Previous articleNext article FreeWhose Hamlet Mocks the Warm Clown?John JowettJohn Jowett Search for more articles by this author Full TextPDF Add to favoritesDownload CitationTrack CitationsPermissionsReprints Share onFacebookTwitterLinked InRedditEmailQR Code SectionsMoreOne Suit of JestsOn 26 July 1602 the stationer James Roberts entered in the Stationers’ Register his title to publish Hamlet. In order to do so, he must have possessed a manuscript of the play by this date. The short First Quarto (Q1) appeared the following year, having been printed by Valentine Simmes for Nicholas Ling and John Trundle. The Second Quarto of 1604–5 (Q2), a much longer text, was printed by Roberts himself for Ling. There is no record of a transfer from Roberts to Ling, though an informal transfer would not have been an irregularity.1 For Shakespeare scholars, these are familiar and secure facts, islands of certainty amidst areas of textual criticism and book history given over to uncertainty and dispute. For many critics, Q1 is a debased text that has gone through some form of disruptive transmission such as memorial reconstruction by actors.2 Others question memorial reconstruction.3 Soon after Q1 was rediscovered, John Thomas Payne and Henry Foss, the antiquarian publishers who issued the first reprint in 1825, attached an “Advertisement” claiming that their book offered Hamlet “as originally written by Shakespeare.” This view has been supported on and off ever since, and in most detail by Terri Bourus, whose argument of 2014 that Q1 reproduces a version Shakespeare wrote in 1589 lends renewed urgency to this most fundamental question in Shakespeare textual criticism.4 It is beyond the scope of a single essay to navigate the entire question of Q1’s origins. I will here simply argue that one block of text that is unique to Q1 belongs to 1602 rather than the 1580s, and that it was not written by Shakespeare. This finding aligns with other arguments in favor of regarding Q1 as a disrupted text of the early 1600s rather than an originary text of the late 1580s.Any account of Q1 based on any theory involving textual disruption needs to negotiate the fact that the text contains writing with no equivalent in the longer versions. Yet there has been little enquiry as to whether this writing was originated by Shakespeare or someone else, or belongs to the late 1580s or the turn of the seventeenth century. The extent of text with no proper equivalent in Q2 and the 1623 Folio (F1) is limited. Just two passages offer a relatively substantial volume of self-contained writing that has little or no equivalent elsewhere and accordingly avoids the heavy intermixture of variant and invariant text in relation to Q2 and F1 that characterizes Q1 as a whole. One of these passages is the stylistically bland dialogue between the Queen and Horatio identified as scene 14 in Ann Thompson and Neil Taylor’s Arden edition.5 The other is a shorter self-contained passage, Hamlet’s description of a “warm clown” in his advice to the players (9.23–38).6 The latter will be the focus of the present paper.In Q1 alone Hamlet admonishes the players with the following criticism:7And then you haue some agen, that keepes one suteOf ieasts, as a man is knowne by one sute ofApparell, and Gentlemen quotes his ieasts downeIn their tables, before they come to the play, as thus:Cannot you stay till I eate my porrige? and, you owe meA quarters wages: and, my coate wants a cullison:And your beere is sowre: and, blabbering with his lips,And thus keeping in his cinkapase of ieasts,When, God knows, the warme Clowne cannot make a iestVnlesse by chance, as the blinde man catcheth a hare:Maisters tell him of it.(F2r–v)Hamlet here describes a feeble stage clown who is restricted to a five-card trick of jokes. His repartee is so well known that gentlemen in the audience have already jotted them down in their writing tables, presumably from earlier theatre visits.8 Who wrote Hamlet’s lines about this clown, and when were they written?Enter HeywoodAn initial test of n-grams using the drama section of the online database Literature Online (see below) placed Shakespeare third out of thirteen dramatists who wrote a substantial number of mainly solo-authored plays and who were working in the period 1587–1603. This is an immediately significant result. The dramatists who scored marginally higher than Shakespeare were Thomas Heywood and Ben Jonson, who both totaled thirty-four hits (tokens) as against Shakespeare’s thirty-two. But the results look as though they are extremely sensitive to canon size, and the Shakespeare dramatic canon significantly exceeds those of all other playwrights. The results were particularly erratic for smaller canons. Arithmetical adjustment of the figures to offset canon size can only exaggerate this unreliability. This preliminary test suggests that a moderately high score is needed for