The Shakespeare Conference: SHK 24.0461 Friday, 27 September 2013
Date: September 26, 2013 9:36:19 PM EDT
Subject: Re: Spanish Tragedy Additions
I’ll try to develop first-reading objections to Brian Vickers’s “Identifying Shakespeare’s Additions to The Spanish Tragedy (1602): A New(er) Approach.” His essay covers the ground with assuredness and new tools; yet if his assumptions don’t hold up, neither might the conclusion:
The results showed that, apart from Shakespeare and Jonson, none of the other possible authors working in the London theatres in the late 1590s were serious candidates. . . . The results for Shakespeare . . . provide overwhelming evidence for his having written these five enlarged scenes in The Spanish Tragedy. As every reader can see, the uniquely Shakespearian matches amount to 116 in the 320 lines of the Additions, a rate of one every 2.5 lines. In the longest, and most memorable scene, that with the Painter (specially advertised on the title page of the 1602 quarto) the frequency drops slightly, to 65 matches in 168 lines, or one every 2.6 lines. There is no other conceivable explanation: Shakespeare wrote these scenes (italics added).
Now, we can conceive that the dish ran away with the spoon; I gather we’re talking probabilities. And yet this article really is presented as if none other than known, professional dramatists could have written the Additions: when the computers are done, all but one is eliminated. The question isn’t really about who wrote “five enlarged scenes” but who “enlarged five scenes” by a total of 320 single-minded lines. “Uniquely Shakespearian matches” refers to the defined parallels between these Additions and the Shakespeare canon, some of which are found in other tested works (that is, some collocations may not be unique to Shakespeare), and many of these parallels are not particularly useful for attribution. “Most memorable” doesn’t tell us much more than “least memorable.” Other references to quality are similarly hedged.
The conclusion rests on “old-fashioned” parallels found or confirmed by “new-fashioned” software with its advantage in comprehensive speed. I’m not averse to properly weighed parallels in authorship studies. Here the investigators break with the software to accept inexact collocations as matches, which I accept in one sense, and question in another.
No doubt, many Addition parallels derive from known Shakespearian usage. As Vickers observes of Stevenson’s comparisons to canonical texts, “such a series of matching interlinked collocations far exceeds the bounds of coincidence.” But that does not force a conclusion that Shakespeare wrote the Additions. After all, Vickers famously refers to Q1 Hamlet, with a quarto-full of collocations, as “Hamlet by Dogberry.” Why aren’t he and his fellows in the mix? Vickers employs a program called “Pl@giarism”: how do we know plagiarism isn’t what it found? This question hasn’t yet been addressed, though it applies to Vickers’s paper. For example, he cites Coleridge’s reported assessment of 1833:
The parts pointed out in Hieronymo as Ben Jonson’s bear no traces of his style; but they are very like Shakespeare; and it is very remarkable that every one of them reappears in full form and development and tempered with mature judgment, in some of Shakespeare’s great pieces. (Woodring 355)
By “style” I guess “parallel phrases” is meant; there isn’t much style otherwise. If all of them show up later in mature form, the Additions must be early (which 1599 is not). Coleridge sees them as undeveloped and immature. Why should that be? If they were written by someone else, we know why.
Byrne’s most demanding criterion remains relevant, that in order to express ourselves as certain of attributions we must prove exhaustively that we cannot parallel words, images, and phrases as a body from other acknowledged plays of the period; in other words, the negative check must always be applied. (24)
The “other plays” criterion does not rule out extensive plagiarism (as we know the term) from one author. “Certainty” requires a negative check for derivative usage (supposing the check might be accomplished) before parallels decide authorship.
However, many of the 106 matches that [Stevenson] claims between the Additions and Shakespeare seem to satisfy Byrne’s other criterion of uniting parallelism of thoughts and words. Several satisfy his own methodological principle that “the strongest proofs of common authorship are not single or isolated parallels, but clusters or interrelated groups of images and phrases which combine to form a distinct pattern” . . . . One curious aspect of discovering these unique matches between the Additions and Shakespeare is to recognize that the original contexts were often radically different.
A “borrower” recalling or referring to canonical lines would incorporate clusters of parallels and thoughts as a matter of course. “Radically different” source material would be OK. And analysis may show the thoughts in the texts are not all that closely related. R. W. Dent (John Webster’s Borrowing, a very instructive book) observed that Webster’s handling of purloined writing could be so different from the originals that meaning was an obscure second in importance to amassing the material itself. That could apply here; for example, Vickers observes:
No other drama text includes the collocation of verb and noun found in . . . Hieronimo’s rhetorical dispraise of a son . . . Being borne, it poutes, cryes, and breeds teeth. (Add. 3.11) which finds a strange echo in Macbeth’s fear of the boy Fleance, who has just escaped being murdered along with his father Banquo:
There the grown serpent lies: the worm that’s fled
Hath nature that in time will venom breed,
No teeth for th'present. (Mac. 3.4.29–30)
The greater probability, from usage and imagery, is that ‘breeds teeth’ is the strange echo. Macbeth’s venom and teeth stand metaphorically for danger that will develop in the cause of vengeance and justice. Cutting teeth is one of listed, normal, infantile things already bred: perhaps a pacifier is indicated here, but no ‘baby rattler.’ This possibility agrees with Dent’s “presence of an element meaningful in the source but not in Webster” (14).
