Shakespeare’s Sonnets: Half of them Dated (Part 2)

In the October 2016 newsletter, we presented the first half of this article, in which W. Ron Hess discussed the difficulties in dating the plays and sonnets, particularly in light of the creative process of playwriting. He made passing reference to Elizabeth Beckwith’s ingenious method of dating the sonnets, which is discussed in more detail below, along with his own view of when the sonnets were written. (This article is a modification by Hess of his article ‘Shakespeare’s Sonnets Dated’ from the Nov. 2015 Journal of Literature and Arts Studies (JLAS).)

I. Beckwith’s 1926 Article

Though her article was short (16 pages), Elizabeth Beckwith packed in a good deal of worthy discussion on the dating of the sonnets. She began with Meres’ notice of the “sugred Sonnets among his private friends,” and noted that the 1599 2nd edition of The Passionate Pilgrim (TPP) had in it close equivalents to sonnets 138 and 144 in the 1609 Shake-speare’s Sonnets project. She noted that Meres’ 1598 notice did not suggest that the sonnets cycle was complete in 1598. (I believe at least 8 of them were yet to be written.)

In her second and third paragraphs, Beckwith noted that Sir Sidney Lee had “practically disposed of the guesswork concerning the ‘Mr. W. H.” of Thorpe’s dedication by noting the existence of printer William Hall (note that the dedication actually read “Mr. W. H. ALL.,” effectively spelling out Hall’s name!). And Lee had gone on to make “an exhaustive study of the Elizabethan sonnet as a type … showing the completely conventional character of most of Shakespeare’s sonnets,” thereby rendering “their autobiographical quality even more dubious.” I would go further by comparing Shakespeare’s sonnets to French and Italian cycles, where the “conventional character” would be even more emphasized.

Beckwith proceeded to note that in 1913, Raymond Alden had “proved with finality… that the [1609] arrangement of the sonnets is not Shakespeare’s, and is neither chronological, nor strictly according to subject…,” based on:

the piratical character of the publication which prohibits the possibility of Shakespeare’s final sanction of the arrangement, and the internal evidence of discontinuity which suggests that [the publisher Thomas] Thorpe received the sonnets in groups of two or three and then made a superficial attempt to arrange them by subject.

This conforms to what has been previously argued, that the Bard had nothing to do with publication or arrangement of the 1609 Sonnets project (because he was dead?).

Beckwith continued to describe “the unjustifiability of supposing either that the date of the sonnets must match some corresponding event in the poet’s private life, or that the chronology must follow the order of the [1609] edition.” (1926), proceeding to describe four general methods of determining the date of the sonnets, and showing the limitations of each:

(1) Lines which allegedly describe public events of contemporary interest (she noted 107’s “mortal moon” and showed that scholars have proposed dates of 1598, 1601, and 1603 for it;

(2) Deciding in advance that either Henry Wriothesley, Earl of Southampton, or William Herbert, Earl of Pembroke, was the “Mr. W. H.” of the sonnets dedication and then interpreting each sonnet as pertaining to the chosen ‘patron.’ She felt this approach particularly fails in terms of explaining who the ‘Dark Lady’ would have been, since identification of reasonable candidates depends on how we date the sonnets themselves (i.e., if dated 1590-1595, they point to Southampton’s scandal, but if 1597- 1601 to Pembroke’s scandal);

(3) Passages paralleled by other authors, such as Sidney’s Astrophel and Stella (pub. 1591), Daniel’s Delia (1591), Drayton’s Idea (1594) and Marlowe’s Hero and Leander (1598) where the question arises of who was the borrower;

(4) The ‘personal opinion of style.’

As it happens, I agree with almost everything Beckwith discusses up to this point, because she supports positions at which I’ve independently arrived. Having dismissed the four most popular approaches to dating the sonnets, Beckwith proceeds to describe her favored approach:

Shakespeare has a habit, to which most writers are more or less prone, leading him often to repeat a word or phrase very soon after first using it, and subsequently ignore it. (1926, p. 229)

She gave a good number of examples, but we’ll let the reader indulge themselves with studying and determining whether they’re valid or not. For this exercise, let us assume her approach is valid, and her findings significant, except that she used dates for the plays that didn’t always conform with dates Chambers would establish four years later; thus, dating of the plays becomes important for dating the sonnets. But, since we’ve established above that Chambers’ approach was likely flawed, what remains is for us to compare a few play-dating systems with Beckwith’s to throw light on our discussion as to how the sonnets might be dated. I will also suggest some additional sonnets that Chambers feels are datable independently.

