Shakespeare’s Sonnets: Half of them Dated – Part 1

To view or download this article as a PDF click here.

I. Problems of Dating Shakespeare in General

In 1926, a little-known article was published by Elizabeth Beckwith1 to present her approach to dating some 52 of the Shakespeare Sonnets. Gathering her data from Sir Sidney Lee’s 1901 and 1905 works, for each selected sonnet she would find a line or two which echoed similar lines in a play, and then adopt the date of authorship of that play as a suggested date of composition of the sonnet. Important to her method was her contention that authors “often use a phrase twice and then forget it” (Beckwith, 1926, pp. 229, 242), as she would show Shakespeare did in selecting her echoes. Her methodology was ingenious, but much of its validity depended on whether she had a solid dating of the plays, and as we know, even today there are differences of opinion about dating individual plays.2

I have long argued that it’s difficult to assign a ‘date written’ to a literary work, because there are milestones or ‘phases’ for each work, ranging from conception (or ‘origination’), revisions, drafts circulated among friends, preliminary rehearsals or private performances, subsequent revamping and edits by others (such as by actors!), to first public performances, first publications, and even subsequent publications, with each version potentially differing greatly from others. Some plays and poems may have originated as the author’s juvenile exercises (‘juvenilia’) while at school, and may have been reworked over and over for decades, even circulated ‘among his private friends,’ such that the extant texts may have come via copies that friends had in some ways modified or corrected. Others may have been influenced by one principal source early on, and then as the author’s voracious reading continued to other sources, even sources for the principal source (e.g., hypothetically, after someone had written the earliest version of Hamlet, Shakespeare might have read Saxo Grammaticus’ 12th century source for Belleforest’s 1570s source for Hamlet, and then applied modifications from Saxo directly and other topical allusions to yield 1604 Hamlet ‘good’ Q2, particularly with regard to the pirate attack and Wittenburg scholars.) And if the author was a polymath, adept in French, Italian, Latin, and possibly other language – wouldn’t he have been able to access original language sources rather than wait for the first English translations? Especially notable, drama is a collaborative process where a play originating in the 1570s to early 1580s, perhaps for Court performances or for private entertainments, might be modified due to the circumstances of each venue, later re- adapted by other companies for other venues, for other patrons, and a welter of modifications and additions by unknown hands may have been made (including by actors), followed by adaptations in the 1590s and beyond, and all of this likely having been beyond the control of the originating author(s) over a period of decades. Finally, and most worrisome, the late Elizabethan stage world was immensely collaborative, with the famous Henslowe’s Diary showing numerous examples of playwrights being paid to write works together or adapt older ones, even collaborate on sequels to earlier successes, with those works often published with no author’s name, if published at all.3

In the case of Shakespeare, nearly all of his plays published up to 1604 had ‘old plays’ that were precursors, often with evidence they had been earlier performed in private venues or by rival play companies to those which Shakespeare had been associated. This was particularly true of the 12 plays listed by Francis Meres in his 1598 Paladis Tamia,4 which also happened to include the first ever mention of Shakespeare’s “sugared sonnets among his private friends …”. Four books (Ellis, 2012; Knutson, 2001; Marino, 2011; and Schoone-Jongen, 2008) are admirable examinations of many of the issues alluded to here with regard to difficulty of dating Shakespeare’s works, and attribution of works to him (e.g., prior to Meres’ list, all Shakespeare plays were anonymous, and the first titlepage to ever have Shakespeare’s name on it was 1598 LLL ‘good’ Q1).

Indeed, Marino’s general opinion (e.g., 2011, pp. 1-17) was that the name ‘Shakespeare’ on literary works was essentially less about actual authorship than it was for locking down ownership of the best texts in their inventory, used by the Lord Chamberlain’s Men under patronage of the 1st and 2nd Lords Hunsdon (‘Shakespeare’s company,’ after 1603 the King’s Men). Many of their titles may have been acted under earlier companies such as Strange’s Men, Derby’s Men, Pembroke’s Men, the Queen’s Men, Oxford’s Men, and even possibly as far back as the 1570s to early-1580s Lord Chamberlain’s Men under patronage of the 3rd Earl of Sussex, or the Boy’s Companies for acting at Blackfriars or at Court. To put it simply, the standard dating system used by most academics today, essentially what was suggested by Chambers in 1930, is a gross oversimplification at best, or a misleading fantasy at worst.5 We could go on with much more discussion about the pitfalls of dating Shakespeare’s works in general, but when we deal with the 1609 Sonnets, it might be worth starting with the question of whether Shakespeare was still living in 1609.