I prie through every crevice of each wall,
Looke on each tree, and search through every brake,
Beat at the bushes, stampe our grandam earth,
Dive in the water, and stare up to heaven,
Yet cannot I behold my sonne Horatio. (Add. 4.17–21)
I pried me through the Crevice of a Wall (Tit. 5.1.114)
Ile leade you about a Round | Through bogge, through bush,
through brake, through bryer (MND 3.1.109–10)
Does Shakespeare resort to senseless dialogue, even in madness? No one squeezes through every crevice of each wall. The words seem borrowed.
My wife Isabella standing by me, with a speaking looke to my sonne
Horatio … and my hand leaning upon his head, thus. (Add. 4.121–4)
As Stevenson noticed, here Shakespeare recalled a visual representation he had constructed a few years earlier in The Rape of Lucrece, the painting or tapestry depicting the Fall of Troy. In it Nestor can be seen addressing the Greek army: “In speech it seemd [that] from his lips did flie | Thin winding breath which purl’d up to the skie” (1405–7) – compare Hieronimo’s question to the Painter: “Canst paint a dolefull crie?” (Add. 4.129). Around Nestor “were a presse of gaping faces” (1408), all jumbled together: Here one mans hand leand on anothers head. (1415)
No similar collocation of thought and language has been found anywhere else in the period from 1580 to 1642 . . . . Stevenson identified some truly idiosyncratic collocations that point unmistakably to Shakespeare’s authorship.
Since 1594 anyone could recall or read Shakespeare’s representation, where Nestor’s “thinly visible breath” in speech has become ‘a speaking look’. Lucrece’s, ‘leand’ follows ‘a press of gaping faces’ to imply that the crowd forced someone’s hand to find an unusual support. Hieronimo suggests no reason for his use of ‘leand’. The poem seems to have supplied words but no proper setting or meaning.
As we have seen, Marston mockingly alluded to the passage where Hieronimo, not yet considering taking personal revenge, affirmed his faith in heavenly justice:
Well, heaven is heaven still.
And there is Nemesis and Furies,
And things called whippes,
And they sometimes doe meete with murderers,
They do not always scape, that's some comfort. (Add. 3.40–4)
That impotent invocation of unfulfilled justice . . . echoes the occurrence of justice fulfilled in 2 Henry VI, where Humphrey, Duke of Gloucester, having exposed the spurious “miracle of St Albans”, by which Saunder Simpcox claimed to have suddenly had his sight restored, and sends the rogue off for punishment:
My masters of St Albons
Have you not Beadles in your Towne,
And Things called whippes? (2.1.132–4)
The closeness of the parallel, in both words and thought, and the similarity in the dramatic context, indicate that both passages come from Shakespeare’s verbal memory.
Again, since 1594 the printed Contention could supply this line. An actor could just as well remember it (and unnumbered others of his choosing). Marston may have purposely alluded to a line (hackneyed by its singular familiarity) that didn’t bear repeating. Further, in those days murderers were hung.
To use the familiar categories established by Saussure, the “frequently repeated” collocations belong to the domain of . . . the shared lexicon of a speech community, while the “unexpected” collocations belong to . . . an individual's lexicon, and are for that reason more valuable for authorship attribution.
That’s not so if the unexpected collocations are “borrowed.”
Yet collocation matching, as demonstrated here, has no difficulty in eliminating a group of rival candidates and establishing Shakespeare as the author.
Mistakes are often not difficult.
Some of these four word matches are discontinuous, such as this (No. 15), between the Additions and The Rape of Lucrece:
Gird in my wast of griefe with thy large darknesse (Add. 150)
And girdle with embracing flames the wast | of Collatine’s fair loue (Lucr. 6–7)
Formally speaking, that match consists of a verb phrase, “gird(le) with”, and its complement (“large darknesse”; “embracing flames”), and a noun phrase in a genitive construction: “wast of griefe”; “wast | of Collatine’s fair loue”. In order to identify unique matches, we must sometimes take careful note of grammar and syntax.
Fair loves have waists, but what is a ‘wast of grief’? The words seem borrowed. A ‘Waste of shame’ is, at least, penetrable.
Five word sequences (pentagrams) are rare in any collocation study, but the matches between the Additions and Shakespeare’s plays and poems include two instances, Nos. 60 and 107:
I know thee to be Pedro, and he Jaques. (Add. 4.44)
The authentically Shakespearian nature of the Additions can be seen from the fact that links occur between them and every play in the canon, apart from Julius Caesar and the two Fletcher collaborations . . . . If we divide the canon at 1599, the approximate date of the Additions, then the 21 earlier plays, from 2 Henry VI to Henry V, account for 135 links. The 20 later plays, from Caesar to The Tempest, account for only 57 links.
Does it make any difference which half Julius Caesar is in? The date of the Additions is unknown; which is true of Macbeth, Lear (and most other of Shakespeare’s plays). But I’m not sure I understand these figures. The majority of “links” don’t pull much weight; any in the late texts which are repeats from early texts are of no account, seems to me. The evidence is in unusual collocations with early or doubtfully dated texts (according to Vickers’s examples). They seem derivative to me. A few more comments:
1) The whole of the Additions should be judged alongside the collocations. Many have faulted the lines’ style. Nothing is easier than equating nonsense with madness, which is what the Additions do. Likewise, death and grief can evoke emotions no matter how weakly they are written up; to feel sorry for Hieronimo is not to make the Additions Shakespearian.
2) Some effort should be made to put an investigation like this into perspective by examining the history of the texts. Something might turn up.
3) The unusual parallels in the Additions strike me not only as inferior renderings amongst inferior lines, but as being too many. An easy “control” would be to test randomly chosen 320-line blocks from the canon to see if they contain as many “unusual collocations” with the works as do the Additions. Maybe they would, maybe not.
Gerald E. Downs