To begin, Table 1 illustrates Beckwith’s dates for plays alongside the dates elsewhere proposed for the same plays.1 Table 2 illustrates a layout of Beckwith’s conclusions, when Chambers-1930 dates are used.2 And Table 3 is another way of organizing Beckwith’s conclusions graphically. Two elements stand out in Table 3, especially in comparison to Table 5:

(1) Based on Beckwith’s method, if her bottom line of 52 total is rolled out to 154 sonnets, it shows that approximately 118 of Shakespeare’s Sonnets likely existed in 1598 when Meres’ Palladis Tamia had the first historical note of the sonnets’ existence.

(2) Based on the extrapolation to 154 sonnets total, 146 sonnets are revealed as having been written in 1603-1604.

Note the oddity of Beckwith’s conclusions in Tables 2 and 3, that Shakespeare’s Sonnets under her system appear to have been mostly written at the same time as his two long poems (V&A and RofL), plus several plays. Is this credible, that the Bard would have overburdened himself in such a short period of time? Of course, her system only models a third of the whole cycle of 154, forming what she calls a ‘skeleton’ for use for further research. So perhaps we could achieve more credible results if we found ways to add a few more dated sonnets to the tally, raising it to over half of 154, and possibly showing a more credible distribution.

II. Adding on a Few More Dated Sonnets

Here we describe the rationale for choosing additional sonnets.

Recall that Beckwith listed four common methods of dating the sonnets, each of them with its limitations. One of them was described as, “Lines which allegedly describe public events of contemporary interest.” We can agree that these often lead to debate among scholars, with multiple potential theories of their meaning and dating. Yet by “contemporary interest,” it is rare that ‘foreign or specialized interest’ has been suggested. If Shakespeare was a cosmopolitan man, possibly adept in the Latinate languages, wouldn’t he have been likely to note exceptional events abroad? And if he was a man with special interests in medicine, especially a less well-accepted line of medicine, wouldn’t that have been notable, as well? Moreover, it is generally thought that the Bard deliberately blurred allusions to avoid offense to living persons, but such reticence wouldn’t have applied to foreigners and foreign events. So here are my added dating suggestions:

A. Hess (2007) argued from L. Hotson (1949) that three sonnets can be dated based on foreign affairs, and that their convergence in dating strengthens his arguments;

B. 1589 for 107 based on the “crescent-shaped” battle formation of the 1588 Armada and many predictions going back for decades that 1588 was supposed to usher-in the end of the world, or at least the fall of empires;

C. 1586-89 for 112 based on the re-erecting in Rome by Pope Sixtus V of obelisks (called “pyramids” in Elizabethan poetry) found buried in the ruins of the ancient forum; and

D. 1589 for 113 based on the assassination of King Henri III of France by a fanatical monk, with Henri having been the favorite son of the infamous Catherine de Medici (i.e., he had been a “child of state”);

E. As to the sonnet based on medical knowledge (an international concern, no doubt), Hess (2007) argued, from R.S. Brazil (2000), that 119 (‘the Apothecary sonnet’) reasonably dated to 1574-1576, based on certain books by Queen’s Physician Dr. George Baker, which argued in favor of Paracelsan medicine (using chemicals, derived of alchemy and herbalism), which is the ancestor of modern medicine and pharmacy. But in the 16th century, Paracelsans were suppressed in some quarters, and discouraged by the usual physician groups who practiced the ‘humours’ and ‘bleeding’ techniques that have today been long discredited. Brazil particularly noted the illustration on the cover of Baker’s 1576 Newe jewell (dedicated to the Countess of Oxford), which had examples of the same things described in 119. But Hess (2007) noted that the 1576 book was essentially a republication of 1574 Oleum magistrale (dedicated to the Earl of Oxford). Thus, the Paracelsan and herbalist interests of the Earl and Countess of Oxford are worth consulting, as well as the two Hess (2011) articles about Thomas Sackville as ‘Literary Mentor’ to Shakespeare, and the MS poem ‘Sackvyles Olde Age,’ (which has in it various lines concerning medical and health issues.) The arguments and comparisons are complex, so for here let’s just say it would take a book to describe them.

