This blog: Authorship attribution in ‘Shakespeare’…
In my last post I noted that in his chapter “Authorship,” in the Oxford Handbook of Shakespeare (2012), Professor Hugh Craig asks: “Which of the works sometimes attributed to Shakespeare are apocryphal? Which plays are in fact collaborations? Which sections of plays outside the canon were in fact written by Shakespeare? How many works can be attributed to Shakespeare as sole author?” (Craig 2012, p.p. 15-16).
As noted by me in a previous post, Hugh Craig advises that attribution studies in Shakespeare necessarily involve two strands:
The main tool for the attribution of a disputed passage to Shakespeare is comparison with well-accepted Shakespeare works, and the same procedures would operate whoever is assumed to be actually holding the pen. But in one case there is a convergence. A manuscript ‘playbook’ of the play Sir Thomas Moore survives. A series of essays in a landmark volume from the 1920s edited by Alfred W. Pollard distinguishes various hands at work in the manuscript. One of them, known as ‘Hand D’, resembles Shakespeare’s signature, which is the only known handwriting of his that survives. On a stylistic side, strong evidence from spelling and shared words and phrases links the linguistic content of this part of the play to Shakespeare. If these two bodies of evidence can be sustained, then the Hand D passages provide for once a link between ‘Shakespeare’ texts and William Shakespeare of Stratford. (Craig 2012, p. 17)
Also noted previously, is that Hugh Craig points out that “Shakespeare as an author can also be defined purely by reference to his language. At the simplest level this is a network of preferences in vocabulary and grammatical constructions. Then there are characteristic expressions, figures of speech, and images…. and Shakespeare’s seductive eloquence” (Craig 2012. p. 22).
In an earlier post again, I wrote that Professor Craig’s “computational analysis has two applications in the field of literature: One, it can help authenticate authorship that is unknown or suspected to have been wrongly attributed, and it can help determine authorship contribution in the most disputed of Shakespeare’s works, and it can help determine authorship in Shakespeare’s works where collaboration is suspected, and where these works have gained significant attention from academic Shakespeareans and literary historians as to warrant further investigation, or are deemed worthy of serious consideration; two, it can be used to build a profile of or define a writer’s particular style, and which is of great importance to scholarship.”
Before I attempt Hugh Craig’s digitised stylistics as explained in his 2009 book co-authored with Professor Arthur Kinney (Hugh Craig and Arthur F, Kinney, eds. Shakespeare, Computers, and the Mystery of Authorship. Cambridge, UK: Cambridge University Press, 2009.), I will now return to his 2012 chapter, “Authorship,” in the Oxford Handbook of Shakespeare (2012) to look further at what this eminent Shakespearean says on the linguistic content in Shakespeare. Professor Hugh Craig is very knowledgeable, and his beautifully-written chapter, “Authorship”, is well worthwhile quoting from extensively.
[With Shakespeare,] all readers and listeners have the experience of hearing an authorial voice in a phrase, or in a favourite unusual word, or in a characteristic transition from one idea to another. It turns out that this kind of linguistic innovation is so marked and persistent that its traces in frequencies and distributions of individual words can be modelled statistically. This allows us to compare our intuitions as readers about the authorship of speeches and scenes with an objective set of measures. It also shows that authorial style, in the sense of highly individualized and consistent language use, is a reality, and not a romantic and sentimental fiction. To illustrate: a Shakespeare passage is twice as likely to include the words gentle and beseech as a passage by one of his contemporaries. Wealth, pride, and lust, on the other hand, are half as likely to turn up in a Shakespeare passage as in the rest. (Craig 2012, p. p.22-23)
Craig says that “There is good reason to think that Shakespeare was involved in the writing of forty-four plays,” and he writes “We need to make distinction among the forty-four plays” (Craig 2012, p. 23):
