We have taken the liberty of reproducing large chunks of Diana Price's article rebutting the content of our Hand D pages (which she does without referring to them or linking to them). You can download the whole article here. Obviously we cannot reproduce and annotate the whole article, so  only the parts on which we have something to say are reproduced here. This isn't optimal, but we want to remain within the definition of Fair Use. And we have quite a lot to say.

As always, allow time for the annotations to appear and see all of them by clicking on the icon in the top right of your browser window.

Complete essay Hand D and Shakespeare’s Unorthodox Literary Paper Trail

Click thumbnails to the different hands in high resolution


The biography of William Shakespeare exerts an influence on various areas of research related to Shakespeare, including textual, bibliographical, and attribution studies. A case in point is the theory that Shakespeare wrote the Hand D Additions in the Sir Thomas More manuscript. That theory is now part of received scholarship, even though many of the assumptions and arguments first published in 1923 have been challenged …  Supporting arguments for the Hand D attribution, in particular those based on orthography, prove vulnerable to challenge.

Keywords: Forensic, ‘Foul Papers’, Hand D, Handwriting Shakespeare, Sir Thomas More

1. Introduction

…  have been designated as Hands A, B, C, D, E, S, and that of the Master of the Revels Edmund Tylney. It is Hand D that is of interest since many, perhaps most, biographers and editors today accept it as Shakespeare’s. This claim has never been front page news. Instead, it has been gradually advanced since 1923, threading its way into the fabric of Shakespearean biography, editions, and studies.

Without the three pages written by Hand D, Shakespeare’s biographical documentation does not include any literary paper trails; that is, he left behind no hard evidence during his lifetime that could support the statement that his occupation was writing. In an exchange concerning his review of my Shakespeare’s Unorthodox Biography (2001), Prof. Stanley Wells acknowledges that the first piece of evidence identifying the man from Stratford as a writer was indeed posthumous. In other words, Shakespeare is the only alleged writer from the time period for whom one must rely on posthumous evidence to support his professional activities as a writer.

In the early 1920s, Alfred W. Pollard recruited a group of scholars to contribute essays identifying Hand D as Shakespeare’s. The collection was published in 1923 under the title Shakespeare’s Hand in the Play of ‘Sir Thomas More’…

… Pollard was attempting to fill the documentary void and put an end to the authorship question. In the early part of the twentieth century, the controversy was gaining momentum. Anti-Stratfordian challenges were coming from J. Thomas Looney and Sir George Greenwood in England, and Mark Twain was popularizing the case in the United States. In his preface, Pollard explained that if it is proved that Shakespeare wrote the Hand D portion of Sir Thomas More, then the theories proposing Oxford, Derby, or Bacon as the author come ‘crashing to the ground’ (1923a, v). There’s his agenda, but the subtext is just as significant. If Pollard thought that Hand D could settle the authorship question once and for all, then he was acknowledging that Shakespeare left behind no evidence during his lifetime that proves he was a writer by profession. Otherwise, Pollard would not have needed Hand D to settle the debate.

These authorship-driven pressures continue today. Hugh Craig describes it:

In many respects attribution …

Hand D passages provide for once a link between ‘Shakespeare’ texts and William Shakespeare of Stratford. (2012, 17)

That ‘link’ is the putative literary paper trail, Shakespeare’s handwritten manuscript, that proves he was a writer. A …

…manuscripts, letters, and inscriptions, as well as signatures. Thompson is trying to lower his readers’ expectations concerning Shakespeare’s literary remains, the evidence that I refer to as literary paper trails. …

…The deficiency is unique to Shakespeare’s literary biography. That deficiency brings the traditional attribution into question, which is why, in his own words, Pollard led the charge to establish Hand D as Shakespeare’s.


Since 1923, the claims that the Hand D Additions were composed by Shakespeare and are in his own handwriting have accelerated. The attribution is now accepted in all the major collected works, many critical editions, Shakespearean biographies, and related scholarship. The claim has been repeated so often that many scholars who might have questioned the original arguments, particularly the handwriting case, have instead accepted it as fact.

The primary argument for identifying Hand D as Shakespeare’s is Sir Edward Maunde Thompson’s case based on palaeography. Thompson compares D’s handwriting with the extant samples of Shakespeare’s penmanship, the six signatures. Thompson was the first Director of the British Museum and a preeminent palaeographer of his time. Harmopn and Holman define palaeography as ‘the study of old forms of handwriting, important to textual studies for establishing texts and deciding authorship’ (1992, 340). Other resources describe palaeography as concerned with ‘ancient’ forms of handwriting. The English secretary hand is certainly an ‘old’ form of handwriting no longer in everyday use, although it is not generally characterized as ‘ancient,’ as are hieroglyphics or Tibetan scripts. However, as I explored Thompson’s palaeographic case, I began to learn about a newer discipline: forensic document examination.

The relevance of the forensic document examiner’s (FDE’s) methods to Thompson’s case for Shakespeare’s handwriting first becomes apparent in the footnotes of Samuel A. Tannenbaum. Tannenbaum was one of the first to challenge Thompson’s palaeographic case, and he cites Albert S. Osborn’s Questioned Documents as, for one example, the ‘authoritative work on the subject’ (1925, 135n.; see also 1927, 8n.).

Osborn’s criteria include those for establishing a control sample (also termed controls for comparison, exemplars, or standards): ‘the best standards of comparison are those of the same general class as the questioned writing and as nearly as possible of the same date. Such standards should, as a rule, include all between certain dates covering a period of time both before and after the date of the writing in dispute’ (1910, 18-19; see also Matley 1990, §5, 17). Osborn’s ‘general class’ rule ensures that the writings being compared belong to the same species; that is, signature to signature, dramatic manuscript to dramatic manuscript, and so on. (Unfortunately, Tannenbaum did not always apply Osborn’s methods or rules, and his 1925 article on Hand D is predicated on his admittedly qualified acceptance of the six signatures as a suitable control sample for his critique of Thompsons’s analysis).

Forensic document examination as a discipline began to emerge in the late 1800s. In 1894 William E. Hagan published Disputed Handwriting, and in 1901 Persifor Frazer published Bibliotics or the Study of Documents. Osborn’s 1910 Questioned Documents continues to be quoted today as a founding text by FDE resources, both in print and online. By definition, the forensic document examiner is concerned with handwriting from the standpoint of providing testimony and evidence in a court of law, but scientific methods are common to both FDEs and palaeographers, involving, as they do, criteria and procedures that can be tested and replicated by others.1

Tannenbaum did not identify himself as a palaeographer. Instead, he adopted Frazer’s term of ‘bibliotics’, considering himself a practitioner of the science of the ‘study of documents and the determination of the individual character of handwriting’ (1925, 135). The term ‘bibliotics’ never really caught on; it is not found in the OED, although in a posting to the online Merriam-Webster Dictionary, William Sutton cites both Osborn and Frazer in the entry for ‘bibliotics’. The word seems to have served more as a bridging term, overtaken by ‘forensic document examinations’ and ‘examiners’ (sometimes termed Questioned Document Examiners or QDEs) whose terms and methods were subsequently adapted to and incorporated into research and analysis in literary, historical, and other disciplines unrelated to legal cases.