F. Pressing on, because 137 and 144 were each included in ‘by-1599’ TPP Q1, it seems reasonable that they date to 1597, if not earlier. Hess (2014) suggested the entire TPP collection may date to 1594-1598, depending on the usual ‘who borrowed from whom?’

G. I’m not the first to suggest that 18’s “darling buds of May” was a reference to the new dating system instituted by Pope Gregory in 1583, adjusting the calendar by 10 days, and thus allowing formerly summer buds to emerge in May.

H. Sonnets 1 to 17 are known as ‘the Procreation sonnets,’ because they urge an unwed person to wed and reproduce. The current reading of these sonnets may seem aimed at a young man, but that may simply be later adjustments made as Shakespeare began to circulate his sonnets “among his private friends” in order to disguise the identity of his original “beloved youth,” or to match the French and Italian pattern of addressing Cupid, the love god. Of course, there are many other theories about the first 17 sonnets. But when we consider that one of the key and most urgent concerns of Elizabethan society 1558 to 1583 was the need for Queen Elizabeth to marry, bear offspring, and thus cement the ‘Elizabethan Religious Settlement,’ it seems most probable that Shakespeare’s first 17 sonnets were derived from earnest pleas that the monarch do her duty for England’s sake. This was both a domestic and international concern, since during that 1558 to 1583 period, there was a long list of foreign princes willing to try their luck, from King Philip II of Spain and Archduke Charles of Austria, to less well known suitors like the Crown Prince of Sweden and the half-brother of Philip II, Don Juan of Austria. Now, in 1583 the Queen turned age 50, and shortly before that her change of life potential began looming greatly, until it was quite unrealistic for her to be considered fertile. Thus, the urgent pleas of sonnets 1 to 17 may have been associated with the period 1579- 1581, when Elizabeth had an actual marriage contract with the Duc d’AnjouAlencon. The Elizabethan court was starkly divided between those for the French marriage to Alencon and those against. At least one duel was nearly fought over the issue, the famous 1579 ‘tennis court brawl’ between Sir Philip Sidney (anti) and the Earl of Oxford (pro). If indeed this did apply to Shakespeare’s first 17 sonnets, quite a range of dates could be selected. But probably the safest would be 1581, when the marriage pact was amicably dissolved, and when Elizabeth’s fertility was near its proverbial ‘bitter end.’ 1581 would have been literally (and poetically) a simple case of ‘do or die barren.’3

I. Analysis of Adm. H. H. Holland was summarized in Hess (2003 pp. 287-289, p. 288 (item b.18)), which noted sonnet 130 parodies part of Thomas Watson’s 1582 Hekatompathia (dedicated to the Earl of Oxford) and a poetry exchange between Oxford and Sir Philip Sidney (died 1586), for which Hess opted for 1583. Holland item b.19 noted from Sir Sidney Lee up to 15 influences on various sonnets from Arthur Golding’s 1567 Metamorphoses (translated while Oxford lived with his uncle A.G. in the house of Sir Wm. Cecil), particularly noting sonnets 55 and 60, for which we’ll opt for 1570 (although reprints occurred in 1584, 1593, and 1603). Holland item b.21 noted an echo of 3H6 and sonnet 121, with an Oxford letter to Cecil in October 1584.