A core group of twenty-eight surviving plays are widely accepted as entirely by his hand, if not entirely without challenge. [  ] Beyond this, Love’s Labour’s Won seems to have been be a single-author play but is lost. Another set of six plays seem to be collaborations in the straightforward sense that in them Shakespeare worked with another dramatist on a joint effort. Thus it is very likely that George Peel wrote part of Titus Andronicus, George Wilkins part of Pericles, Thomas Middleton part of Timon of Athens, and John Fletcher parts of Henry VIII and Two Noble Kinsmen. A third likely collaboration with Fletcher, Cardenio, is lost. With five plays we believe Shakespeare to have written a portion, but are uncertain of the number or identity of his collaborators. Measure for Measure and Macbeth seem to be Shakespeare single-author texts with additions or revisions by Middleton. Finally, there is reason to believe there are two surviving plays to which Shakespeare added passages sometime after their original performance: The Spanish Tragedy, more speculatively, and Sir Thomas Moore, now beyond reasonable doubt. (Craig 2012, p. 23)
It seems to me that, in a way, Shakespeare is not unlike the Venus de Milo and Leonardo Da Vinci’s Mona Lisa in that they all have an air of mystery about them and possess a timelessness that crosses social and cultural divides, a quality that seems to emanate from within the works themselves. Yet I sometimes feel that the mystique surrounding the Mona Lisa and the statue of the Venus de Milo would almost dissipate on one level if everything about them–who, why, when, how–were to be revealed even though these works of fine art will ever be priceless treasures and a connection to our past as humans. To my mind, Shakespeare’s exquisite turn of phrase and unusual use language and combinations of words, and his uncanny ability to decipher the human psyche and portray human behaviour accurately, gets at the heart of the matter; Shakespeare speaks to us as ordinary people, as men and women, and shows us to ourselves as individuals–how we can or do behave or what we are each capable of–and it shows us “ordinary” and personal life in action. In part, this gives his works an added mystique to that surrounding the Venus de Milo and the Mona Lisa, and one which would survive (and more than likely increase,) if he were to come back to life and offer lectures on his work.
Imagine the mad scramble there would be! Scholars and academics (me included), and a host of thousands would climb all over each other to get tickets to attend! It would be war!
In his chapter “Authorship,” after making a distinction between the forty-four plays listed in the Shakespeare canon, Hugh Craig writes: “This estimate of the Shakespeare canon rests on centuries of work by an extraordinary band of individuals” (Craig 2012, p. 24).:
In Shakespeare’s lifetime… his works consisted primarily as a large collection of play-scripts belonging to his theatre company … some of which had been printed in various degrees of care and accuracy, and as a smaller assortment of printed and manuscript poems. The first attempt at collecting the plays was in 1619, when Thomas Pavier and William Jagger, printers and stationers put ten plays into a common format so that they could be bound together in a single volume, or sold separately. This set included two plays now thought to be by others and several to which Pavier and Jagger did not in fact own publishing rights. Four years later two of Shakespeare’s fellow-actors and shareholders, John Heminges and Henry Condell, published the First Folio, presenting thirty-six plays, eighteen never before published, and many of the others in new versions. In their dedication, and again in the preface, they say that since Shakespeare did not live to publish his writings himself, the task of collecting the plays and putting them in print has fallen to them as his friends. Heminges and Condell thus present themselves as Shakespeare’s literary executors, and this strong connection is supported by Shakespeare’s bequest to them, along with Richard Burbage, of money for funeral rings. (Craig 2012, p.p. 23-24)
None of the plays included in their collection has been excluded from the modern canon, though scholars now agree that several of them contain work by other writers. (Craig 2012, p.p. 23-24)
Hugh Craig writes that, “Over the last two and half centuries, since, say, the founding of the New Shakespeare Society in 1874, two broad tendencies have been evident in work on Shakespeare’s canon” (Craig 2012, p. 24):
One of [these tendencies] is to confirm the integrity of the thirty-six plays in the Folio, to see a single authorial controlling influence through this stable set of dramatic works, and a largely uniform progress through time with each play as a milestone, surviving more or less intact from the moment of its first creation….
The other tendency is to see the Folio canon as a more arbitrary and questionable collection. Adherents of this second view argue that the volume may include sections, or at least layers, or work by others, beyond the well-attested collaborations. like Henry VIII, which do appear within its covers….