The standards set by examiners are conservative because someone’s innocence or guilt hangs in the balance. Of course, the possibility that Shakespeare did

1 On reproducible tests, procedures, and conclusions, see Huber and Headrick 1999, 261.

not himself pen the Hand D Additions raises no ethical or criminal issues. Nevertheless, legal standards of proof are relevant to several scholars with respect to Hand D. In 2013, Douglas Bruster proposed that Shakespeare wrote revisions for the 1601 quarto of The Spanish Tragedy. His argument inferring manuscript idiosyncrasies is based on the assumption that Shakespeare wrote the Hand D Additions. In his essay on ‘Authorship’, Hugh Craig mentions both theories:

There is reason to believe there are two surviving plays to which Shakespeare added passages some time after their original performance: The Spanish Tragedy, more speculatively, and Sir Thomas More, now beyond reasonable doubt. (2012, 23)

The phrase ‘beyond reasonable doubt’ implies that the case for Hand D as Shakespeare’s could meet the standard of evidence required to obtain a conviction. At some level, Gary Taylor is aware of the weakness of the case, admitting that the case for Hand D as Shakespeare’s might not hold up in criminal court (1989, 102). He is preceded by Pollard:

If we think of the use which might be made of Sir E. M. Thompson’s arguments in a trial at law it is obvious that they are much more valuable for defence than for attack. Let it be granted that if an estate were being claimed on the evidence adduced to show that the two hands are identical, a jury would probably refuse to award it. (1923b, 13-14)

In modern times, cases involving questioned handwriting are likely to relate to fraud, forgery, and other crimes. High profile cases include the trial of Bruno Hauptmann and the Lindbergh kidnapping ransom notes (Osborn was one of the expert witnesses), the Hitler Diaries hoax, and Clifford Irving’s forgery of Howard Hughes’ signature on publishing contracts for his ‘autobiography’.

However, during the first half of the twentieth century, palaeographic studies began to incorporate the FDE’s and ‘Questioned Documents’ terminology and techniques. An overview of this cross-pollination is provided by Jeffrey Abt:

Although the forensic scientists laid the groundwork for scientific investigations into manuscripts and books and the effective documentation of their findings, general knowledge of this work remained confined to legal circles. The first to synthesize this body of research and, along with studies in other fields, apply it to historical questions raised by library materials was Reginald B. Haselden (b. 1881), then curator of manuscripts at the Huntington Library. In the preface to his seminal Scientific Aids for the Study of Manuscripts (1935), Haselden remarks: ‘In recent years scientific knowledge has extended its sphere of usefulness to almost all fields of endeavor. The question is whether this knowledge can be utilized and brought to bear on the complex problems encountered by the paleographer and the student of literary and historical manuscripts’. (1987, 29)

Palaeography and bibliotics are two of seven distinct ‘auxiliary sciences’ that Haselden incorporated into his study of manuscripts (Haselden 1935, 1-4).2

In the ensuing decades, scholars studying various types of historical manuscripts integrated into their ‘new palaeographic approach’ the basic practices of ‘forensic handwriting analysis and how [they] may be applied outside the courtroom’ (Dalton, et al. 2007; see also Stokes 2007-2008). Yet most of the Hand D literature continues to refer only to the early palaeographic case, without comparing Thompson’s methods and standards to those of the FDE.

Fig. 1 – Signature n. 1 on the Mountjoy affidavit (1612)

Fig. 2 – Signature n. 2 on the Blackfriars Gatehouse purchase deed (1613)

Fig. 3 – Signature n. 3 on the Blackfriars Gatehouse mortgage (1613)

2 Haselden includes R.B. McKerrow in his acknowledgements (1935, x).

pFig. 4 – Signature n. 4 on page 1 of Shakespeare’s Last Will (1616)


Fig. 5 – Signature n. 5 on page 2 of Shakespeare’s Last Will (1616)

Fig. 6 – Signature n. 6 on page 3 of Shakespeare’s Last Will (1616)

At the beginning of his 1923 essay, Thompson identifies some obvious problems in conducting a palaeographic analysis, including the paucity of specimens available for comparison, that is, the control sample; the degrees to which the signatures vary in formations and method of writing; the interval of up to twenty or more years separating the penning of Hand D and the signatures; and the supposed illness of the testator affecting at least three of the signatures (71-72). Thompson references his earlier essay in Shakespeare’s England in which he further explains why the three signatures on the will comprise unlikely exemplars:

The three subscriptions present great difficulties which are almost beyond explanation. In the first place, they differ from one another to such a degree that it is not going too far to declare that, were they met with on three independent documents, they might not unreasonably be taken, at first sight, for the signatures of three different persons. (1962, 1: 304; see also 1916a, 12)

Signature n. 4 is so degraded as to be useless, so scholars have long relied on George Steevens’s facsimile of 1776, despite Thompson’s misgivings about other facsimiles ‘engraved from drawings by Steevens’ (1916a, x). Elimination of n. 4 reduces the control sample to five signatures.

In his book on Shakespeare’s Handwriting, Thompson further reduces the control sample to fewer than three:

Practically these two signatures [#1 and #2] are the only specimens from among the six which afford sufficient data for forming an opinion on the character of Shakespeare’s handwriting. The third signature … is too formal to serve as a criterion … The three signatures to the will are likewise of little value for general comparison, with the exception of the first three words [of #6, ‘By me William’]. (1916a, 28)

Because the words ‘By me William’ are noticeably better formed than the surname, Greenwood reasonably posits that a scrivener wrote them and that Shakespeare wrote only his last name (1920, 32). According to the author of Forensic Handwriting Identification, the specimen writing is defined as ‘writing the authorship of which must be known if it is to be used by the FDE for comparison purposes’ (Morris 2000, 129, original emphasis). The obvious difference in the penmanship of ‘By me William’ and the surname suggests that there cannot be certainty that the three words are in Shakespeare’s handwriting.

In addition, as L.L. Schücking observes, the B in the word ‘By’ is unlike the majuscule Bs in D’s Additions (1925, 41). Thompson excuses the ‘malformed’ capital B as ‘owing to [Shakespeare’s] infirmity’ (1923, 105). To this layperson, it does not necessarily look ‘malformed’, perhaps just differently formed. Roy A. Huber notes that the letters h, p, and s of the signatures are formed differently in the Additions. He also points out that none of the letters i in the signatures are dotted, whereas D consistently dots his i’s (1961, 62, 64). These are some of the dissimilarities that make a decisive identification difficult.

At best, all but three capital letters in the alphabet, W, S, and B (the latter of which may have been written by a scribe), are missing from the signature specimens. Thompson notes that the ‘majority of capital letters’ (actually thirteen) in the alphabet are present in D’s Additions and that these letter formations can be used ‘to conjecture the character of the letters which are wanting’ (1923, 103). Conjecture would seem to be of little use to compensate for missing specimens in a handwriting analysis.