J. Finally, what shall we do with the 8 sonnets (18, 19, 43, 56, 75, 76, 96 and 126) that were in 1609 Sonnets but weren’t in 1640 Poems? Above, I suggested they should be considered to be sonnets not yet finished before the MS later used for 1640 had been separated from that later used for 1609, and thus I dated them to 1601-04. Certainly, it would be easier to date them if they ‘hung together’ around a common theme, perhaps oriented to ‘time.’ And yet, we’ve just suggested that 18, the first of them, can be dated to circa 1583 with the change to the Gregorian calendar abroad. It appears 19 is also consumed with exploring ‘time,’ but not the seasonal type. Then 43 is concerned with transitions from day to night. Near its end, 56 does return to the seasonal concern, as 75 does briefly at its beginning, and 76 returns but lightly to times of day towards its end. Sonnet 96 has no particular relationship to time other than the foibles of youth. And 126 does relate to time as in the approaching end of life and judgment day. So even concerning ‘time,’ there doesn’t seem to be a solid link between the eight sonnets other than that they weren’t in 1640 Poems. However, some or all of them could have ‘originated’ in the 1583 timeframe that 18 was, and then withheld until the 1609 project was underway. Still, the dating for sonnets 19, 43, 56, 75, 76, 96 and 126 just isn’t strong enough to use for this exercise, even if we do assume they could be dated to circa 1601-04. I’ve explored the possibility that 1640 Poems had only 146 sonnets because Shakespeare’s sonnets project was intended to respond in general to the 146 of the 1573-83 Diane sonnets by French Poet Laureate Philipe Desportes but haven’t yet come to firm conclusions, except that if this were true, the additional 8 appearing in 1609 would be pirated additions to what had been intended by a living Shakespeare before his pre-1609 death or retirement.

III. Conclusion

Instead of being content with Beckwith’s 52 sonnets (a third of the 154 total), by adding a few more based on their orientation toward international matters, we can increase dating to 79 sonnets, or more than half of the 154 total of 1609 Sonnets. By redating Beckwith’s sonnets based on an alternative system for the plays from that established by Chambers in 1930, we redistribute the pattern of the sonnets such that in Table 5 there is less of a pile-up in the 1591-1595 period than there was for Beckwith in Table 3. Shakespeare’s Sonnets thereby become a decades-long exercise rather than focusing on very few years.

Several of my Tables here refer to ‘Extrapolation from 154,’ which simply means dividing 154 by the grand total (in Table 3, 52 sonnets and in Table 5, 79 sonnets) to yield a percentage, which is then multiplied by other totals to yield projections for how many cumulative sonnets there might have been for any given year. For example, in Table 3 from Beckwith, we can extrapolate that the number of sonnets existing in 1598 were 118, and that circa 1603-04, there were 146 sonnets. By comparison, in Table 5 from the ‘Alternative Plays Dating,’ we extrapolate that in 1598, there were 129 sonnets, and circa 1600-01, there were 146 sonnets. These two comparisons (1598 and 146) are not very far apart from Beckwith’s and the ‘Alternative’s’ results.

The main difference in judging between Beckwith’s system (Table 3) vs. the ‘Alternative’ (Table 5) is how many sonnets were accounted for (52 vs. 79), and the percentage comparison of 154 sonnets (34% vs. 51%). So, the question is this, if we are to accept Beckwith’s invitation to set up “a skeleton around which the remaining sonnets can be safely built,” is it better that the skeleton be limited to only 34% of the body of sonnets or that it be expanded to 51%? Assuming our rationale for dating each sonnet used is sound in each method, wouldn’t the larger number in the ‘skeleton’ be better?

Another difference is the odd distribution yielded for Beckwith’s ‘skeleton,’ with her system wanting us to believe (Table 3) that 115 sonnets were written 1591-1595, during which time Shakespeare also produced his two long poems (V&A and RofL), plus assorted plays, and the plague was rampant, to boot. By contrast, Table 5 shows only about 30 extrapolated sonnets were written during essentially the same period (1592- 1595), while sonnets were being written in the 1570s and 1580s, and were more reasonably distributed throughout, even with major peaks and troughs along the way.

Beckwith was working on the problem back in 1926, with only bulky concordances, Lee’s books and so forth, to draw on. Her initiative was impressive and her innovative dating method for the sonnets deserves more attention than it has had in the past 90 years. Even though her dates for plays deviated from the system formalized by Chambers in 1930, she still came up with a close match. And thus, the peculiar distribution that she generated (i.e., Table 3) can be interpreted as a failure of the Chambers dating system for the plays. And that is the most important conclusion that we can make!