any test along these lines to begin to function. Although several other dramatists including George Chapman and Thomas Dekker would score more highly than Shakespeare if the numbers were adjusted to reflect canon size, attention must be confined to Heywood and Jonson. Jonson is certainly a more plausible author of the passage than Shakespeare, and in many respects an attractive contender. To give just one example, he writes the 2-gram “my porridge” collocated with references to “your eating player” and “apparel.” His use of the word “cullison” will be considered further below. But when Jonson is assessed against Shakespeare in the way described for Heywood, the number of n-grams exclusive to Shakespeare proves to be significantly higher than with Heywood. It is therefore Heywood who is examined in more detail below.Obviously the combination of strong results and the convenient testability of Heywood as the author of a substantial canon offers no guarantee that a less prolific writer did not contribute the passage. However, there are a number of factors relating to biography, style, and ideation that make Heywood a plausible contender. In the weeks between 24 November 1602 and 7 January 1603, Heywood, at this time a leading member of Worcester’s Men, was writing a play called The Blind Eats Many a Fly, the title being a proverbial tag very similar to “as the blind man catcheth a hare.”9 Heywood was responsible by his own account for “an entire hand, or at least a maine finger” in 220 plays.10 Most important, he was a specialist in writing dialogue for clowns. A short sample from a speech of the Clown Betts in Heywood’s revision of Sir Thomas More, probably executed in about 1603–4, shows language similar to that in Q1’s description of the warm clown:11 “Come, come, we’ll tickle their turnips, we’ll butter their boxes! Shall strangers rule the roast? Yes, but we’ll baste the roast. Come, come, aflaunt, aflaunt!”12 Here too we see the expression of resentment and social conflict in terms of symptomatic foodstuffs, and self-expression by way of repeating quasi-proverbial phrases. The traits of referencing homely food and quoting proverbs are thoroughly typical of Heywood’s clowns. So are references to parts of the face—“Charge me to the mouth,”13 “I will set out a throate,”14 “from lip to lip, and from nose to nose”15—and to apparel: “and his suite is the old suite still, and his cloaths the old cloaths.”16 So the verbal register fits well. Heywood also wrote against the “dull” clown:This hobinall, this rusticke, this base clowne;I finde him of a dull and brain-lesse eye,Such as I know a golden-headed shaftWill never enter; of a skin so thicke,As pointed silver hath no power to pierce:17He refers to surly clowns—“’Tis hard to teach them manners that are Clownes”18—and to a clown being paid “wages”—“Madam I haue a yeeres wages before hand.”19 His clown also might drink beer: “Why, ale is a kind of juice made of the precious grain called Malt, and what is malt? Malt’s M, A, L, T, and what is M, A, L, T? M much, A ale, L little, T thrist, that is, much ale, little thrist”;20 “that should giue mee a Pot of Ale, that should drinke to me, and drinke vp all.”21 Heywood associates a clown specifically with writing tables: “Enter the Clowne with his Table-bookes.”22 Indeed, the title page engraving to A Maidenhead Well Lost illustrates figures from the play including the Clown actually carrying writing tables in his hand.23 And it was Heywood who famously complained when one of his own plays was, as he says, published in a bad text taken down in the theatre by the shorthand technique of stenography:… but ill nurst,And yet receiv’d, as well perform’d at first,Grac’t and frequented, for the cradle age,Did throng the Seates, the Boxes, and the StageSo much; that some by Stenography drewThe plot: put it in print: (scarce one word trew:)And in that lamenesse it hath limp’t so long,The Author now to vindicate that wrongHath tooke the paines, upright upon its feeteTo teach it walke, so please you sit, and see’t.24Heywood was therefore prolific as bit-writer, specialist in clown roles, and commentator on the uses of both writing tables and stenography in the theatre. Nothing is yet proved by this evidence, but even without detailed reference to quantifiable links Heywood begins to look more plausible as author than Shakespeare, and there are grounds to justify further enquiry. Given that mode of writing in Hamlet’s words about the warm clown is deficient in Shakespearian phrasing, might they be positively Heywoodian?N-Gram Test ProcedureWith this question in mind, I searched Early English Books Online (EEBO) for words and fixed sequences of words—n-grams—in the passage, identifying those that are found either in the works of Heywood but not Shakespeare, or in the works of Shakespeare but not Heywood.25 A crucial and generally well-recognized issue for the