A related controversy in studies of the canon is between those who give credence to internal, stylistic evidence and those who do not. [S.] Schoenbaum [(1996)] gives the first two and a half decades of the twentieth century the somewhat ironical title of the ‘Golden Age’ of attribution based on style rather than documentary evidence. Too often, as he shows, the table of statistics of metrical patterns and the lists of rare words, parallel passages, and image clusters were merely ‘impressionism rationalised’….Often enthusiasts failed to carry out what M. St C. Byrne [(1932)] called the ‘negative check’ to see if a phrase or wording was really characteristic and not a commonplace. She points out, too, that they often overlooked the fact that that if a parallel might be a sign of commonplace authorship, it might also be a plagiarism or an coincidence. (Craig 2012, p.p. 24-25)
With advancements in computer technology in more recent times, correct analysis of authorship attribution in Shakespeare is becoming more reliable:
Brian Vickers has demonstrated how often scholarly work going back to the middle of the nineteenth century arrived at what now seem to be accurate divisions of collaborative plays between Shakespeare and other authors. With searchable text provided by collections like Literature Online and Early English Books Online, present-day scholars have something like comprehensive coverage of surviving plays, so that negative controls can be watertight. Countable electronic text, allowing a statistical approach to word frequencies, offer a further step forward. With these resources it should be possible at last to pursue authorship questions ‘upon a general and disinterested method, rather than along the casual lines of advance opened up by the pursuit of an author for this or that suspected or anonymous play. (Craig 2012, p.p. 25-26)
Looking in more detail at some of the outstanding problems in the canon, we can start with doubts about work usually printed in a collected edition. A Lover’s Complaint is a case in point. There is strong external evidence connecting this poem with Shakespeare. It was published with the Sonnets, and is attributed to ‘William Shakespeare’ on its own separate title-page. On the other hand, if it is by Shakespeare, it is a departure from his regular style. (Craig 2012, p. 26)
This matter, says Hugh Craig, became the subject for argument:
Colin Burrow, editing the poems in 2002 for the Oxford edition, declared that the studies by Kenneth Muir and MacDonald P. Jackson in the 1960s had concluded the attribution debate in favour of Shakespeare. However, another eminent figure in in Shakespeare authorship studies, Brian Vickers, has recently argued for John Davies of Hereford as the more likely author. (Craig 2012, p. 26)
In one section of his [2007) book on the subject, says Craig, Vickers “sets out to show that the poem is distinct from Shakespeare in its vocabulary, its syntax, its verse, and its use of some rhetorical figures and metaphor,” and he argues that “the Lover’s Complaint poet is much less skilful than Shakespeare…”;
A statistical study by Ward Elliot and Robert J. Valenza applying a series of empirical tests to the poem, also declares the poem to be outside the range of Shakespeare’s practice. The debate is thus resolved. Shakespeare studies in general seem to be able to tolerate this uncertainty. Complete Shakespeare editions almost invariably include the poem, and critical studies continue to declare confidently that ‘Shakespeare’ includes five printed poems, the two narrative poems, the Sonnets, The Phoenix and the Turtle, and A Lover’s Complaint, but only the occasional critical enterprise could be said to depend on the attribution for its validity. (Craig 2012, p. 26)
In recent years, there have been other debates amongst scholars about authorship contribution in Shakespeare’s works and about authorship where collaboration is suspected. Craig advises that, “If Lover’s Complaint illustrates that areas of doubt in Shakespeare attribution remain, even after the application of the most sophisticated and modern methods, then another area, dramatic collaboration, shows how some long-standing debates can reach closure” (Craig 2012, p. 26).
In his book Shakespeare, Co-Author Vickers deals with the five collaborations…–one each with Peele, Wilkins and Middleton, and two with Fletcher–and … reviews previous studies, and adds new ones to show a convergence of differing approaches to a consensus not only on the partnerships involved, but on the divisions of the plays between the collaborators.
The most straightforward cases of collaboration involve … John Fletcher 91579-1625). The title page of the 1634 Quarto of Two Noble Kinsmen says the play was ‘Written by the memorable Worthies of their time Mr John Fletcher and Mr William Shakespeare Gent[lemen]… [ ] (Craig 2012, p. 26)
This play, The Two Noble Kinsmen, says Craig,
does not appear in the First Folio. A series of tests such as metre, the use of contractions, and vocabulary converge on a division of the play agreed on by modern scholars. There is also general agreement the Henry VIII is a collaboration between the two playwrights; the division of the play proposed by Spedding in 1850 stands up well to modern testing. The relative ease with which the divisions are established suggests that the two men generally worked on separate sections of the play, rather than jointly writing scenes or acts. It seems likely that there was a third Shakespeare-Fletcher collaboration, Cardenio [  ]… which was published in an English translation in 1612. A play of this name was performed twice at court in 1613, although there is no evidence of publication. (Craig 2012, p. 26)
Professor Craig says, “It now seems clear that Shakespeare worked as an anonymous collaborator on plays early in his career”(Craig 2012, p. 27). Hugh Craig then lists works which have now been proven by scholars as collaborations, and other works in which Shakespeare may have had some hand, and other works that are still in doubt, or which are surrounded by confusion.