For Table 4 and Table 5, E, Beckwith’s data is to be applied to an Alternate Dating regime from Hess (2003) (pp. 298-99, the first column of dates labelled “Hess 2002”). Because that column gives an “origination phase” for each play, followed by key ‘revisions’ periods for each, we essentially use the last revision date for this exercise. Of course, that has limitations, since the last revision date might be for fairly minor changes vs. major ones done earlier. But as long as we know what choices we’ve made, we at least can avoid the pretence that Shakespeare wrote a complex play all in a few weeks or even a single year. Moreover, Tables 4 and 5 feature more datable sonnets than those that Beckwith’s process generated. Above, Section IV described the rationale for each additional sonnet, nearly all of them relating to Shakespeare’s awareness of international affairs of his time. This of course mixes in a different rationale among Beckwith’s findings. And to a certain extent it’s like mixing ‘apples and oranges.’ Still, if Beckwith’s method is largely valid, and the method(s) for choosing additional sonnets is largely valid, why shouldn’t combining them be valid?






  1. In addition to the dating systems featured in endnote 2 above, in Table 1 a column has been inserted for comparison with Gilvary-2010, but not used in this article. Gilvary’s was a massive dating effort for the plays that had been underway for at least a decade. Baconian Vincent Mooney Jr., until his death in 1998, taunted Oxfordians that their theory failed to produce a dating system for Shakespeare’s plays that would have no play written after 1604. As a result, Hess began his own research independently the De Vere Society in England, drawing participants from around the world. It took over a decade and the tremendous efforts of Kevin Gilvary to complete the project. An early state of that project was reported in Hess-2003, pp. 298-99, as ‘De Vere Soc. 2000’ (about half the plays were left undated from 2000 until Gilvary inherited the project), so those early findings were considered in the Hess-2003 effort. The participants in the 2000-2010 effort were familiar with the Hess-1999’s dating regime, and Gilvary’s results list the Hess-1999 dates. The Hess-1999 System of dates was not pretending to be definitive or even well-considered. But it did insert an important concept that Gilvary’s results also show, that there should be a ‘likeliest date’ inside of a ‘range’ of possible dates, rather than pretending that all phases of writing a complex literary work could be accomplished in a single year. Unfortunately, Gilvary’s results seem to have taken that goal of a range far too literally to make themselves useful – in most plays there’s a beginning-of-range date in the 1560s to 80s followed by a 1590s or later end-of-range date, which tends to ignore the ‘likeliest date’ concept. This shows that even the best of dating regimes have their flaws.
  2. Near the end of Beckwith’s article (pp. 240-241) she listed her conclusions. She gave a sequence of dates (1591- 1607) with one or more sonnet in Roman numerals to the right of each date. Her well-conceived article might have gotten better response if she or her editor had transferred that data into a chart, similar to Table 2 here. On the other hand, it would have made it easier to see the oddity of her conclusions, the clumping of so many sonnets into a very fewyears.
  3. Hess (2011) was about Thomas Sackville as Shakespeare’s “Literary Mentor” and presented an alternative theory that Sackville’s lost sonnets cycle, praised by Skelton in 1560, may be part of a core of sonnets adapted into what became Shakespeare’s 1609 Sonnets. If so, note that Sackville was the Queen’s 2nd cousin through Boleyn blood, and apparently in the 1560s it had been suggested to her that she should consider marrying Sackville, who was 4 years her junior. This is implied by her response, which was a curious backhanded slap at her cousin, for she publicly announced that were she to consider marrying, she would only wed Sackville’s father (her Chancellor of the Exchequer, then in his 70s, and her 1st cousin)! There is also the tradition that Shakespeare “would have been a king had he not played one on the stage”. Sackville and Thomas Norton penned the revenge play Gorboduc for performance at Lincoln’s Inn and at Court, and it was later published without their authorization in 1565, the plot somewhat similar to that of King Lear, with a kingly father having to deal with three sons whom he had made kings, but who then rebelled against him. If Sackville had acted as one of the four kings in his own play, and had close associations with Shakespeare, he may well have inspired such a tradition. Another interesting insight into the “French Marriage” issue is Hess (2009), about the wooing of “Avisa” (i.e., Queen Elizabeth) by a panoply of suitors in 1594 Willobie His Avisa (WHA). There I suggest dating the “origination” of WHA to 1581-1583, between the end of the “French marriage” contract and up to the death of the Duc d’Anjou-Alencon, who is modeled as “the French suitor” in WHA.


