professor Klein did not cite Jonathan Hope's ground-breaking study, published by Cambridge University Press in 1994.

    Having failed either to report or evaluate this scholarly tradition adequately, Klein endorsed the theory, which he believed was ‘first proposed by Wilhelm. Wendlandt’ in 1888, and given ‘its most powerful and persuasive presentation by Una Ellis-Fermor’ in 1942, that Timon is an unfinished play (pp. 65–6). However, this theory had been discussed by Nikolaus Delius in 1846, by Charles Knight in 1849, and Delius again in 1867, both of whom rejected it. Knight observed that ‘the differences of style … prevailing in the successive scenes of this drama’ indicated the presence of two hands, easily distinguishable by regularly recurring linguistic features, and that Timon contains ‘parts not only out of harmony with the drama as a whole … but altogether different from anything [Shakespeare] had himself produced in his early, his mature, or his later years’ (Knight 1849, pp. 70, 76). In 1846 Delius had already dismissed the ‘unfinished play’ argument by pointing to the ‘striking contrast in style, versification, and characterization that occurs throughout the whole drama’, with ‘the weaker passages’—or those standing ‘in glaring contradiction’ to the other style—being not limited to one part of the play but alternating throughout. Delius found it inconceivable that Shakespeare could have left the ‘finished’ and ‘unfinished’ parts so inextricably intertwined (Delius 1867, pp. 335–6). It is regrettable that Karl Klein, reviving the ‘unfinished play’ explanation, should have failed to address the cogent objections made by a German scholar one and a half centuries earlier. As MacDonald Jackson judged, this theory simply ‘fails to account for the evidence’ of two distinct styles in the play, one of which has all the characteristics of Middleton (1979, pp. 63, 89–90, 155–6).

    With Timon of Athens, as with the other plays studied here, the impressive feature of authorship studies from, the 1840s to the 1990s is that widely differing methodologies have converged, supporting sound attributions and discrediting others. The newer statistical methods build on, and should be taken in conjunction with, older approaches through verse styles, verbal collocations, and linguistic preferences.40 All these methods agree in assigning to Middleton a substantial part of Timon, and Shakespearians who continue to deny this point risk forfeiting their scholarly credibility.

    Notes:

    (1) CHS, ii. 95, 255; v. 141; vi. 361–5.

    (2) C. Knight, Studies of Shakspere (London, 1849).

    (3) N. Delius, ‘Ueber Shakespeare's Timon of Athens’, Shjb 2 (1867): 335–61. All translations mine.

    (4) N. Delius, Die Tieck'sche Shakespearekritik beleuchtet. Ein Supplement zu Shakespeare's dmtnati & chen Werken (Bonn, 1846; Hildesheim, 1981). For information on German Shakespeare criticism see the admirable annotated bibliography by Hansjurgen Blinn, Der deutsche Shakespeare (Berlin, 1993).

    (5) F. G. Fleay, ‘On the Authorship of Timon of Athens’, TNSS 1 (1874): 130–51; repr, in Fleay, Shakespeare Manual (London, 1876), pp. 187–208.

    (6) J. C. Maxwell (ed.), The Life of Timon of Athens (Cambridge, 1957), p. 87 n.; J. Q. Adams, ‘Timon of Athens and the Irregularities in the First Folio’,JEGP 7 (1908): 53–63.

    (7) The Revenger's Tragedy, ed. R. A. Foakes (London, 1966).

    (8) E. H. Wright, The Authorship of Timon of Athens (New York, 1910).

    (9) William Wells. ‘Timon of Athens’, NQ 12th ser. 6 (1920): 266–9.

    (10) Quotations have been cheeked against G. R. Price (ed.), ‘Michaelmas Term’ and ‘A Trick to Catch the Old One’ (The Hague, 1976). Cf. also Shortyard's assurance that Easye will ‘not want money’ since he feels bound ‘to see you furnisht’ (2.1.15–17).

    (11) B. Vickers, The Artistry of Shakespeare's Prose (London, 1968; repr. 1979), pp. 373–4.

    (12) The phrase recurs in a Dekker–Middleton play, The Roaring Girl (1608), this time with a bawdy meaning. Moll Cutpurse, dressed as a man, having drawn her sword against Sebastian, is handed a viol instead, with the words ‘Hold, there shall need no weapon at this meeting’, and replies: ‘it shall neʼer by said I came into a gentleman's chamber and let his instrument hang by the walls!’ (4.1.76–87). See the Revels edition by P. A. Mulholland (Manchester, 1987), pp. 10–11, for the ascription of this scene to Middleton.

    (13) H D. Sykes, ‘The Problem of Timon of Athens’, in Sidelights on Elizabethan Drama (Oxford, 1924), pp. 1–48 (originally as essays in NQ, 1921), at p. 1.

    (14) Namely: ‘… invite them without Knives:|Good for their meat, and safer for their lives’ (1.2.44–5); ‘Methinks false hearts should never have sound legs’ (233–4); ‘O that men's ears should be|To counsel deaf, but not to flattery!’ (249–50)

    (15) ‘Including one case of had (= he had)’. Cf. 3.2.34; 3.3.13, 23; 3.5.62; 4.3.450, 469.

    (16) E. K. Chambers, William Shakespeare: A Study of Facts and Problems, 2 vols. (Oxford, 1930), i. 481.

    (17) Una Ellis-Fermor, ‘Timon of Athens: An Unfinished Play’, in Ellis-Fermor, Shakespeare the Dramatist (London, 1961), pp. 158–76; originally in RES 18 (1942): 270–83.