Four of the letters in the signatures (i, l, r, and y) are insufficient in Thompson’s view to ‘afford criteria’ for comparison with Hand D (1916a, 57). While this decision is surely a good one (the y occurs only in By, some letters are replaced with marks of contraction), it has the unfortunate result of further shrinking the control sample. Missing letters would seem to present an impediment to a meaningful comparison, as defined in a FDE textbook:

Many of these discriminating elements [of writing] will involve specific letters or combinations of two or more letters in particular relationships to one another. As a result it is important, if not imperative, that the writing standards with which comparisons will be made consist of similar letters and combinations of letters as will occur in similar words, names, texts, or signatures written under comparable circumstances. (Huber and Headrick 1999, 249)

With few exceptions, such discriminants are not available for comparison with D’s Additions. The opportunities available to Thompson for comparing combinations of letters in D’s Additions are obviously limited to those found in the signatures, such as the pe in n. 1 with peace or speake in the Additions, and ha with that or chartered. D’s words makst and forsaks allow for a slightly longer string for comparative purposes. However, these limited letter combinations do not inspire confidence in the fulfilment of Huber’s and Headrick’s injunctions.

Further, the palaeographers in the early 1900s disagreed among themselves as to the spellings in the signatures. With respect to signature n. 1, Thompson spells it Willm Shakp (1923, 59; a line over the letter m indicates abbreviation); Sidney Lee spells it Willm Shak’p (1968, 519); C.W. Wallace (who discovered the signature) spells it Willm Shaks (1910, 500); C.J. Sisson spells it Shak­with no s or p, the hyphen indicating abbreviation (1961, 77n1); Tannenbaum cannot be sure whether it is Wilm or Willu and Shakper or Shaksper (1925, 157).

These palaeographers are basing their transcriptions on a difficult-to-read script so it is not surprising that they propose different spellings. What undermines a meaningful handwriting comparison is the uncertainty concerning the presence or absence of certain letters since, as we have seen, handwriting analyses include the comparison of combinations of letters in both the control and in the questioned document.3 Uncertain combinations include ll, ks, aks, and pe.

If it is not possible to agree on the spelling of a signature and if spellings and letter formations and methods of writing differ from signature to signature, how can any one of those signatures serve as the exemplar? Which one is to be chosen as the standard against which all the others are compared and either accepted or rejected? Or are all of them to be accepted in all their variations by virtue of their presence on the legal documents? In his study of Shakespeare’s Handwriting, Thompson decided to accept signatures n. 1, n. 2, and ‘By me William’ in n. 6 as the control, but his decision did not prevent him from comparing letters written by D with letters in the other signatures (e.g., 1923, 92, 94).

Hagan provides a rule-of-thumb for determinations about signatures

3 On combinations, see Matley 1992, §, 43.

as controls:

Too much care cannot be exercised in the examination of the signatures produced as standards from which to make the comparison as to their character … the time at which they purport to have been written as compared with the date of the contested signature; and where there are several standards presented for comparison they should be by analysis determined as the writing of the same person before comparing them with the contested writing. (1894, 83-84)

It is surely legitimate to question the origin of the words By me William.

Thompson describes the capital B in n. 6 as ‘malformed’ due to Shakespeare’s illness or infirmity (1923, 105). Yet elsewhere he describes the words By me William as written firmly and legibly, in contrast with the ‘weakness and malformation’ of the surname which he attributes ‘certainly to the condition of the dying man’ (1916, 13; see also Hays 1975a, 245-46), a scenario repeated by biographers (e.g., Schoenbaum 1975, 246). There is no external evidence of Shakespeare’s alleged infirmity or ‘writer’s cramp’ as there is, for example, concerning Philip Henslowe’s final illness and palsy, which resulted in him probably dictating his declaration as testator; his will is authenticated with the words ‘Signu mdicta Philip Henslowe’ (Rendle 1887, 157; Sisson 1929, 311; Honigmann and Brock 1993, 103-104).

According to FDE Thomas W. Vastrick, ‘one’s handwriting can change or evolve over long and even short periods of time. Handwriting and signature specimens should be dated as close as possible to the date of the purported writings – ideally, from a few months before to a few months after to offset this phenomenon. This is particularly important if the purported writer is elderly, ill, or sustained an injury around the date of the writings’ (1992, 3, emphasis added).4 Thompson’s explanation of the contrast between ‘By me William’ and the surname on n. 6 is called into question by such cautionary guidelines.

The prospects for a meaningful comparison of handwriting in Shakespeare’s case are even more fraught with difficulties because

when evidence nears the lower limits for positive conclusions that examiners have arbitrarily set for themselves … we see divergence occurring in the findings of examiners. As the strength of the evidence diminishes, conclusions such as ‘a very strong possibility’, ‘a strong probability’, and so on, are qualified with diminishing degrees of probability. (Huber and Headrick 1999, 262)

One might characterize the evidence for Shakespeare’s penmanship as near ‘the lower limits’. Thompson himself comments on the handwriting ‘being of an ordinary type and presenting few salient features for instantaneous

4 On age-related disabilities, see Osborn 1910, 24.

recognition’ (1916a, 29). Yet these six signatures, representing less than half the letters of the alphabet, comprise Thompson’s control sample for comparison with Hand D. And in his opinion, only two of the signatures plus By me William are useful as exemplars.

Again, from Handwriting Identification:

Handwriting comparisons require samples of writing from those individuals who are considered to be potential authors, that … are sufficient in number to exhibit normal writing habits in executing the questioned text or parts thereof, and to portray the consistency with which particular habits are executed. (Huber and Headrick 1999, 247)

Are Shakespeare’s signatures ‘sufficient in number’ to ‘exhibit normal writing habits’? One authority recommends that ‘five or six pages of continuous writing should be adequate for comparison with questioned extended writings, and twenty or more separate signatures should be adequate for comparison with questioned signatures. Others have suggested less, perhaps only half those numbers’ (249).5 Even by the latter measure, six signatures are insufficient in quantity to comprise a control sample that can ‘exhibit normal writing habits’.

It was not until Wallace discovered the Mountjoy signature (Fig. n. 1), which he described as ‘rapid, abbreviated’ (1910, 502), that a signature by Shakespeare exhibited any fluency. Thompson thinks the signature is the best written of the six, as it was ‘inscribed with freedom’ and ‘devoid of [the] hesitation or restraint’ found in the other five signatures (1923, 61; 1916a, 1, 9-10). According to Thompson, ‘if the later signature alone [n. 2, the Blackfriars purchase] had survived, we should have been inclined to judge Shakespeare’s handwriting to have been that of an imperfectly educated man of inferior rank’ (1916a, 27), in striking contrast with the fluent handwriting he sees in signature n. 1. Thompson’s reaction should have set off his own alarm bells:

Since fluency is so important in the determination of genuineness it must be noted that the signature is the single element of one’s writing that is done more automatically, hence more fluently, and with less awareness of the writing process. Even the poorest of writers of other material can have reasonable fluency in their signatures. (Huber and Headrick 1999, 297)

Yet Shakespeare’s signatures did not otherwise exhibit fluency.

Natural variations in ‘normal writing habits’ present another hurdle, since they

5 See also Osborn 1910, 18-19 and Matley 1990, §6, 19.

may be broad or narrow, depending on the individual and the circumstance. Should they be broad, as occurs in less skillful writings, only greater quantities of standards … will properly portray its nature and its range. The variables affecting writing have a greater influence on less skillful writing than on skillful writing. (Huber and Headrick 1999, 250)

And so the task of identifying suitable controls becomes yet more difficult, as Shakespeare’s signatures are not likely to be described as ‘skillful’, not even the one exhibiting some fluency.

