Shakespeare's Sonnet 17 is, literally, cryptic:
Who will believe my verse in time to come
If it were filled with your most high deserts?
Though yet heaven knows it is as a tomb
Which hides your life and shows not half your parts.
If I could write the beauty of your eyes
And in fresh numbers number all your graces,
The age to come would say “This poet lies:
Such heavenly touches ne’er touched earthly faces”.
So should my papers, yellowed with their age,
Be scorned like old men of less truth than tongue,
And your true rights be termed a poet’s rage
And stretched metre of an antique song:
But were some child of yours alive that time,
You should live twice, in it and in my rhyme.
There is much prior evidence (brought out in The Biography in Shakespeare's Sonnets) that this sonnet was addressed to Henry Wriothesley, third Earl of Southampton. The opening quatrain suggests that the poem is like a tomb, which hides the addressee's life and shows some proportion of his “parts”. Might these parts be certain letters of his name, concealed within the “tomb” of the poem? The hint is strengthened in the closing couplet, which suggests that, through offspring and this rhyme, he will live twice, or, dependent on how the wording and punctuation are construed, four times.
The latter interpretation has Wriothesley living two lives as flesh (his own and that of a descendant) and two more as manifestations of the poem. In a paper, published in Oxford Journals, Roy Winnick suggests that there are four anagrams of “Wriothesley” built deliberately into compact sections of the poem: two in line 4 (each short of one letter) and two in line 9.1
Winnick's general thesis – that the presence of many Wriothesley anagrams within compact sections of the Sonnets was deliberate – is, as he says, often strengthened by two features of such occurrences:- (1) there are accompanying verbal hints of this style of wit and (2) when translated into their Wriothesley form, the anagrams provide coherent messages in line with the plain text.
In Sonnet 17, there is certainly an accompanying verbal hint (as brought out above). However, the translation of the purportedly deliberate anagrams does not really deliver any coherent message. Also, we end up with too many “Wriothesleys”, given that one of the Earl's manifestations in the poem must reasonably be as its subject.
However, it is possible to “solve” the purported clues in another way. Let us look more closely at the final couplet (or rhyme), presented in its original Quarto spelling and punctuation:
But WeRe some chIlde Of yours alive THat timE, You Should LivE twise in it, and in mY rime.
Each letter of Wriothesley is included in sequence, within the 69-letter rhyme. On this basis, he appears twice in the poem: (1) as its unnamed subject/theme and (2) within the letters of the relevant rhyme – in this case, the punch-line couplet.
What are the odds of this phenomenon, together with its “clues”, occurring other than deliberately?
Based on letter frequencies in Shakespeare's plays, the odds of finding w-r-i-o-t-h-e-s-l-e-y in sequence within a 69-letter script (which is random in everything except those frequencies) are approximately ten to one against.2 However, this approach is far from perfect, because the structure of English and the constraints of fashioning a coherent message in rhyming iambic pentameter removes much of that randomness of selection, on which the calculation depends.3
A better indicator is the frequency of occurrence within the 154 Sonnets. In these sonnets there are 159 rhyming couplets (including 6 in Sonnet 126). Only one of these (the punch-line couplet of Sonnet 17) delivers all eleven letters of Wriothesley in sequence. This suggests that the true odds of an accidental w-r-i-o-t-h-e-s-l-e-y sequence within a couplet of the Sonnets are over one hundred to one against.4
We should now take account of the odds against the poem also containing appropriate clues - in this case, in the opening quatrain and in the punch-line couplet. Taking the couplet first, we can find a reasonably similar theme (of the addressee living on in the verse) in seven other punch-line couplets within the Sonnets.5 This suggests odds against of more than 10 to 1.6 Themes suggesting possible concealment of the addressee in the poem, such as that of lines 3 and 4 of Sonnet 17, are, however, rarer. Allowing some strain of interpretation, I suggest that such a theme occurs within lines of only four other sonnets – 18.12, 54.14, 55.3 and 81.9 - indicating odds against of more than 20 to 1.7
The odds against all three events occurring by accident in one sonnet is the product of the individual probabilities - in this derivation, more than 20,000 to 1.8 Not insignificant.
1 This paper may be read at http://litimag.oxfordjournals.org/content/early/2009/10/04/litimag.imp049.full#xref-fn-9-1
2 Letter frequencies are: w 0.025; r 0.06; i 0.07; o 0.085; t 0.09; h 0.065; e 0.12; s 0.0065; l 0.045; e 0.12; y 0.025, based on information kindly provided by Dr G.R. Ledger. The product of these values is then multiplied by the standard formula for deriving a maximum number of combinations, N!/n!(N-n)!, where N = 69 and n = 11.
3 This works in both directions. Most random combinations of letters would produce gibberish, which excludes them from acceptable script. This feature of random selection indicates that the odds in intelligible script are significantly longer. However, some letter sequences, such as t-h-e, occur more frequently than predicted under individual random selection, thereby tending to offset such lengthening of odds.
4 These odds are comfortably supported by the incidence of wriothesley sequences within pairs of consecutive sonnet lines (in their original Quarto spellings). In the first 70 sonnets I found only five hits: at lines 11.7-8, 17.13-14, 51.14-52.1, 55.12-13, 63.3-4. Taking into account the couplets of the remaining sonnets (already assessed above), this suggests odds of 5 in [(70x14) + 83 + 6] = 1 in 214. Boredom and eyestrain stopped me at this point, given the substantial cushion against counting and statistical error.
5 Sonnets 15, 18, 19, 55, 63, 65 and 107.
6 8 sonnets in 126 (restricting the latter to the Fair Friend sequence) points to odds against of approx 16 to 1, giving a substantial cushion against interpretive and statistical error.
7 5 sonnets in 126 points to odds against of 25 to 1, giving a reasonable cushion against statistical error.
8 100 x 10 x 20 = 20,000.