methodology mentioned already is that the corpora of the two writers are not of equivalent scope and size. The Shakespeare canon is far more predominantly in the form of dramatic dialogue. Where the Shakespeare canon is dominated by plays, the Heywood canon consists of a high proportion of prose works. Preliminary searches indicated that the corpus size has a significantly distorting impact on the results. Heywood is favored, because, once his prose and non-dramatic verse are added to the plays on which the initial scoping test was based, the canon of his works is significantly larger than Shakespeare’s. The corpora were therefore equalized by restricting the Heywood canon to his dramatic works, supplemented by other works to make up a corpus roughly equivalent to works and parts of works of Shakespeare’s uncontested authorship.26 These works, which include translations, were selected by date of publication, favoring dates up to and including 1625 that are closer to the composition of the warm clown passage, with the exception of Gynaikeion (1624) on account of its extraordinary length.27The adopted procedure responds to the fact that the speech amounts to no more than 113 words. It is short enough to check every sequence of words it offers. I have nevertheless assumed that anything less than a clear-cut differentiation would not be significant. Lower summative numbers that produce positive differentiations are preferred over higher numbers capturing examples of imprecise, malleable, or obfuscating data. Consequently, all cases where both authors use a particular sequence of words but there is a preponderance of one over the other were rejected. This followed from trial searches of other passages that suggested that such evidence produced an unacceptable number of false positives. Only words and sequences of words unique to one of the writers were therefore accepted. This approach has two benefits: that the evidence is controlled as to quality, and that the findings are reasonably commensurate with each other, as befits any analysis that depends on counting items as though they were equal.In order to minimize the possibility of accidental manipulation of the results, the test was based on strict correspondences, to the exclusion of morphological variants and analogues depending on ellipsis, transposition, or substitution. For the most part, EEBO’s application of spelling variants, which is ultimately based on the OED’s documentation, was followed without modification. The results were thinned by rejecting any hit that occured in a passage of a collaborative work attributed to another writer, or any recurrence of a hit in a reprint or variant version. Like other examples of this kind of investigation, the test captures some seemingly insignificant sequences of words from a semantic point of view, but such cases have to be accepted on the unqualifiable basis that they are simply there as data; it may be added that what is seemingly insignificant may have a significance that is not obvious purely by intuition, and that such verbal combinations often constitute a part of a larger and more meaningful sequence. To illustrate the latter point, in the control test of a passage from Heywood discussed below we find two Heywood parallels to the 3-gram “his people to.” This is a meaninglessly sub-phrase in its own right, but the contexts establish a phrasal parallel that is disguised by the clipping that produces the n-gram. The control passage has the phrase “he hath taught his people to sow”; this compares with similar passages elsewhere in Heywood: “causing his people to land again,” “he commandeth his people to stay his return,” and “encourageth his people to carry a valiant conceit.” The manifest shared but sub-phrasal element is a symptom of a shared semantic structure.For the purpose of counting, special attention was paid to additional n-grams falling within the same longer n-gram, as with “a great way off,” “great way off,” and “way off.” These are formally separate n-grams, but they do not have the same significance as tokens or types occurring in different locations. Therefore the longest n-gram is counted and the additional ones considered separately. In the lists of n-grams these additional instances are placed in square brackets and not included in the n-gram count. In the summative tables they are counted in a separate column headed “n(+n)-gram.”This test, with its dependence on n-grams, loses potential evidence, and it lacks any attention to ideation. Looser parallels that depend on the manipulation afforded by allowing morphological variables, ellipsis, transposition, or synonym substitution are therefore considered separately. Though such items are sometimes quantified in other studies, here they are not, because their quantitative significance is insecure. But they undoubtedly have qualitative significance, and indeed they can