He may well have contributed a section, but only a section, to the Raigne of Edward III, which was printed in 1595 but seems to have been performed earlier. Timothy Irish Watt has recently … augmented the case for Shakespeare’s part-authorship. The three plays of Henry VI are dated to this period also. Confusion surrounds their authorship, however, and even the order in which they were written. Versions of part 2 and part 3 were published, without any indication of authorship, in 1594 and 1595. All three were included in the First Folio, so there is a prima facie case that Shakespeare was involved, but there is no agreement on how much, or about who his collaborators were ….
Since the middle of the seventeenth century there has been a persistent strand of commentary linking Shakespeare with the anonymous play Arden of Faversham, first printed in 1592. Claims that the play is entirely by Shakespeare have been refined to suggestions that only some sections are his. There is by no means consensus … but there are some strong connections with known Shakespeare in terms of style and imagery, confirmed by quantitative work in stylistics. (Craig 2012, p. 27)
Hugh Craig also notes other works in which Shakespeare may have had a hand:
Shakespeare may well have written the series of additions to Thomas Kidd’s pioneering revenge play The Spanish Tragedy which were published in the 1602 edition. In this case the external evidence points to Ben Jonson a the writer. Payments to Jonson for revisions to the play are listed in the diary of the theatre manager Philip Henslowe in 1601, but the additions include speeches of whimsical, ironical mental instability quite unlike anything in Jonson. Coleridge thought they were Shakespeare’s work, and they share an unusual words and phrases with Shakespeare plays and poems. A statistical analysis of patterns of word use, both function words and lexical words, supports the attribution to Shakespeare. (Craig 2012, p. 28)
As well, Craig points out that there are poems that have been forward for inclusion in the Shakespeare canon, but which fall short of being included because the evidence presented does not stand up:
In the late 1990s the claim that ‘Shakespeare’ should be extended to include a 1612 funeral elegy for William Peter, a little-known Devonshire gentleman, renewed the debate about the role of internal evidence in attribution. Donald W. Foster, the main proponent of the attribution, presented extensive data in a 1989 book showing that the poem fitted well within the ‘Shakespeare’ range on a number of linguistic markers such as the frequency of some common words and the frequency of some figures of speech. Foster says in fact he was unable to find a Shakespeare test that the Elegy would not pass. He summed up the case thus: the poem ‘belongs hereafter with Shakespeare’s poems and plays …because it is formed from textual and linguistic fabric indistinguishable from that of canonical Shakespeare’. The only external evidence of any substance was the appearance of the initials ‘W. S.’ on the dedication to the poem. Most readers, meanwhile, agreed that the poem was laboured and dull, and saw no obvious connections with ‘Shakespeare’ in style, theme, or artistic stance. (Craig 2012, p. 28)
Professor Craig: “Richard Abrams, a second proponent of the attribution, responded that given the strength of the evidence for the inclusion of the poem in the canon, the rest of Shakespeare’s works would just have to be read differently from now on” (Craig 2012, p. 28).
Craig advises that argument over whether the Elegy was Shakespeare or not, went back and forth: on one hand “doubters had had to rely on their impressions that the Elegy’s style was ‘un-Shakespearean’, and these were increasingly discounted”; on the other hand “the poem began to appear in American Shakespeare editions like Norton and the Riverside”; and then “the momentum was abruptly reversed in 2002 …when G. D. Monsarrat published strong evidence in favour of another candidate, John Ford–mainly words and phrases in common with Ford poems written about the same time. Foster and Adams conceded shortly afterwards” (Craig 2012, p. 28).
He points out that “This case illustrates some important methodological inclusions” (Craig 2012, p. p. 28-29):
Where an attribution relies on internal evidence, and connections are relative and comparative, one author may be the most likely candidate from those tested, but there is always the possibility of a new author from outside the set being stronger still. Once that author is included, or just taken seriously–Ford was in Foster’s original control set, but not given anything like the same attention as Shakespeare–the claims of the first author look much less conclusive.