II. Posthumous Sonnets Project

In 2009, in response to the 400th anniversary of the 1609 Sonnets, the President of the Shakespeare Oxford Society (now Shakespeare Oxford Fellowship (SOF)), Matthew Cossolotto, initiated a ‘Posthumous Sonnets Project,’ to gather evidence that the author of the Sonnets was actually dead in 1609 (or at least perhaps profoundly disassociated with his works?). The most comprehensive of his several articles in that year was a ‘Letter to the Editor’ of The Oxfordian (2009, pp. 4-10). The list of evidence he had gathered at that point, in my summary from Cossolotto’s several articles, included:

  1. Shakespeare scholars fail to agree on many questions, including how the Sonnets came to be published in the first place (e.g., were they ‘pirated’ and never intended at all by the author to be made public?) and whether Shakespeare authorized the publication (i.e., were they published without his knowledge and consent?). Scholars tend to agree, but are still unable to explain, the apparent absence of the poet in the publication and proofreading process (i.e., the 1609 Sonnets were shot full of mis-wording, oddly placed punctuation, and other errors not corrected until 1640).
  2. The 1609 Sonnets does not include a dedication from the poet. If the author was alive in 1609, why didn’t he write a dedication to a volume of poetry that he believed would be eternal (per sonnet # 107’s line 11, “Since spite of (Death) I’ll live in this poor rhyme”)?
  3. If the 1609 Sonnets had been pirated unlawfully, and if he was still alive at the time, why wouldn’t Shakespeare have complained or asserted his legal right to the poems? Because there are only 14 extant copies, there is a possibility that the 1609 Sonnets was a failed project, and even may have been recalled by authorities shortly after initial publication. Meres mentioned them in 1598, and two of them were published in 1599 The Passionate Pilgrim (TPP), but even after their 1609 publication, they were not quoted or publicly acknowledged until the corrected publication of 146 in the 1640 Poems by Mr. Wm. Shake-speare.
  4. Following the publication of the Sonnets in 1609, William Shakespeare of Stratford made no documents respecting anything vaguely literary, even having no mention of manuscripts or books in his 1616 will. Subsequent records of him in presumed retirement in Stratford show only financial and real-estate dealings – some scholars have even suggested that his palsied signatures may indicate that he was mentally afflicted or disabled.
  5. The title of the 1609 book – SHAKE-SPEARES SONNETS – is itself unusual. First, the gratuitously hyphenated (i.e., not forced by an end-of-line) name could mean it was a recognized pseudonym. For example, ‘Martin Mar-prelate,’ or ‘Tom Tell-troth’ were certainly pseudonyms, and the hyphenated ‘Shake-speare’ often continued to be used even up to the last two dedications (by Leonard Digges and ‘I.M.’) in 1623 F1. Second, the title implies finality, as if the 154 sonnets in that collection were the end of the line for Shakespeare’s sonnet-writing career. If Shakespeare were still alive in 1609, why would the publisher suggest by this title that the poet would write no more sonnets?
  6. The publisher’s dedication to 1609 Sonnets referred to “our ever-living poet,” strongly implying that the poet was dead, or immortal (there may have been one or two uses of ‘ever-living’ for living authors, but the vast number of examples were for dead ones). Several sonnets imply the poet was nearing his death and expected his name to be forgotten, or ‘buried’ after his death (e.g., # 72’s “My name be buried where my body is, / And live no more to shame nor me nor you.”) This made no sense if the poet’s real name was Shakespeare, a name extremely famous in 1609. I add that the ‘funereal arrangement’ of the words in the dedication (i.e., dots between each word, as in carvings on tombstones) may also show the author was dead.

These are only the beginnings of Cossolotto’s initiative, which has continued since 2009. So, for the purposes of this article, we’ll assume that the 154 sonnets were all written by a year prior to 1609, earlier perhaps by 5 years or more. Because of the manifest differences in the texts of 1609 Sonnets vs. 1640 Poems (there were 154 sonnets in 1609 and only 146 in 1640), we’ll also assume that the 1640 project had its roots in an MS that was earlier and separate from the one used in 1609. Whereas the 1609 project has many reasons to assume the author was dead, the publisher of 1640 Poems, John Benson, famously stated in his dedication to the Reader that the collection he had prepared had works which (underline added): “…appeare of the same purity, the Authour himselfe then living avouched; they had not the fortune by reason of their Infancie in his death, to have the due accomodation of proportionable glory, with the rest of his everliving Workes…” Note that we should take this claim of Benson’s to apply particularly to the 146 sonnets it featured, as much of the rest of the poetry attributed to Shakespeare in 1640 Poems derived directly from the 1612 3rd edition of TPP, which was disputed by Thomas Heywood’s claim that the two long poems in it were by him rather than the Bard.6