There is yet another impediment to Thompson’s case. Following B.A.P. Van Dam and L.L. Schücking, Gerald E. Downs questions an underlying assumption on which Thompson’s case for ‘Hand D’ is based: that D’s Additions are authorial, representing original composition. Downs identifies characteristics in the handwriting, including eyeskip (at lines 127, 130) and mistaken anticipation (the deleted and at line 85),6 both of which are consistent with scribal transcription (2000, 5, 8-9). Hand C, an unnamed playhouse scribe, was transcribing, not composing, and Michael L. Hays reconsiders the possibility that Hands C and D are one and the same (1975b, 69; see also McMillin 1987, 153-154). In addition, if the Hand D Additions are Shakespeare’s so-called ‘foul papers’, they are unique specimens in the More manuscript; other portions of the play are fair copy, whether authorial or scribal.

It is not necessary to prove that Hand D was copying his own composition or that of another. If there is a possibility that Hand D was copying, rather than composing, then there can be no case for Hand D as Shakespeare’s in the throes of composition. As Hays points out, if D’s Additions are fair copy, then ‘paleographic distinctions reflecting changes in the creative process evaporate’ (1975a, 247). In addition, if D’s Additions could be scribal copy, then the field of candidates necessarily expands to other mostly unknown hands, Hand C being a possible exception. An argument that D’s Additions are in Shakespeare’s hand in the act of copying (as proposed by, e.g., Grace Ioppolo 2012, 94) whether his own or somebody else’s work, is still dependent on a valid control sample of his handwriting.

Few of the FDE’s criteria are met in the palaeographic analysis set forth by Thompson. Obviously, modern day handwriting resources were not available to him or his colleagues. But the work of Albert S. Osborn (1910) was available, and also that of William Hagan (1894) and Persifor Frazer (1901). While neither Thompson nor Greg cites early texts by these specialists, both demonstrate an awareness that the time interval between the composition of Hand D and the penning of the six signatures represents an impediment. The dates of composition of both the original text of and the additions to Sir Thomas More remain subjects of disagreement. John Jowett argues for a date of composition of the original Munday-scribed text ca. 1600, while acknowledging that most prior scholarship proposes earlier dates of 1593-1595; he proposes a date of

6 Line numbers are from Jowett 2011, 404-412.

1603-1604 for the additions and revisions, again despite earlier dates of 1593­1597 from prior scholars (Jowett 2011, 424-425; see also Melchiori 1989, 95; Melchiori and Gabrieli 1990, 27). While Jowett’s proposed later dates reduce the time interval between the inscribing of the signatures and the Hand D Additions, they do not reduce it enough.

W.W. Greg dealt with the problem of the time interval by employing a double standard, as his method of identification of Hand E illustrates. Hand E, who wrote some additions to Sir Thomas More, has been identified as that of playwright Thomas Dekker. A number of writing samples by Dekker survive, including a 1616 letter addressed to the actor Edward Alleyn (Greg, et al. 1925­1932, §IX, §X). Greg placed the date of composition of the Additions somewhere between 1593 and 1597. Therefore, Greg did not use Dekker’s letter of 1616 as a basis for comparison to the Hand E Additions, because it was written at least nineteen years later, or, in Greg’s own words, ‘too late for useful comparison’ (Greg 1923a, 53; on time intervals, see Osborn 1910, 145). To use it would violate one of the palaeographer’s rules. However, the first three words of signature n. 6 on Shakespeare’s will, penned in the same year as Dekker’s letter, are admitted to the control as, evidently, not too late for useful comparison. Signatures n. 1 and n. 2 are likewise separated from D’s Additions by an interval only three to four years fewer.

To summarize, the handwriting analysis is impeded by a control sample that is insufficient in quantity and quality to exhibit ‘normal writing habits’ for comparative purposes. The signatures were written ‘too late for useful comparison’. Signatures belong to a different species than dramatic manuscripts. Thompson’s palaeographic arguments do not fare well when considered alongside the methods and rules imposed by others in his field and those in the then-emerging field of FDEs. Yet the Hand D Additions have been tacitly or explicitly elevated to full status as a literary paper trail, and D’s writing is cited to explain how Shakespeare wrote.

3. Analysis by an FDE

In 1961, Roy A. Huber published a paper about Hand D that was first delivered at a Shakespeare Seminar in Canada. He was not a Shakespeare specialist; he was a forensic document examiner who had served in the Royal Canadian Mounted Police for twenty years. He went on to serve as the 24th President of the American Society of Questioned Document Examiners and was the 2003 recipient of the Albert S. Osborn Award of Excellence.

The challenge to Huber was to re-examine Thompson’s case. Huber qualifies his findings, especially since he did not have an opportunity to examine the original manuscript – a serious drawback. He also gives due deference to the palaeographer’s jurisdiction, so to speak, over a case involving the comparison of secretary hands, yet he also hopes that his contribution might ‘suggest areas for further consideration and study’ (1961, 55). Such further study could revisit not only questions about Hand D but also the case for Thomas Heywood as Hand B and the theory that Hands C and D are the same.

Huber examines the principal points of similarity that Thompson identified between the control sample and the questioned document (such points including the ‘spurred’ a, formations of the letters k, s, and p), explains his reasons for downgrading their significance, and concludes that ‘a positive identification’ of Shakespeare as D is not possible (1961, 66). As I read Huber, the different degrees of inconsistencies between D’s Additions and the signatures constitute his most significant reason.

Huber’s brief analysis of the palaeographic case prompted two papers by Shakespeare scholars who attempted to build on his findings and presumably reopen the case in the Shakespearean community. The first paper was by Michael L. Hays in 1975, the second by Paul Ramsey in 1976. Hays identifies an important point:

In the past thirty years, reviews of the problem have offered balanced summaries of both paleographic and literary considerations, generally implying that the weaknesses of the one are remedied by the strengths of the other … This strategy is, however, somewhat disingenuous. First of all, nonpaleographic arguments may reach the same conclusion as paleographic ones, but they cannot strengthen the paleographic arguments themselves. (1975a, 241-242)

Hays is explaining why the ‘full force of cumulative evidence’7argument is flawed. Hays also published on watermarks in some leaves of the More manuscript, and in his conclusion he touches again on the shortcomings of the handwriting case (1975b, 69).

Hays’ essay concerning Huber’s analysis has been largely ignored. In his critical edition of Sir Thomas More, John Jowett does not cite it (or Huber or Ramsay), and he relegates Hays’ ‘Watermarks’ article to a dismissive footnote (2011, 363, n. 2). It is not surprising that Paul Werstine criticizes Jowett’s ‘summary of scholarship on the Shakespeare attribution [as] bent on marginalizing what it demonstrates to be widespread recent scepticism about his authorship of the Hand-D pages’ (2013, 345n29).