    (18) D. J. Lake, The Canon of Thomas Middleton's Plays (Cambridge, 1975), pp. 279–80.

    (19) H J. Oliver (ed.), Timon of Athens, New Arden Shakespeare (London, 1959; rev. edn. 1963).

    (20) See the memoirs in NQ 221 (1976): 194–7 following his accidental death.

    (21) ‘But by the very thoroughness of his analysis, Wright demonstrated the weakness of his theory. A second writer had been called in to remove irregularities and inconsistencies in the play as Shakespeare left it, but all the irregularities and inconsistencies which Wright detected arose, according to him, precisely from the process of addition and revision’ (p. x).

    (22) Cf. pp. 121 (on an ellipsis at 2.2.18 and in Middleton), 127 (on 2.2.224, a similar usage in The Phoenix).

    (23) See e.g. p. 133, on 3.4.2–9, 13–17; and pp. 160–1, on 4.3.464, 471–4, 477, 512: twice here Maxwell adopts emendations from Pope, transposing prose into regular verse.

    (24) Oliver's lack of concern for authorship matters can be seen in his report that Charles Knight argued that Timon is ‘a reworking by another dramatist of a tragedy originally by Shakespeare’ (1959, p. xxii), when Knight argued the opposite case.

    (25) G. R. Hibbard (ed.), Timon of Athens (Harmondsworth, 1970).

    (26) M. P. Jackson, Studies in Attribution: Middleton and Shakespeare (Salzburg, 1979), p. 59.

    (27) Jackson's figures (hath 29, has 33) differ slightly from Lake's (has 31), and also from those given by F. O. Waller: see Jackson 1979, p. 65 n. 15, and p. 66 n. 31.

    (28) C. Hinman, The Printing and Proof-Reading of the First Folio of Shakespeare, 2 vols. (Oxford, 1963), ii. 280–5. Lake (1975, p. 286 n.) had also cited Hinman.

    (29) Cf. T. J. B. Spencer, ‘Shakespeare Learns the Value of Money: The Dramatist at Work on Timon of Athens’, ShS 6 (1953): 75–8. Spencer accepted the Chambers/Ellis-Fermor theory of the play being unfinished.

    (30) R. V. Holdsworth, ‘Middleton and Shakespeare: The Case for Middleton's Hand in Timon of Athens’, Ph.D. Diss. (University of Manchester, 1982). In the Textual Companion (1987) Holdsworth was credited with a ‘forthcoming’ book, Middleton and Shakespeare: The Case for Middleton's Hand inTimon of Athens (TxC, p. 128). I have twice attempted to obtain Dr Holdsworth's permission to read his thesis, but without success.

    (31) R. V. Holdsworth, ‘Middleton's Authorship of A Yorkshire Tragedy’, RES 45 (1994): 1–25, at p. 23.

    (32) W. M. J. Nowottny, ‘Acts IV and V of Timon of Athens’, ShQ 10 (1959): 493–7. This is basically a discussion of moral and Christian ‘themes’, ‘symbols’, and ‘world picture’ in the play, towards the end of which Nowottny declared her belief that the play was ‘the work of two hands’. She could find ‘no compelling case for supposing Shakespeare to have written any more of Acts 1 to 3 than the dialogue of Timon and Flavius in Act 2, scene 2’ and doubted whether Shakespeare contributed much after 4.3.464 (p. 497). These assertions showed no knowledge of authorship studies.

    (33) M. Tarlinskaja, Shakespeare's Verse: Iambic Pentameter and the Poet's Idiosyncrasies (New York, 1987), p. 104.

    (34) A. Oras, Pause Patterns in Elizabethan and Jacobean Drama: An Experiment in Prosody (Gainesville, Fla., 1960), pp. 68, 75, 38.

    (35) J. Hope, The Authorship of Shakespeare's Plays: A SocioLinguistic Study (Cambridge, 1994), PP. 100–4.

    (36) Cf. B. Vickers, ‘CounterfeitingShakespeare (Cambridge, 2002), ch. 4.

    (37) A. D. Nuttall, Timon of Athens (Hemel Hempstead, 1989).

    (38) A. Thompson et al., Which Shakespeare? A User's Guide to Editions (Milton Keynes and Philadelphia, 1992), pp. 164–5.

    (39) Timon of Athens, ed. Karl Klein (Cambridge, 2001). Professor Klein died in 1997, and his edition was completed by colleagues at the Universität des Saarlandes. According to the publisher's note on the half-title, he ‘establishes Timon as one of Shakespeare's late works, arguing, contrary to recent academic views, that evidence for other authors besides Shakespeare is inconclusive’ (my italics): but these ‘views’ are 160 years old. Klein's all-too-brief discussion of the play's date (p. 1) ignored the many reliable studies of Shakespeare's chronology, verse style, and vocabulary, concluding that ‘dating it will always remain a matter of conjecture’. He did, however, record the claim by F Brownlow that it was ‘written after 1614’: no evidence would support this late date.

    (40) M. W. A. Smith, in ‘The Authorship of Timon of Athens’, Text, 5 (1991): 195–240, applied purely statistical techniques to examining the case made by D. J. Lake and MacDonald Jackson for joint authorship. His methods (using a function word test, type-token ratios, and the first-word-of-speeches test) confirmed Shakespeare's part-authorship but left some doubts about Middleton's (pp. 202, 221–3). Since Smith limited his enquiry to statistical techniques, ignoring all the stylistic, linguistic, and lexical evidence for Middleton's authorship, we can conclude that stylometry on its own is not conclusive—as indeed Smith has frequently stated.