constitute a rich network of association.Control TestsThese procedures were tried out on a passage that has not been questioned as to its Shakespearian authorship, and another passage that has not been questioned as to its Heywoodian authorship. For the Shakespeare control test I sought a passage as similar to the clown passage as possible: the same speaker in the same scene of the same play, the same use of prose,28 and same absence of interlocution. Thus I selected the first 113 words of the opening of the scene, as printed in the quarto of 1604–5, but here modernized in spelling:Speak the speech, I pray you, as I pronounced it to you, trippingly on the tongue. But if you mouth it as many of your players do, I had as lief the town crier spoke my lines. Nor do not saw the air too much with your hand, thus. But use all gently. For in the very torrent, tempest, and, as I may say, whirlwind of your passion, you must acquire and beget a temperance that may give it smoothness. O, it offends me to the soul to hear a robustious periwig-pated fellow tear a passion to tatters, to very rags, to split the ears of the groundlings, who for the most …The words and phrases unique to one writer or the other, separated by n-gram, are as follows:29Shakespeare Control Passage In Shakespeare but not Heywood:1-gram 3-gram trippingly1it to you7robustious1it as many1pated5nor do not12-gram hand thus but1and beget1soul to hear1tear a2 to very2 split the1 In Heywood but not Shakespeare:1-gram 3-gram tatters1may give it12-gram the soul to1to split4who for the2it offends24-gram and as I may1 View Table Image In this passage, the n-grams types (the different phrases as listed, excluding variant forms) favor Shakespeare by seven in Heywood to twelve in Shakespeare. The tokens (the individual occurrences of the phrases) produce a clearer differentiation of twelve in Heywood to twenty-four in Shakespeare. The Shakespearian tokens cluster in two types, “it to you” and “to split.”The Heywood control passage was selected from a small menu of uninterrupted speeches spoken by a clown. The first 113 words, thus the sample to be tested, here modernized from the quarto text, are:Let his virtues speak for himself: he hath taught his people to sow, to plough, to reap corn, and to scorn acorns with their heels, to bake and to brew; we that were wont to drink nothing but water have the bravest liquor at court as passeth. Besides, he hath devised a strange engine, called a bow and arrow, that a man may hold in hand and kill a wild beast a great way off, and never come in danger of his clutches. I’ll tell you a strange thing, nurse: last time the King went a-hunting, he killed a bear, brought him home to be baked and eaten. A gentlewoman of the Court …30For this passage, the words and phrases unique to one writer or the other are as follows:Heywood Control Passage In Heywood but not Shakespeare:1-gram bake and brew1acorns1we that were1a-hunting2a bow and12-gram [great way off (+1)] bake and1have the bravest1and brew1danger of his1but water1[you a strange (+2)] wild beast1time the king1[way off (+2)] brought him home13-gram gentlewoman of the3for himself he1for himself he1hath taught his14-gram his people to2a great way off1to sow to3[tell you a strange (+1)] to plough to15-gram corn and to1I’ll tell you a strange1 In Shakespeare but not Heywood:2-gram 3-gram last time1that were wont1arrow that1he hath devised1beast a1hand and kill1last time4never come in5[be baked (+1)] to be baked1and eaten14-gram that a man may2 View Table Image The test yields twenty-three types and twenty-nine tokens for Heywood, as against eleven types and nineteen tokens for Shakespeare (see Table 1). The n-grams, restricted though they are by reduction of the Heywood canon and insistence on recording only items unique to one author, yield a significant discrimination.Table 1. Counts and Percentages of N-Grams in Control Passages Count of Types Heywood: ShakespeareCount of Tokens Heywood: ShakespearePercentage of Types Linked to HeywoodPercentage of Tokens Linked to Heywoodn(+n)-gramShakespeare Control7:1212:2436.8%33.3%0:0Heywood Control23:1129:1967.6%60.4%6:1 View Table Image When Gary Taylor ran a comparable control test to investigate passages of Shakespeare and Middleton, the text produced the anticipated result for the Shakespeare passage but failed to identify the Middleton passage as Middleton.31 There are inherent limits to the testing that can be employed on a short passage, but Taylor was able nevertheless to persevere with his test to identify the hand of Middleton in a short passage of Macbeth as a result of the differential between the Middleton control results and those for Shakespeare. With the present control tests Heywood is at least as strongly identified as Shakespeare. To be sure, an adjustment of the test as regards the limitations imposed on the Heywood corpus or the searching techniques