The only major collection of essays on Sir Thomas More subsequent to Pollard’s was published in 1989 and edited by T.H. Howard-Hill. Following

G. Harold Metz’s contribution in that collection, the analysis by Huber and the subsequent papers by Hays and Ramsey have all but disappeared from view. Metz cites the editor of The Riverside Shakespeare on the Hand D attribution (1989, 25):

The real strength of the case for Shakespeare’s authorship of these two passages rests, then, not on any single piece or kind of evidence but on the quite remarkable manner in which several independent lines of approach support and reinforce one another pointing to a single conclusion – the ‘hand’ of Shakespeare. (in Evans 1974, 1684)

7 See Metz 1989, 39, n. 26.

In quoting Evans, Metz is also referencing the ‘full force of cumulative evidence’ argument with literary, palaeographic, and bibliographic arguments strengthening each other. They follow Greg, who argued in 1927 that the case rests on ‘the convergence of a number of independent lines of argument … and not on any one alone’ (200; see also Jowett 2011, 438-439). Werstine summarizes the fallacy:

Authorship of the work is credited to Hand D, to whom Shakespeare’s works can be assigned only through an argument from ‘cumulative evidence’ – all of which evidence has been dismissed as inconclusive by Shakespeareans themselves. (1999b, 141)

Even if arguments based on style or linguistics turn out to be 100 percent correct and the author of Hamlet can be proven to have composed the Hand D Additions (and possibly the 21-line soliloquy in C’s hand), there is no way to identify the penman, whether author or scribe. As Eric Rasmussen points out in connection with the attribution of the manuscript of The Second Maiden’s Tragedy, ‘whether or not Middleton’s handwriting appears in the manuscript has no bearing on his authorship of the play’ (1989, 8, n. 24). In other words, the case for Shakespeare’s authorship of the Hand D Additions is independent of the case based on handwriting. Despite the various ‘cumulative’ arguments, only handwriting tests can prove or disprove that Shakespeare inscribed the Hand D Additions, and the only specimens available as controls are inadequate for the purpose.

Metz references the insufficiency of Shakespeare’s handwriting specimens:

The inadequacy of the control [sample] is ineluctable and does in fact constitute a substantial problem to be faced in a palaeographical investigation. This circumstance is the reason Thompson, as Greg noted in his 1927 re-examination, felt constrained to search for minute bits of support and thus overextended some aspects of his argument … Authenticated Shakespearean handwriting is beyond question small in quantity, but it is not negligible. To deny the identification because of a paucity of control exemplars is a refusal to face the problem. (1989, 16)

One should question the identification of Hand D as Shakespeare because of the ‘paucity’ of suitable exemplars.

Metz marginalizes Huber’s analysis as ‘inconclusive’ (17). What Huber actually concludes is that ‘the evidence is not sufficiently strong to justify a positive identification’ of Shakespeare as D (66). In this case, his ‘inconclusive finding’ contradicts Thompson’s attempt at a positive identification (Thompson 1923, 71). Any conclusion finding a degree of probability lower than 100 percent constitutes an ‘inconclusive’ finding, and not in the sense of an inadequate argument.8 Huber stated up front that a positive identification is ‘of necessity a matter of probability’

8 On conclusions and probabilities, see Morris 2000, 138, 216-219; see also Matley 1992, §3.1.1, 34.

(1961, 57), and it would have been helpful if he had quantified his opinion. Ironically, Metz acknowledges that for ‘a significant minority’ of Shakespeareans, there is ‘insufficient evidence to arrive at a decision’ (1989, 17).

Yet, perhaps Huber should have declined the assignment altogether. According to Osborn, ‘many errors in the examination of questioned writing are due to the fact that an adequate amount of standard writing is not obtained before a final decision is given. The competent examiner will decline to give any opinion until a satisfactory basis for such an opinion is available’ (1910, 19).

4.‘Foul Papers’

Many biographers of Shakespeare describe the Hand D Additions as simply part of a play in his handwriting, without further specifying the nature of the manuscript. However, in 1931 W.W. Greg categorized the Additions as ‘foul papers’ (1969, 200).

In his 2013 book-length study of early English playhouse manuscripts, Paul Werstine traces the genesis of Greg’s concept of ‘foul papers’ and its influence on Shakespeare studies. The term is found in at least two Jacobean records. In 1613, the playwright Robert Daborne refers to sending along to Philip Henslowe his ‘foule sheet’ instead of ‘yefayr I was wrighting’ (quoted in Chambers 1963, 1:96). In an annotation ca. 1619-1624 concerning John Fletcher’s play Bonduca, Edward Knight, who copied the play, wrote that ‘this hath beene transcrib’d from the fowle papers of the Authors wch were found’ (quoted in Greg 1925, 152).

Knight’s reference to ‘fowle papers’ led Greg to develop a hypothetical definition of the term. Greg imagined ‘foul papers’ as ideals, intended to serve as proxies for the lost manuscripts (until such time as any might be discovered) that were submitted as printer’s copy for publication. If correctly defined and if any actual ‘foul papers’ could be found, then the features they would contain could explain problems and corruptions in certain printed texts. Among the important features that Greg inferred from the Bonduca transcript were misplaced passages, lacunae, and illegible handwriting (Werstine 2013, 13, 38, 41, 98). However, Greg’s original 1927 essay on the topic was rejected by The Library and remained uninspected until Grace Ioppolo found it at the Huntington Library and published parts of it in 1990.

In his subsequent works, Greg’s discussions about ‘foul papers,’ while grounded in his unpublished hypothetical definition, regularly made reference to ‘rough drafts’ (e.g., 1925, 156; 1931, 195-197, 199; 1951, 27, 31). Beginning in the 1930s, most, perhaps all, of the editors and critics employing the term ‘foul papers’ did so without reference to or knowledge of Greg’s unpublished definition. In 1955, Greg again asserted that Hand D’s Additions were Shakespeare’s ‘foul papers’ (108-109). He also identified numerous Shakespeare texts as based on hypothetical ‘foul papers’ serving as printer’s copy, such as the ‘good’ quartos of Romeo and Juliet, Much Ado About Nothing, and King Lear. In the absence of any known printer’s copy of any description, his claims are speculative. But they have remained largely unchallenged, Werstine’s work being a major exception.

Today the term ‘foul papers’ is usually employed without reference to Greg’s original formulation and usually without acknowledgement that there are no extant manuscripts that fit his formulation (Werstine 2013, 34-38). Instead, the editors of, for example, the Oxford Textual Companion offer a general definition as ‘an author’s first complete draft of a play’ (Wells et al. 1997, 9). They claim that Shakespeare’s handwriting in his ‘foul papers’ is ‘implied by errors in printed editions’ and with reference to D’s handwriting (124, 510; see also Jowett 2007, 99, 101). The Arden editor of Much Ado About Nothing identifies in the text Shakespeare’s ‘characteristic lightness of punctuation’ and ‘characters designated variously both by function and by given name’, among other idiosyncratic elements contained in Shakespeare’s ‘foul papers’ (McEachern 2007, 128-129). Thus an editor may claim that certain features and imperfections can be explained by Shakespeare’s ‘foul papers’ even though nobody knows what they looked like, and the identity of the penman who inscribed the Hand D Additions remains unknown.

Ioppolo’s confidence in the number of printed Shakespearean texts ‘almost certainly’ based on authorial ‘foul-papers’ – at least seventeen – reflects the widespread acceptance of the term as synonymous with ‘rough’ or ‘working draft’ (2012, 93). She is able to refer to ‘extant autograph foul papers’, plural, because she explicitly replaces Greg’s ‘constricted definition of foul papers’ with ‘the working draft by the author(s)’ (Ioppolo 2006, 7; 2012, 88). Her definition is even more flexible than the Oxford editors’ ‘first complete draft’ and contains features such as false starts, duplications, and ‘confusions in character names and interactions’, among others (2012, 91). None of the manuscripts in her discussion of ‘foul papers’ can be shown to have served as printer’s copy, and many characteristics in her definition, including duplication and ‘confusions in character names’ are also present in other types of manuscripts; e.g., transcripts for playhouse use and, as significantly, the longhand transcription of a shorthand report (Downs 2007-2008, 126; Werstine 2013, 9 and passim).