In my opinion, the division of the MS of 146 sonnets carefully preserved for use by the 1640 project (versus that of 154 sonnets sloppily used by the 1609 project) dates back to circa 1601-04. Thus, it’s likely that fewer sonnets than 146 existed in 1598 when Meres noted them. Shakespeare then went on to write at least 8 more sonnets, some that may now be lost, and the survivors are the extra 8 sonnets in the 1609 collection that weren’t in the 1640 Poems. The separate paths of two separate MS versions of the sonnets helps explain why the 1609 texts were so adulterated, why the 1640 texts were not, and why the 1640 texts in general are what are used today for modern Sonnets renditions.

Let’s examine the question “Why did Shakespeare aim to produce about 150-or-so sonnets?” The answer that seems closest to truth is that there were precisely 150 Psalms in the Old Testament, poems ascribed to King David. As it so happens, in 1599 TPP, the first two sonnets were later found in 1604 Sonnets (#s 138 & 144) and 1640 Poems, plus two more sonnets in TPP had been included as incidental poetry imbedded in Shakespeare’s 1598 good quarto of Love’s Labour’s Lost (LLL), along with a fifth poem which may have been adapted from a sonnet.7 Adding-in those four or five poems undeniably by the Bard to the 146 later in 1640 Poems comes right close to 150 (particularly if we count some of the poetry in Sonnets to Sundry Notes of Music, which was in TPP 2nd & 3rd eds.). That does seem to be more than coincidental, considering the 150 Psalms. And if more than coincidental, doesn’t it mean that much of the layout for 1640 Poems was Shakespeare’s intent?

Go to Part Part II DVS Newsletter, January 2017


1 See Beckwith (1926). Following her signature to the article, Beckwith signed ‘Cambridge, Mass.’ An Elizabeth Beckwith is listed ( in the Harvard University Library’s ‘OASIS’ system, with the following entry for Radcliffe College: ‘Series 1, Box 9. Beckwith, Elizabeth Ann (Mrs. B. M. Nicholds), AB New York University ‘21, AM’ 25, 1923-1925, n.d.’ Thus, she seems likely to have been the more famous Elizabeth Nicholds who published in 1933 the celebrated farming memoire, Thunder Hill, followed in 1934 with the pamphlet The Goats of Thunder Hill, Forsaking City Streets for a Rural New York Retreat, Four Women Brave the Complications of a Cowless Dairy Farm. Although she didn’t receive a Ph.D. from Radcliffe, she apparently kept writing, because a 1960 book A Primer of Social Case Work is credited to an Elizabeth Nicholds, living in Ohio, but published by Columbia University Press in NY City. Those three publications are listed by searching under her name on under ‘books,’ from which I bought my copies. If for some reason you’re having trouble locating Beckwith’s 1926 article, contact me by e-mail ( and I’ll happily send you a transcript I’ve made of it.

2 Hess (2003), Figure B.1, pp. 298-99, features columns comparing dates assigned to each play by E.K. Chambers in 1930, F. E. Halliday in 1952, G. E. Bentley in 1961, and Riverside in 1974 (E. G. Blakemore, Ed.).’ Yes, their dating schemes are remarkably similar (derivative might be a better term), but there are differences. The column for Bentley is divided into columns for first performance and first publication, which neatly skirts the typical need to date ‘when written.’ A handy tool is Riverside’s listing (pp. 51-56) of ‘Sources’ for each play, thus providing an excellent glimpse at how hollow are the pretences that ‘dates of authorship’ for the plays can be dated with any real ‘certainty.’ For most plays, the sources come from the 1560s to 80s or even earlier, and if Shakespeare knew French and Italian, many could have come from earlier sources yet.