There was a precursor to the Elegy episode in a controversy over a much shorter untitled poem beginning ‘Shall I die? Shall I fly?’. Gary Taylor, one of the editors of The Oxford Shakespeare, found the poem in the Bodleian Library in Oxford. The catalogue attributed the poem to Shakespeare. The poem was included in the 1987 Oxford edition, but its evident clumsiness and derivativeness and the paucity of persuasive parallels to ‘Shakespeare’ kept it out of the canon in any more general sense. (Craig 2012, p. 29)
Hugh Craig says that these types of cases are not altogether unknown.
A considerable number of anonymous and even well-attributed plays have been proposed for inclusion in the Shakespeare canon as wholly or partly by him. Like the apocryphal books in the Bible, they form a penumbra to the canonical works. One instance is Edmond Ironside, a late sixteenth-century history play which has survived in manuscript. E. B. Everitt put a case for Shakespeare’s authorship in a 1954 book, mainly on the basis of verbal parallels with early canonical Shakespeare, and subsequently Eric Sams made is own arguments for the idea in a book and series of articles in the 1980s, again on the basis of internal evidence from vocabulary. Most other scholars reject the attribution. (Craig 2012, p. 29)
In the conclusion to his chapter Craig says: “Because so much is at stake, Shakespeare authorship throws the methods for arriving at the truth in a range of questions into extraordinary relief… there is no doubt that some fine intellects have given their best efforts in the quest to resolve some of the perplexing questions that arise” (Craig 2012, p. 30).:
It is also worth noting that after several centuries of endeavour there has really been not so very much added, nor much taken away, from the first monument of Shakespeare authorship–the Folio volume of 1623, with its thirty-six plays presented as the work of a fondly remembered friend and colleague dead not seven years before. (Craig 2012, p. 30)
Craig says, “No one has succeeded in ruling any of the Folio plays out of a collected Shakespeare: our best understanding is that he was involved in the majority of them as sole playwright, and in a minority as collaborator, or as the author of an earlier version later revised or supplemented by another” (Craig 2012, p. 30).
Moreover, “no whole play has been added to the canon, though there are a series of parts of other plays that can now be attributed to him …” (Craig 2012, p. 30).
The Folio also remains the best guide to Shakespeare authorship in terms of the identity of the author, and the idea of authorship itself. William Shakespeare is in the title, is presented throughout as the author, and is identified through the dedication, preface and commendatory poems with William Shakespeare the actor and King’s Men shareholder, born in Stratford-upon-Avon. (Craig 2012, p. 30)
Further to this, Professor Hugh Craig advises that “No challenge to these straightforward links between this individual,” William Shakespeare author and playwright from Stratford-upon-Avon, “and these works has been sustained” (Craig 2012, p. 30):
In the Folio volume Shakespeare appears neither as a solitary genius creating in isolation, nor as a mere functionary in a larger productive enterprise. He is presented simply as an exceptional theatre professional, admired by his peers both for his collegial ties and for his extraordinary talent. (Craig 2012, p. 30)
[  ] In his chapter, Craig footnotes these plays “in order of composition given in Stanley Wells and Gary Taylor in William Shakespeare: A Textual Companion (Oxford: Clarendon Press, 1987): Two Gentleman of Verona, Taming of the Shrew, Richard III, Comedy of Errors, Love’s Labour’s Lost, Midsummer Night’s Dream, Romeo and Juliet, Richard II, King John, Merchant of Venice, Henry IV Part 1, Merry Wives of Windsor, Henry IV Part 2, Much Ado About Nothing, Henry V, Julius Caesar, As You Like It, Hamlet, Twelfth Night, Troilus and Cressida, Othello, All’s Well That Ends Well, King Lear, Antony and Cleopatra, Coriolanus, Winter’s Tale, Cymbeline, and Tempest.”
 Not my italics here.
 I referred to this play earlier in this blog about a different matter.
Final image of Shakespeare in this post is by Samuel Cousins, and was copied from the net from a page displaying varies images of Shakespeare.
Craig, Hugh, and Arthur F, Kinney, eds. Shakespeare, Computers, and the Mystery of Authorship. Cambridge, UK: Cambridge University Press, 2009. Print.
Craig, Hugh. “Authorship.” In The Oxford Handbook of Shakespeare. Ed. Arthur F. Kinney. Oxford: Oxford University Press, 2012. p.p. 15-30. Print.