Disagreement concerning definitions is illustrated with reference to Thomas Heywood’s manuscript of The Captives. Werstine classifies Heywood’s manuscript as ‘a playhouse [manuscript] used for performance’ (2013, 305), but Ioppolo classifies it as ‘annotated foul papers’ (2006, 95; 2012, 91-92). However, if Heywood’s rough draft, his ‘foul papers’ were sufficiently legible for submission to the playhouse, then by definition, they served as ‘fair copy’ (Heywood’s labour-saving strategy may not necessarily be shared by other dramatists, such as Daborne). McMillin describes this alternative with respect to D’s Additions, which could be ‘first-draft writing [‘foul papers’] which turned out to be usable without copying’ (1987, 144). Werstine makes the point that only upon transcription did a dramatist’s original draft become his ‘foul papers’ (2013, 98-99, 100). It would therefore be unlikely that any of the extant theatrical manuscripts, including the Hand D Additions, are ‘foul papers’ as Greg conceived of them.

Further complicating the designation of Hand D’s Additions as ‘foul papers’ is Ioppolo’s decision to classify them as ‘authorial fair copy’ (2006, 104; 2012, 94). D’s Additions alone cannot decide the question for Shakespeare: Jowett agrees with the palaeographers who describe them ‘as showing a writer in the immediate process of composition’ (2011, 440), that is, ‘foul papers’ (2007, 99); Ioppolo proposes authorial fair copy; Downs suggests a scribal transcript. However, if Ioppolo is correct to classify the Hand D Additions as ‘authorial fair copy’, then either all arguments asserting or implying that D’s Additions represent Shakespeare’s ‘foul papers’ are mistaken, or alternatively, a definition of ‘foul papers’ is a matter of choice. Werstine describes the term as ‘a matter of uncertain interpretation’ (2013, 33), which allows an editor to select features observed in extant manuscripts to be included in a revised definition of ‘foul papers’. These less precise and overly inclusive definitions illustrate the circularity summarized by H.R. Woudhuysen: ‘ “foul-paper” texts can be identified by the presence of those features which are characteristic of “foul-paper” texts’ (1998, 320).

No printer’s copy survives. No ‘foul papers’ as conceived by Greg have been discovered, and definitions remain elastic, or in Ioppolo’s word, ‘fluid’ (2012, 87). In his Arden edition of King Lear, R.A. Foakes describes his method for attempting to peer through the so-called ‘veil of print’: ‘The only evidence we have for the copy that lies behind the text printed in Q is to be found within it; the nature of the manuscript has to be inferred, and arguments can never be conclusive in the absence of external proof’ (2000, 119).

The Hand D Additions are part of a manuscript intended for use in the playhouse, not in the print shop, and they contain characteristics consistent with a scribal transcript. They do not contain the principal features enumerated by Greg in his hypothetical definition of ‘foul papers’. The penman’s identity cannot be proven on the available evidence. In the meantime, many editors continue to accept D’s Additions to More as Shakespeare’s ‘foul papers,’ and many biographers continue to accept them as a literary paper trail in Shakespeare’s handwriting.

5. Orthography

In the case for Shakespeare as D, a significant part of the ‘force of cumulative evidence’ argument is the demonstration of ‘Shakespearean’ spellings. The case based on orthography was introduced in 1923 by John Dover Wilson. Wilson could barely contain his excitement when he discovered that Scilens, a rare spelling for ‘Silence’, appears in both Q 2 Henry IV (1600) and in the Hand D manuscript (128-129). The spelling of Scilens was clearly, in Wilson’s view, an authorial choice, because Scilens is the name of a Shakespearean character, ergo a sacrosanct designation, not just an idiosyncratic spelling for a common word. But Q offers Wilson no support that Scilens is an authorial choice. The name of the silly Justice occurs in Q forty times (including in the cancelled leaves). In dialogue, as a speech prefix or in a stage direction, the Justice’s name is spelled Scilens eighteen times, Silence three times, and Silens nineteen times (as well as in abbreviated form in the cancelled leaves). The variations mean that a particular preference of spelling cannot be argued, regardless of whose supposed ‘preference’ it might represent – author’s, scribe’s, compositor’s, or editor’s.

In the Hand D Additions, the word scilens at line 59 is an interjection, not a proper name, and the word ‘silenced’ at line 78 is spelled by D as sylenct. Similar but not identical variations are found in the 1611 manuscript of The Second Maiden’s Tragedy, which contains four instances of silence and three instances of scilence.

The editors of the Revels edition of the play gloss the word as a spelling ‘found also eighteen times in 2HIV, and nowhere else in Elizabethan texts’ (Melchiori and Gabrieli 1990, 98, n.; see also Schoenbaum 1966, 105; Jowett 2007, 13; Jowett 2011, 442). It is surely significant, however, that the spelling occurs nowhere else in Shakespearean texts, either. The word ‘silence’ occurs dozens of times in the Shakespeare corpus, but in no other instance is it spelled scilens. Other texts supposedly set from Shakespeare’s ‘foul papers’ and containing the spelling silence include Q A Midsummer Night’s Dream (six instances), Q Much Ado About Nothing (two instances) and Q Merchant of Venice (five instances). In essence, to identify this ‘Shakespearean’ spelling, scholars are comparing a rare spelling of a character’s name, found multiple times amongst two other spellings of the same word, in a quarto produced by compositors using unknown printer’s copy, with a single instance of the rare spelling found in a manuscript penned by a (possibly authorial) scribe.

Further, if a particular preferred spelling of a character’s name is a hallmark of Shakespeare’s authorship, as Wilson suggests, then Hand D fails as Shakespearean. D usually spells the title character’s name as ‘moor’, but he also spells it ‘moore’, ‘more’, and ‘moo’. Greg acknowledges that speech prefixes in the Hand D Additions were omitted on the penman’s first pass, and that the speech prefixes subsequently added by D are ‘perfunctory’ in nature (1923b, 229).

Even while identifying possible phonetic error in Q2 Romeo and Juliet (2013, 98), Arden editor René Weis’ argument goes back to Shakespeare’s spelling preferences as found in D’s Additions:

As we know from Sir Thomas More and other foul papers using words ending in –ce, Shakespeare’s spelling practice is to drop the final e, hence ‘obedyenc’ (6.47), ‘obedienc’ (at 107 and 129), ‘insolenc’ (92), ‘offyc’ (112) and frraunc (143) in Sir Thomas More. (100)

However, other spellings in the Hand D Additions of words ending in ce (with the final e) include audience (47), elevenpence (2), and violence (132). Other texts presumably based on ‘foul papers’ with final e spellings include Q A Midsummer Night’s Dream (obedience; disobedience), Q1 Lear (France, offence; notice; office); and Q 2 Henry IV (office, obedience, Prince).

