The Shakespeare Conference: SHK 29.0425  Thursday, 6 December 2018

 

[1] From:        Hardy M. Cook <This email address is being protected from spambots. You need JavaScript enabled to view it.>

     Date:         Thursday, December 6, 2018

     Subj:         SHAKSPER: NOS 

 

[2] From:        Gabriel Egan <This email address is being protected from spambots. You need JavaScript enabled to view it.>

     Date:         December 6, 2018 at 3:52:45 AM EST

     Subj:         Re: SHAKSPER: NOS

 

[3] From:        Brian Vickers <This email address is being protected from spambots. You need JavaScript enabled to view it.>

     Date:         December 6, 2018 at 10:47:43 AM EST

     Subj:         Re: SHAKSPER: NOS 

 

 

[1]-----------------------------------------------------------------

From:        Hardy M. Cook <This email address is being protected from spambots. You need JavaScript enabled to view it.>

Date:         Thursday, December 6, 2018

Subject:    SHAKSPER: NOS

 

I hope that we can keep to the issues without the snarky remarks.

 

I welcome private responses on the level of discourse in this thread. They will be just for me and not for publication.

 

Hardy

 

[2]-----------------------------------------------------------------

From:        Gabriel Egan <This email address is being protected from spambots. You need JavaScript enabled to view it.>

Date:         December 6, 2018 at 3:52:45 AM EST

Subject:    Re: SHAKSPER: NOS

 

Dear SHAKSPERians

 

Pervez Rizvi alludes to but does not disclose the method by which he extracted the lemmas from the XML-format files of the 527 plays his recent work is based on. He writes that it took him “months of patient editing to turn those XML files into a database fit for N-gram searching”. Okay, but how exactly?

 

He says he knows how to do it: “I did it because I have the technical know-how, not because I have some secret software that I haven’t ‘disclosed’ to Gabriel”. But he really hasn’t disclosed how he did it. And it matters, because any reader of his work (not just me) wants to be sure he did it correctly.

 

Rizvi describes me as “an English scholar, not a programmer”. This comes close to telling me that I shouldn’t worry my pretty little Humanities head about the technical aspects of the problem. Rizvi suspects that I wouldn’t in any case understand how to parse XML files: “It may well be that when Gabriel looks at those files, he doesn’t know what to do”.

Actually, I teach undergraduate and postgraduate courses in computer programming for the purpose of analyzing texts. I’m asking Rizvi the questions I would ask any student who presented the results of a project without showing all of their method.

 

The lemma for each word in the XML files Rizvi uses is stored as an attribute called @lemma on the element <w> (for ‘word’).  One way to get at them is to write a program in the language XQuery, which is probably the way I would teach students to do it. Another way would be to create an eXtensible Stylesheet Language Transformation (XSLT) to transform the entire document into some other format that throws away the great deal of XML data that we are not interested in and leaves just those @lemma attributes.

 

Yet another way would be to ignore the XML structure and to write a program in some non-XML-aware language (such as C, or Python, or Perl) that just looked for the literal string ‘lemma=”wxyz”’ and pulled out the ‘wxyz’ bit. This is a quick-and-dirty way that non-XML-experts sometimes resort to, and it entails dangers that are not obvious when you first try it. It’s awfully easy to get the wrong results when you use a quick-and-dirty approach.

 

I don’t suppose that Rizvi means to be insulting when he says that I’m not a programmer, but it was an unwarranted and mistaken assumption about me. This assumption about my limitations makes him misread my serious interest in the technical details of what he did as my attempts to “divert attention” from his achievements.

 

I’m genuinely interested. I’m a programmer too.  And I think the technical details are germane to the evaluation of his work. That’s why in my emails I asked for the materials on which it is based and why I’m asking for the technical details now.

 

So, which was it: XQuery, XSLT, or some scripting in a non-XML-aware language? Something else I haven’t thought of?

 

I’m disappointed to hear that some materials underpinning the New Oxford Shakespeare Authorship Companion were not available to someone wishing to follow up an argument made there. I can’t legislate for what my collaborators decide to do with their materials and I can’t require Oxford University Press to make available materials that it owns, but I stand by the principle of total transparency in science and if the reader who was rebuffed would like to get in touch with me I’ll see what I can do to furnish the materials they’re after.

 

Regards

Gabriel Egan

 

[3]-----------------------------------------------------------------

From:        Brian Vickers <This email address is being protected from spambots. You need JavaScript enabled to view it.>

Date:         December 6, 2018 at 10:47:43 AM EST

Subject:    Re: SHAKSPER: NOS 

 

Letter to SHAKSPER 6.12.18

 

Gentles all,

 

Gabriel Egan’s latest posting is a classic example of his polemical method. It completely evades the issue, the fact that the essay by Pervez Rizvi on which I reported has demolished the “micro-attribution” method used by Gary Taylor and his colleagues in the English Department of Florida State University. It did so by using mathematical logic and drawing evidence from a corpus of 527 plays. The demolition is absolute and will never be refuted. The NOS’s attempted deformation of Shakespeare’s canon has suffered a severe blow, for many pages of the Authorship Companion are rendered valueless.  

 

Being unable to challenge its findings, Egan resorted to his familiar weapons: false logic, nit-picking, misrepresentation, and the appeal to authority. For the first, take his argument that if the small sample that Taylor used (63 words in Macbeth)

 

is just too small to test … it becomes impossible to show that Taylor et al. are wrong in his attributions. Vickers’s argument leads to an admission of defeat—nobody can tell who wrote such a small passage—not to the routing of attributions Vickers doesn’t like.

 

But the point is that the passage is too short to yield a reliable result. It does not follow that ‘nobody can tell who wrote such a small passage’, since we have it on the authority of the First Folio that Shakespeare wrote it. Ergo, Taylor’s attempted re-assignment to Middleton has failed.

 

For the nit-picking, Egan objects to my statement that Rizvi has “published” a database of plays, since Rizvi took over the texts from sources he duly acknowledged (I pass over Egan’s other pedantries). As Hamlet says about the Gravedigger, “We must speak by the card”. I should have written that “Rizvi has published a corpus of plays that he has marked up with programs of his own devising, enabling users to compare the two main types of verbal repetition in any two plays in the period between 1542 and 1657”. Please, teacher, will that do?

 

For the misrepresentation, that is a stock feature of Eganian polemic, as both myself and Darren Freebury-Jones have complained in this forum. It is also astonishingly short-sighted, since the authors so treated are all alive and their work is in the public domain. Having been a regular user of Rizvi’s database for some months, I can confirm that his account of Egan’s distortions is accurate.

 

Finally, the appeals to authority. In her review of the NOS Authorship Companion, Lois Potter described Egan as a “law-enforcer”: yes, but only of the laws he has laid down. He thinks that attribution scholars should use the widely available Literature Online database, which would “reproduce the results that the investigators claim”, although he concedes that users “have to trust that the hits returned correctly represent what is present in the plays it is meant to be searching.” I have recently compared the results produced from LION as used by some NOS scholars (MacDonald Jackson, John V. Nance, Anna Pruitt) to justify their attributions, with the results I achieved by using anti-plagiarism software. In every case I found more hits, so either LION doesn’t work very well, or those scholars didn’t use it diligently enough. In any case, compared to the Rizvi database, where the searches have already been made and are wholly reliable, LION is now obsolete.

 

If Egan continues to defend the NOS attributions by such devious methods, I shall continue to expose them.

 

Warm regards.

 

 

 

Subscribe to Our Feeds

Search

Make a Gift to SHAKSPER

Consider making a gift to support SHAKSPER.