3 For example, on March 8, 1598/9 Henslowe paid Anthony Munday, Michael Drayton, Robert Wilson, and Richard Hathway to write Pt. 1 of Sir John Oldcastle, which was published in August 1600 with no author’s name, but republished in 1619 with Shakespeare’s name on the title-page (if a Pt. 2 ever existed, it is now lost or else folded into Henry V). The play’s relationship to Shakespeare is doubted, but it may have been a source for Henry IV Pts. 1 and 2, and for Henry V; or, as is my preference, the four authors may have simply dressed up an old MS play that was intermediary from the origination state of those 3 Shakespeare plays (1&2H4 and H5) – possibly the MS was in Shakespeare’s own hand, or possibly it was a copy of earlier MSS in a lineage going back to the 1570s to early-80s. Other examples might be the alleged collaborations of Shakespeare with John Fletcher in Henry 8 and the now-lost Cardenio, and possibly other plays, for which the chief evidences were play lists of the 1650s and 60s naming the two as collaborators. If Shakespeare ever collaborated with Fletcher, I believe it was in similar fashion to the Oldcastle project, where an old MS (possibly derived directly or indirectly from Shakespeare) was later presented to Fletcher for revamping, and then later attributed to Shakespeare long after the Bard was dead and unable to confirm or deny the credit. More to the point, Shakespeare may have been dead, or never had met Fletcher, at the time that Fletcher allegedly ‘collaborated’ on these works – where Fletcher simply ‘completed’ earlier draft plays that may or may not have involved Shakespeare. See Hess (2002), Chapter 7, pp. 313-42, for a description of the ‘Passive Collaboration’ process which I believe explains much now in dispute, where independently-made changes over decades had revisers ‘collaborating’ with prior authors unknown to them.

4 Palladis Tamia was entered in the Stationer’s Registry (i.e., copyrighted) on July 7, 1598, and is famous for unusual attention paid to works of Shakespeare, and its list (p. 282) of 12 plays (underlines added): “…for comedy, witness his Gentlemen of Verona, his Errors, his Love Labours Lost, his Love Labours Won, his Midsummer’s Night Dream, & his Merchant of Venice; for tragedy, his Richard the 2, Richard the 3, Henry the 4, King John, Titus Andronicus, and his Romeo and Juliet…” Notice that “Henry the 4” covers two extant plays (1H4 & 2H4) that were included in the 1623 First Folio (F1). There is some debate that an MS Thomas of Woodstock play was the 1st part of a 2-part Richard 2 play (see (play)), and possibly Meres referred to that combo. Scholars debate if Love Labours Won was an earlier name for The Tempest (a title never heard of until it appeared in F1), another extant comedy, or was a now-lost play; yet it did in fact exist, since in Aug. 1603 a play list had that title for sale. Scholars debate whether any play that didn’t get into Meres’ list hadn’t yet been written in 1598; yet, Meres was clearly using 6 Comedies and 6 Tragedies (including some Histories) to compare with Greco-Roman comedy and tragedy writers, so there may have been many others in 1598, or even earlier, that Meres just didn’t use in his comparison exercise.

5 Riverside (1994)’s listing (pp. 51-56) of ‘Sources’ for each play revealed that for some plays “style elements” or ‘stylistics’ were employed, probably to bridge the great gulf between the early dates of sources and the first dates of known public performances or publication. These included: All’s Well (54), Timon (55), Coriolanus (55), and Cymbeline (56), which doesn’t seem like much, except that other plays were then dated in relation to these, and others in relation to those. Meanwhile, Halliday (1952, pp. 203-204) had endorsed use of the ‘Feminine Endings’ style element “for establishing the chronology (of plays it) is one of the most helpful of the verse-tests.” And yet, in the same book, after various discussions of style endings, Halliday admitted (p. 680), “The evidence of verse tests is treated less seriously today”. These matters were discussed in Hess(1999, p. 27), which attempted to confront the 1990s revival of ‘stylistics’ by demonstrating that modern stylists are suspect or wrong in much of their methodology and conclusions. In the 1940s ‘stylistics’ were discredited, and remained so for half a century, only to be revived in concert with attempts at ‘computer-aided’ analysis. Such analysis involves setting up a computer ‘black box’ with magical rules and attributes inside, poorly described to the general public, or even to other ‘experts.’ The process generates results which have the aura of veracity because ‘they are computer generated,’ as if computers have never been wrong or misused. And when we’re through with ‘stylistics,’ there are other serious problems that we have no space to discuss here, such as whether or not Henslowe’s use of ‘ne’ in the margin of his Diary meant ‘new’ or just his playhouse at ‘Newington Butts;’ and the issue of whether Meres’ failure to list a play meant it didn’t exist in 1598 yet. It’s obvious why we should be skeptical about modern wizardry concerning ‘stylistics,’ and about the standard dating regime which depended on it to a surprising extent.