Beckwith, E. (1926). On the chronology of Shakespeare’s sonnets. The Journal of English and Germanic Philology, 25, 2, 227-242. Champaign, Illinois: University of Illinois Press. Retrieved from

Bentley, G. E. (1961). Shakespeare, a biographical handbook. New Haven, Connecticut: Yale University Press.

Brazil, R. S. (2000). The true story of the Shakespeare publications, Vol. I: Edward de Vere and the Shakespeare Printers. Retrieved from

Chambers, E. K. (1930). William Shakespeare: A study of facts and problems. Oxford University: Clarendon Press.

Cossolotto, M. (October 2009). Letters to the Editor: The Sonnets at Four Hundred, The Oxfordian (TOX), 11, 4-10 Retrieved from HERE. Two earlier 2009 articles by Cossolotto on this subject are (Retrieved from here and here).

Ellis, D (2012). The truth about William Shakespeare, fact, fiction and modern biographies, Edinburgh: Edinburgh University Press.

Evans, G. B. (Ed.). (1974). Riverside Shakespeare. Boston: Houghton Mifflin.

Gilvary,K. (Ed.). (2010). Dating Shakespeare’s plays: A critical review of the evidence, Kent: Parapress, ISBN 978-1-898594-86-4. Retrieved from

Halliday, F. E. (1952). A Shakespeare companion 1550-1950. London: Gerald Duckworth Co.

Hess, W. R. (Oct 1999). ‘Shakespeare’s dates: Effects on stylistic analysis’ (with Bloch, Howard W. and Chow, Winston C.). The Oxfordian (TOX), II, 25-59. Retrieved from here.

Hess, W. R. (2002). The dark side of Shakespeare, Vol. I of III: An iron-fisted romantic in England’s most perilous times, Lincoln, NE: Writers Club Press, ISBN # 0-595-24777-6.

Hess, W. R. (2003). The dark side of Shakespeare, Vol. II of III: An Elizabethan courtier, diplomat, spymaster, & epic hero, Lincoln, NE: Writers Club Press. ISBN # 0-595-29390-5.

Hess, W. R. (Spring 2007). ‘Searching Under the Lamp-posts for Dating Shakespeare’s Sonnets.’ Shakespeare Oxford Society Newsletter, 43 (2), 14-20. Retrieved from here.

Hess, W. R. (November 2009). ‘Discussing 1594 Willobie his Avisa.’ Hess webpage article #14. Retrieved from here.

Hess, W. R. (March 2011). ‘Did Thomas Sackville influence Shake-speare’s sonnets?’ DeVere Society Newsletter, 18(1), 21-30. (Abstracted from Hess Sept 2008 webpage article #11). Retrieved from here.

Hess, W. R. (September 2011). Sackvyles Olde age. Hess webpage article #15 Retrieved from here.

Hess, W. R. (July 2012). ‘Oxfordian Musings On Medical Matters.’ Hess webpage article #17 Retrieved from here.

Hess, W. R. (November 2014). ‘Hypothetical Earlier Dating for The Passionate Pilgrim and The First Folio.’ Journal of Literary and Arts Studies (JLAS) 4(11) 916-940. Retrieved from here

Hotson, L. (1949). Shakespeare’s sonnets dated and other essays. London: Hart-Davis.

Knutson, R. L. (2001). Playing companies and commerce in Sh’s time. Cambridge: Cambridge University Press.

Lee, Sir Sidney (1901). A Life of William Shakespeare. London, Macmillan.

Lee, Sir Sidney (1905). Shakespeares Sonnets. Oxford, Clarendon Press.

Marino, J. J. (2011). Owning William Shakespeare. Philadelphia: University of Pennsylvania Press.

Meres, F. (1598). Palladis Tamia: Wits Treasury, Being the Second Part of Wits Common wealth. London: Peter Short for Cuthbert Burby.

Schoone-Jongen, T. (2008). Shakespeare’s companies. Burlington,VT: Ashgate Publishing Co.

Shakespeare, W. (1609). Sonnets by William Shake-speare. EEBO image STC-22353-1034:13.

Shakespeare, W. (1640). Poems: Written by Wil. Shake-speare. Gent. EEBO image Harl.5927- E4:1[180f].


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