could lead to a noticeable adjustment to the aggregated figures. But any credible adjustment is unlikely to cancel the strong differentiations seen in the present test.The Warm Clown Passage TestedWe can now ask whether the diction of the warm clown passage from Q1 Hamlet is intrinsically more like Shakespeare or Heywood, knowing the efficacy of the text, and also see whether the degree of separation compares with that found in either of the control passages. The findings reported in the following discussion are based on the same procedure as that outlined for the control passage, except that the results were refined through an additional check in Literature Online, which caused two potential Heywood items to be discounted. As will be seen, more of the types are exampled in Heywood but not Shakespeare than in Shakespeare but not Heywood.Q1 Warm Clown Passage In Heywood but not Shakespeare:1-gram 3-gram catcheth321before they come1 (+2)2-gram unless by chance1some again14-gram one suit1before they come to1 (+1)33a quarter’s1they come to the1quarter’s wages25-gram wages and1before they come to the2your beer1 In Shakespeare but not Heywood:1-gram 4-gram cinquepace343as the blind man12-gram is sour1 3-gram cannot you stay351 to the play1 [as the blind (+1)] [the blind man (+1)] him of it1 View Table Image The examples amount to twelve Heywood-linked types as against six Shakespeare-linked types. This is closely in line with the Heywood control test. The eight tokens relating to Shakespeare as against fourteen relating to Heywood offer no contradiction. From a statistical point of view, the positives in favor of Shakespeare need no special explanation, as they are significantly smaller in volume than the false positives found in the control test of a passage from Heywood and are to be expected. Table 2 summarizes the counted ratios and proportions.Table 2. Counts and Percentages of N-Grams in Control Passages and the Warm Clown Passage Count of Types Heywood: ShakespeareCount of Tokens Heywood: ShakespearePercentage of Types Linked to HeywoodPercentage of Tokens Linked to Heywoodn(+n)-gramShakespeare Control7:1212:2436.8%33.3%0:0Heywood Control23:1129:1967.6%60.4%6:1Q1 Warm Clown12:614:866.7%63.6%3:2 View Table Image Clearly there is no resemblance between the percentages of Heywood markers in the Shakespeare control passage and those in the Q1 passage: the links to Heywood are by percentage over twice as high as in the Shakespeare control passage. When the Heywood control is compared, the volume of authorial positives is lower for both writers, but the percentages are far closer to the Heywood control: still higher for types, but lower for tokens.As for the separated n(+n)-grams, the control tests suggest that they are more typical of Heywood than of Shakespeare, and this may be a broader indication of a stylistic difference between the two writers. The score 3:2 in the warm clown passage is based on low numbers but maintains the pattern favoring Heywood. But the two Shakespeare examples come within the one acknowledged proverb in the passage (see below), and therefore from a lexical sequence that is not Shakespearian. If these items, despite their borderline status, are nevertheless counted in with the types, the percentage links to Heywood become: Shakespeare control 36.8%, Heywood control 70.7%, Warm Clown 65.2%. These figures reconfirm the passage’s affinity with Heywood. Thus a rather striking example of the malleability of this kind of testing fails to counter-indicate the conclusion. By any criteria based on this test, Heywood is the favored author.Further EvidenceAs has been suggested, the restriction of the above test to true n-grams without morphological variables and the limitation of the Heywood canon involve a loss of evidence, in the name of generating controlled and comparable data. If we move from this controlled kind of counting to a wider compilation of parallels, the case for Heywood against Shakespeare becomes far stronger. Further, it becomes possible to see how lexical data can be re-envisaged as markers of style and technique that relate an authorial persona.Further Heywood Parallels in the Full Heywood Corpus but not in Shakespearesome again that1keeping in1they come to the+1is known by1his lips and1man is known by1unless by chance+1catcheth+2Heywood Parallel for a Phrase in Shakespeare but not in the Restricted Heywood Corpusto the play Variant Forms: HeywoodAnd then you have some again that: the further instance of “some again that” noted above, in The Hierarchy of the Blessed Angels (1635) occurs in the phrase “and some again that,” with the same clausal structure as in Q1. There are two further instances of “some again” in non-corpus Heywood, making four in all.as a man is (as a man was)beer is (beer being)blabbering (blabber)masters tell him (master tell him, master telling him)Then you have some: then you have one (spoken by a clown).gentlemen “quotes … thus:” Heywood refers to “Censurers, such as continually carpe at other mens labours, and superficially perusing them, with a kind of negligence and skorne, quote them by the way, Thus.”36 The collocation “quote(s) … thus” followed by a series of examples is all the more striking as Shakespeare never uses “quote” to mean “note, jot, write.”suit of jests: Heywood but not Shakespeare has “[n.] of jests,” is a passage giving a similar character sketch: “Adue fond humorist, Parenthesis of iests, / Whose humour like a needlesse Cypher fils a roome.”37in their tables: compare “his name he in his Tables hath inscrib’d,”38 referring again to writing tables. Shakespeare has no example of “in [possessive pronoun] tables.”“your beer” near “[pronoun] lips.”catcheth a hare: Compare “catcheth the hare,” as an example of “active nimbleness.”and thus keeping in his: varied as “and thus speaking in his.” EEBO identifies no other example of “and thus fby.3 [followed within three words by] in his” in the work of any dramatist in the period.Variant forms: Shakespearethey come to the play: they are coming to the playLess Strict Parallels: Shakespearemy porridge: Shakespeare but not Heywood has three examples of “[possessive pronoun] porridge.”Supporting evidence in favor of Shakespeare is conspicuously thin. Under the variant forms listed above, one phrase is unexampled in other dramatists: “they come to the play.” But it is compromised as a Shakespeare marker, because the Shakespeare parallel occurs in Hamlet itself.39 In other words, it might be a displacement from one part of the text to another; it can therefore be explained in two conflicting ways, authorship or disrupted transmission, and offers no differentiation between them. Q1 Hamlet has various other examples, such as the line in the closet scene, “To liue in the incestuous pleasure of his bed,” which echoes Hamlet’s “Or in th’incestious pleasure of his bed” in the Q2 version of the preceding prayer scene. Indeed, in the case of “they come to the play,” the degree of dramatic separation is smaller, as the Q2 parallel comes from the very same scene. Parallels that are internal to the texts of Hamlet itself are thus of no discriminatory value.Author Versus DateTaken as a whole, the evidence for Heywood presented here heavily outweighs that for Shakespeare. Yet two items stand out as exceptions to this overall pattern. They are, potentially at least, of too much weight to be discounted as random false positives. I will argue that they run against the manifest direction of the other evidence for reasons that relate to the date at which the Q1 text was prepared.The clown’s use of “cinquepace” looks like an exceptional mark of Shakespeare’s authorship and would be crucial to any defense of the passage as Shakespearian in its stylistic quality. Of the locutions favored by one dramatist over the other in the passage, it is among the rarest, as can be seen in Table 3, ranking the phrases unique to one author or the other by order of rarity in early modern English over a century from 1550 to 1650.Table 3. Markers of Authorship by Rarity in Early Modern English 1550–1650 HeywoodShakespeareElsewhere1unless by chance20+15402and some again that10+203your beer10+27414cinquepace4203+335a quarter’s20+366some again that10+57 View Table Image These are all decidedly low counts, considering that they are based on searches of over 50,000 texts. Five of the six rarest items point to Heywood. “Unless by chance,” “cinquepace,” and “a quarter’s” are particularly significant for the additional reason that they occur more than once in the work of the corresponding author. This is all the more reason to dwell on “cinquepace,” which runs against the pattern otherwise seen in Table 3. The figure for “cinquepace” excludes examples in Middleton written before but published after 1642. The dancing-master Cinquepace (spelt “Sinquapace”) is named in full eight times in No Wit/Help Like a Woman’s (performed 1611, published 1657), and also, in abbreviated form (“Sinq.”), twenty-nine times in speech prefixes. If this text had been taken into account, “cinquepace” would not look so unusual.Whether or not one sets aside this anomaly in the figures, the word’s weight as an authorial marker demands special consideration, as it is the only 1-gram in the vocabulary unique to Shakespeare. Indeed it is the only 1-gram exclusive to either writer if “catcheth” is excluded as a mere inflection. What does its 1-gram status indicate? As a new item in any author’s usage, “cinquepace” would be an established but unusual and distinctively foreign word belonging to a restricted area of diction relating to the art of dancing. As such, it is a consciously deliberated and knowing choice, and has the inevitable character of a borrowing, a word that has come to an author from a different author’s lexis. For instanc