A similar argument is advanced by an Oxford editor of A Midsummer Night’s Dream:

There is more than enough evidence, according to the canons of bibliographical proof, to show that the copy for Q1 was autograph ‘foul papers’ … While Elizabethan compositors often varied the spelling of words, a number of the more unusual spellings in Q1 agree with the spellings used by Hand D in the manuscript of the play of Sir Thomas More, a handwriting usually held to be Shakespeare’s own. In particular Hand D and Q1 share a preference for using ‘oo’ (in, for instance: prooue, hoord, boorde, shooes, mooue). (Holland 2008, 11, 114)

Q1 Lear contains thirteen instances of prove or proves, none spelled with the ‘oo’. Q2 Romeo contains seventeen instances of move and seven instances of prove, none spelled with the ‘oo’. Other supposedly ‘Shakespearean’ spellings such as deules (devils) or Iarman (German) are as easily disproved.

No one can know how many agents intervened between an author’s manuscript and the printed text. Honigmann explains that an editor ‘wants to know how many scribes and compositors copied and set the text’ since they ‘normally changed spelling and punctuation’; even straightforward reprints introduced spelling changes in the printing house (1998, 353). Arguments concerning ‘Shakespearean’ spellings cannot be sustained when spellings in all texts supposedly based on ‘foul papers’ are tabulated.

6. Conclusions

Shakespeare’s handwritten leaves in the Sir Thomas More manuscript would fill an evidentiary vacuum: If Hand D is a written specimen of Shakespeare’s, he not only left behind a literary paper trail, he left one of the highest quality, a manuscript in his handwriting, whether ‘foul papers’ or authorial fair copy. However, the evidence and the arguments based on handwriting, spellings, and assumptions about ‘foul papers’ do not support such conclusions.

The handwriting case for Shakespeare as D cannot be made on the available evidence: the control sample is inadequate in quantity and quality, signatures and dramatic compositions belong to different classes or species, and the years between the penning of D’s Additions and the signatures render comparisons less useful. The related case that Hand D is an example of ‘foul papers’ relies on a term that is variously defined without reference to any surviving manuscript that served as printer’s copy and without reference to Greg’s original conception; the term remains open to interpretation. ‘Shakespearean’ spellings are based on selective comparisons. With the exception of arguments and data based on stylistics, such as collocations of words and imagery (which are independent of the handwriting analysis), the ‘force of cumulative evidence’ argument is instead comprised of disproved, unproven, or unprovable assumptions.

In the years since 1923, many scholars, editors, and critics have claimed Hand D as Shakespeare’s, and the mere repetition of that claim has bestowed on it a misplaced legitimacy. David Hackett Fischer identifies the logical fallacy as ‘proof by repetition’ (1970, 302-303). Yet despite deficient evidence and faulty arguments, the case for Hand D not only has survived, as of 2015, it is thriving beyond Pollard’s wildest dreams.

Works Cited

Abt Jeffrey (1987), ‘Objectifying the Book: The Impact of Science on Books and Manuscripts’, Library Trends, Summer, 23-38, <https://goo.gl/DhCeYB>, accessed 16 December 2014.

Bruster Douglas (2013), ‘Shakespearean Spellings and Handwriting in the Additional Passages Printed in the 1602 Spanish Tragedy’, Notes and Queries, 19 July, <http:// goo.gl/mgmoZE>, accessed 19 August 2012.

Chambers E.K. (1963 [1930]), William Shakespeare: A Study of Facts and Problems, London, Oxford University Press, 2 vols.

Craig Hugh (2012), ‘Authorship’, in A.F. Kinney, ed., The Oxford Handbook of Shakespeare, Oxford, Oxford University Press, 15-30.

Dalton Jacob, Tom Davis, and Sam van Schaik (2007), ‘Beyond Anonymity: Paleographic Analyses of the Dunhuang Manuscripts’, The Journal of the International Association of Tibetan Studies 3, December, <http://goo.gl/cVa1CZ>, accessed 20 November 2014.

Downs G.E. (2000), ‘Hand D & The Book of Sir Thomas More: By the Nature of Your Error’, pamphlet privately printed following presentation at ‘The Second Shakespearean Research Symposium’, Detroit, October 2000. Bibliographical note: A later version of this paper was announced by Studies in Bibliography for vol. 57 (2005-2006). The editor withdrew the paper since shortly before publication a version was published in 2007 without permission as ‘A Question (not) to be Askt: Is Hand D a Copy?’ in Shakespeare Yearbook 16, Lewiston, The Mellen Press, 241-266.

Downs G.E. (2007-2008), ‘Memorial Transmission, Shorthand, and John of Bordeaux’, Studies in Bibliography 58, 109-134.

Evans G.B., ed. (1974), The Riverside Shakespeare, Boston, Houghton Mifflin Co.

Fischer D.H. (1970), Historians’ Fallacies: Toward A Logic of Historical Thought, New York, Harper Perennial.

Foakes R.A., ed. (2000 [1997]), King Lear, London, The Arden Shakespeare.

Frazer Persifor (1901), Bibliotics or the Study of Documents, Philadelphia, J.P. Lippincott Company.

Gore-Langton Robert (2014), ‘The Campaign to Prove Shakespeare Didn’t Exist’, Newsweek, 29 December, <http://goo.gl/Z2FJTC>, accessed 1 January 2015.

Greenwood George (1916), Is There a Shakespeare Problem?, London, John Lane, The Bodley Head.

Greenwood George (1920), Shakspere’s Handwriting, London, John Lane.

Greg W.W. (1923a), ‘The Handwriting of the Manuscript’, in W.W. Greg, ed., Shakespeare’s Hand in the Play of ‘Sir Thomas More’, Cambridge, Cambridge University Press, 41-56.

Greg W.W. (1923b), ‘Special Transcript of the Three Pages’, in W.W. Greg, ed., Shakespeare’s Hand in the Play of ‘Sir Thomas More’, Cambridge, Cambridge University Press, 228-269.

Greg W.W. (1925) ‘Prompt Copies, Private Transcripts, and the “Playhouse Scrivener” ’, The Library, fourth series, 6, 2, 148-156.

Greg W.W. (1951), The Editorial Problem in Shakespeare: A Survey of the Foundations of the Text, Oxford, Clarendon Press.

Greg W.W. (1955), The Shakespeare First Folio, Oxford, Clarendon Press.

Greg W.W. (1966 [1927]), ‘Shakespeare’s Hand Once More’, in J.C. Maxwell, ed., W.

W. Greg Collected Papers, Oxford, Clarendon Press, 192-200. Greg W.W. (1969 [1931]), Dramatic Documents From the Elizabethan Playhouses, Oxford, Clarendon Press, 2 vols. Greg W.W., et al. (1925-1932), English Literary Autographs: 1550-1650, Oxford, Oxford University Press.

Hadfield Andrew (2013), ‘Theorizing Shakespeare’s Authorship,’ in P. Edmondson and S. Wells, eds, Shakespeare Beyond Doubt: Evidence, Argument, Controversy, Cambridge, Cambridge University Press, 64-66.

Hagan W.E. (1894), A Treatise on Disputed Handwriting, Albany, Banks & Brothers.

Haselden R.B. (1935), Scientific Aids for the Study of Manuscripts, Oxford, Oxford University Press.

Hays M.L. (1975a), ‘Shakespeare’s Hand in Sir Thomas More: Some Aspects of the Paleographic Argument’, Shakespeare Studies 8, 241-253.