6 Heywood’s complaint about 1612 TPP was discussed at length in Hess (2014) in JLAS. There were four major conclusions: (1) We don’t know the date of the TPP 1st ed., because unlike the 1599 date on the 2nd ed’s title-page, the sole extant copy of the 1st ed. is missing a title-page and some of the poetry. Hess-2014 proposed a date of circa 1594-1599 for TPP 1st ed. based on the early dating for most of the poems contained within it, plus the earliest evidence of William Jaggard’s printing career. (2) Heywood’s dispute, or complaint, was on the last page of his 1612 Apology for Actors, where he claimed that he brought the matter of Jaggard’s misuse of two of Heywood’s poems (first published in 1609 Troia Britanica) to the attention of Shakespeare, who was shocked at Jaggard having mis-attributed the poems. In any case, nearly every line in that complaint is garbled by what appeared to be Heywood’s hurried amendments made to it. For one thing, Shakespeare was likely not alive (or was effectively retired) in 1612, as we’ve seen demonstrated above. (3) Hess (2014) proposed a reinterpretation of Heywood’s complaint to mean that the Jaggard mischief had actually occurred for the ‘by-1599’ TPP 1st ed., not the 1612 3rd ed. If so, Heywood’s complaint had originally been prepared for that earlier occasion, set aside after Jaggard removed the offending two poems in the 1599 2nd ed., then hastily amended for reuse in 1612. In the interval, Shakespeare was either dead or effectively retired, and Jaggard had gone from a newly-begun printer with little influence in circa 1594-99 to become one of the great printing powerhouses by 1608 and later, particularly after he had been named Official Printer for London City by the Bishop of London (Jaggard’s shop would go on to print the 1623 F1 project). In any case, in 1612 Jaggard likely had little to fear from Heywood’s accusations, so he reinserted the disputed two poems into his 1612 project, a sort of thumb in Heywood’s eye for what had transpired nearly 2 decades earlier about the 1st ed. (4) By comparing Leonard Digges’ (1558-1635) dedication used in 1640 Poems with the one he contributed to 1623 F1, Hess (2014) suggested that the two dedications were two parts of one long poem prepared over a decade before F1 was published. Where the short dedication in F1 oddly had more to do with Shakespeare as a poet, the longer dedication in Poems had more to do with Shakespeare and Ben Jonson as dramatists, quite the opposite of what we should expect. Nearly all of the Jonson drama alluded to dates in the 1610- 1612 timeframe. Thus, Hess-2014 suggested that Digges’ two-part poem had been prepared for a circa-1612 grand project to publish all of Sh’s plays and poetry combined—or whatever was not otherwise copyrighted by other stationers—and present it all in one huge project. That project fell apart for a number of possible reasons (e.g., the 1612 death of Prince Henry, who may have been the patron of the Grand Project), leading to F1 and Poems having separate publication paths. But it does serve to explain why Jaggard published the 3rd ed. of TPP in 1612, because that was the part of Shakespeare’s poetry to which he already had full title, and he could publish it without negotiating and sharing profits with other stationers.

7 Recall, 1598 Love’s Labour’s Lost (LLL) good Q1 was the first-ever play to have Shakespeare’s name associated with it. The poetry in “by-1599” The Passionate Pilgrim (TPP) Q1 and its 1599 Q2 was discussed in Hess (2014) (in JLAS). In addition to close versions of sonnet #s 138 & 144, the 3 other clearly Shakespeare sonnets or poems which TPP contained were from 1598 LLL: (IViii) “Did not the heavenly…,” (IVii) “If love make me forsworn…,” and (IViii) “On a day (alack the day)”. The last was an 18-line non-sonnet (in LLL it has an additional two lines for 20 total), but contains within it what may have once been a sonnet, or could have been intended to be reworked into a sonnet, and it is often listed under Sonnets to Sundry Notes of Music in modern scholars collected works of Shakespeare. These five undeniable works of Shakespeare are the only parts of TPP which modern scholars grudgingly allow were by the Bard, although Hess (2014) argued that each of the rest could indeed have been by Shakespeare if we follow the simple principle that Deloney, Barnfield, Griffin, Marlowe, and Raleigh were more likely to have ‘borrowed’ from Shakespeare rather than vice versa.

Go to Part Part II DVS Newsletter, January 2017

To view or download this article as a PDF click here.