Hays M.L. (1975b) ‘Watermarks in the Manuscript of Sir Thomas More and a Possible Collation’, Shakespeare Quarterly 26, 1, 66-69.

Holland Peter, ed. (2008 [1998]), A Midsummer Night’s Dream, Oxford, Oxford University Press.

Harmon William, C.H. Holman and W.H. Thrall (1992), A Handbook to Literature, sixth edition, New York, Macmillan.

Honigmann E.A.J., ed. (1998 [1997]), Othello, London, The Arden Shakespeare.

Honigmann E.A.J. and Susan Brock (1993), Playhouse Wills 1558-1642, Manchester, Manchester University Press.

Howard-Hill T.H., ed. (1989), Shakespeare and ‘Sir Thomas More’: Essays on the Play and its Shakespearian Interest, Cambridge, Cambridge University Press.

Huber R.A. (1961), ‘On Looking Over Shakespeare’s “Secretarie” ’, in B.A.W. Jackson, ed., Stratford Papers on Shakespeare 1960, Toronto, W.J. Gage, 51-70.

Huber R.A. and A.M. Headrick (1999), Handwriting Identification: Facts and Fundamentals, New York, CRC Press.

Ioppolo Grace (1990), ‘The Final Revision of Bonduca: An Unpublished Essay by W.W. Greg’, Studies in Bibliography 43, 62-80.

Ioppolo Grace (2006), Dramatists and Their Manuscripts in the Age of Shakespeare, Jonson, Middleton, and Heywood: Authorship, Authority and The Playhouse, London, Routledge.

Ioppolo Grace (2012), ‘Revision’, in A.F. Kinney, ed., The Oxford Handbook of Shakespeare, Oxford, Oxford University Press, 85-99.

Jowett John (2007), Shakespeare and Text, Oxford, Oxford University Press.

Jowett John, ed. (2011), Sir Thomas More, London, The Arden Shakespeare.

Lee Sidney (1968 [1931]), A Life of William Shakespeare, New York, Dover Publications.

Looney J.T. (1920), ‘Shakespeare’ Identified in Edward De Vere, the Seventeenth Earl of

Oxford, New York, Frederick A. Stokes Co.

Matley M.B. (1990), Exemplars: Genuine Samples for Comparison with Questioned Writing and Documents, San Francisco, Handwriting Services of California.

Matley M.B. (1992), Studies in Questioned Documents: Number Seven: Reliability Testing of Expert Handwriting Opinions, San Francisco, Handwriting Services of California.

McEachern Claire, ed. (2007 [2006]), Much Ado About Nothing, London, The Arden Shakespeare.

McMillin Scott (1987), The Elizabethan Theatre and ‘The Book of Sir Thomas More’, Ithaca, Cornell University Press.

Melchiori Giorgio (1989), ‘The Book of Sir Thomas More: Dramatic Unity’, in T.H. Howard-Hill, ed., Shakespeare and ‘Sir Thomas More’: Essays on the Play and its Shakespearian Interest, Cambridge, Cambridge University Press, 77-100.

Melchiori Giorgio and Vittorio Gabrieli, eds (1990), Sir Thomas More, The Revels Plays, Manchester, Manchester University Press.

Metz G.H. (1989), ‘ “Voice and credyt”: The Scholars and Sir Thomas More’, in T.H. Howard-Hill, ed., Shakespeare and ‘Sir Thomas More’: Essays on the Play and its Shakespearian Interest, Cambridge, Cambridge University Press, 11-44.

Middleton Thomas (1909 [1611]), The Second Maiden’s Tragedy, ed. by W.W. Greg, Malone Society, Oxford, Oxford University Press. Morris Ron (2000), Forensic Handwriting Identification: Fundamental Concepts and Principles, London, Academic Press. Osborn A.S. (1910), Questioned Documents: A Study of Questioned Documents with an Outline of Methods, Rochester, Lawyers Co-operative Publishing Company. Oxford English Dictionary (2000), second edition, CD 3.0 including Additions Series Volumes 1-3. Pollard A.W. (1923a), ‘Preface’, in A.W. Pollard, ed., Shakespeare’s Hand in The Play of Sir Thomas More, Cambridge, Cambridge University Press, v-vi.

Pollard A.W. (1923b), ‘Introduction’, in A.W. Pollard, ed., Shakespeare’s Hand in The Play of Sir Thomas More, Cambridge, Cambridge University Press, 1-40. Price Diana (2001), Shakespeare’s Unorthodox Biography: New Evidence of an Authorship

Problem, Westport, Greenwood Press. Ramsey Paul (1976), ‘Shakespeare and Sir Thomas More Revisited: Or, A Mounty on the Trail’, Papers of the Bibliographical Society 70, 333-346. Rasmussen Eric (1989), ‘Shakespeare’s Hand in The Second Maiden’s Tragedy’, Shakespeare Quarterly 40, 1, 1-26.

Rendle William (1887), ‘Philip Henslowe’, The Genealogist IV, n.s., 149-159.

Schoenbaum Samuel (1966), Internal Evidence and Elizabethan Dramatic Authorship: An Essay in Literary History and Method, Evanston, Northwestern University Press. Schoenbaum Samuel (1975), William Shakespeare: A Documentary Life, Oxford, Clarendon Press. Schücking L.L. (1925), ‘Shakespeare and Sir Thomas More’, Review of English Studies 1, 1, 40-59. Shakespeare William (2011), Sir Thomas More, ed. by J. Jowett, London, The Arden Shakespeare. Shakespeare William (1990), The Second Part of King Henry the Fourth, Malone Society Reprint, Oxford, Oxford University Press.

Sisson C.J. (1929), ‘Henslowe’s Will Again’, Review of English Studies 5, 308-311.

Sisson C.J. (1961), ‘ “Postscript” to R.A. Huber’s “On Looking Over Shakespeare’s ‘Secretarie’ ” ’, in B.A.W Jackson, ed., Stratford Papers on Shakespeare 1960, Toronto, W.J. Gage, 70-77.

Stokes P.A. (2007-2008), ‘Palaeography and Image-Processing: Some Solutions and Problems’, Digital Medievalist 3, <http://www.digitalmedievalist.org/journal/3/ stokes/>, accessed 16 December 2014.

Sutton William (2013), Shakespeare Institute posting on Merriam Webster Dictionary at <http://www.merriam-webster.com/dictionary/bibliotics>, accessed 24 November 2014.

Tannenbaum S.A. (1925), ‘Shakespeare’s Unquestioned Autograph and the Addition to Sir Thomas Moore’, Studies in Philology 22, 133-160.

Tannenbaum S.A. (1927), The Booke of ‘Sir Thomas Moore(A Bibliotic Study), New York, The Tenny Press.

Taylor Gary (1989), ‘The Date and Auspices of the Additions to Sir Thomas More’, in

T.H. Howard-Hill, ed., Shakespeare and ‘Sir Thomas More’: Essays on the Play and

its Shakespearian Interest, Cambridge, Cambridge University Press, 101-129.

Works spectacularly left uncited

Everything of importance written on Hand D, including the now unequivocal attribution by The British Library, with the exception of Pollard's 1923 collection of essays. Also previous Oxfordian work on Hand D, probably now thought dangerously ridiculous.

Annotated Articles