Plagiarism-Detection Software Confirms Shakespeare Play

mi tips us that software intended to help essay graders detect plagiarism has been used to attribute to Shakespeare — with high probability — a hitherto unattributed play, 'The Reign of Edward III.' It seems that the work was co-authored by Shakespeare and another playwright of the time, Thomas Kyd. "With a program called Pl@giarism, Vickers detected 200 strings of three or more words in 'Edward III' that matched phrases in Shakespeare's other works. Usually, works by two different authors will only have about 20 matching strings."

Comments

I call bullshit

[popo]

This is a very unscientific study, with far more potentially meaningful variables than they have accounted for here.

For example, these matching strings could just as well be common turns of phrase of the day. There doesn't seem to be any indication that the software was re-configured for common expressions of old English.

The study would be more plausible if works by two different authors IN ENGLAND IN THE YEAR 1600 contained 20 or so matching strings. But since that control group is missing -- so is the validity of the conclusion.

Re:I call bullshit

[osu-neko]

For example, these matching strings could just as well be common turns of phrase of the day. There doesn't seem to be any indication that the software was re-configured for common expressions of old English.

This is gibberish. The software isn't configured for common expressions of modern English, either. If you understand what it's doing, you should understand why no such configuration is necessary, as long as the two works being compared are contemporaneous. (Or heck, even if they aren't -- correlation should go down in that case, a high score is even more indicative when comparing non-contemporaneous authors.)

The study would be more plausible if works by two different authors IN ENGLAND IN THE YEAR 1600 contained 20 or so matching strings. But since that control group is missing -- so is the validity of the conclusion.

This is just misinformed. They've compared works by both the same author and different authors in England around 1600. It turns out it's just as true then as it is today that works by different authors contain significantly smaller sets of common wording. Indeed, this technique is used to identify which 60% of the play was written by Kyd (by comparing with his other work) and which 40% comes from The Bard. Comparing known works of either Kyd or The Bard with other works by the same author produce the same high correspondence, and comparing known works between the two different authors produces the same low correspondence.

Re:So what they're saying is that...

[wvmarle]

According to WP, copyright started with the Statute of Anne in Britain in 1710. International copyright recognition came only later.

Also according to WP, Shakespeare lived from 1564 to 1616.

So actually I think Shakespeare's plays were never copyrighted in the first place.

Re:Or...

[osu-neko]

From TFA there were parts of the text that were very strongly Shakespearian, and parts not. There is no word on whether they did a plagiarism test on this script vs Kyd's work.

Actually, there is, and they did. About 60% of the work does match Kyd's other known works, as well as four other unattributed plays that are believed to be by Kyd as well (and this result would lend further credence to that).

Being pedantic

[Kupfernigk]

Shakespeare didn't write Old English. He actually wrote modern English. Old English is Anglo-Saxon. Even Chaucer (Whan that Aprille with his shoures soote) wrote in English, though he was sometimes unsure as to how many esses to use.

Why the pedantry? Because, if you didn't know that, you really shouldn't be pontificating on linguistics or linguistic analysis.

Re:Being pedantic

[jonadab]

> Shakespeare didn't write Old English. He actually wrote modern English.

No, he wrote early English. It *is* English (unlike Old English, which is not the same language at all), but it is most definitely not the modern form thereof.

> Old English is Anglo-Saxon.

Yes, that's right. Actually, "Anglo-Saxon" is a much better name for it, because it's not really anything you would recognize as English. It's much more closely related to Germanic and Scandinavian languages.

> Even Chaucer wrote in English,

Chaucer wrote in Middle English, which is more similar to English than Old English is, but still most definitely not the same language. In fact, English is less similar and less closely related to Middle English than French is to Latin. The relationship to Old English is even more remote.

To get from Old English (Beowulf) to early English (Shakespeare) you have to stir in such generous quantities of loan words (mostly from French, Latin, and Greek) that fewer than 10% of the words in the language trace their ancestry back to Old English. You also have to make considerable adjustments to the morphology of the language, significantly alter the orthography (taking the basic spelling system apart and putting it back together differently), completely change the phonology of all the vowels and several of the consonants, alter the grammar in a number of significant ways, and run through several rounds of vulgarization (i.e., let the street lingo of the common people diverge so substantially from the written form of the language that it essentially becomes a creole, then get enough authors to start writing in the common language of the people that it becomes accepted in educated circles; rinse and repeat several times), among other things.

It's sort of like the relationship between Classical Latin and Haitian Kreyol, except that English has had a larger number of external influences on its vocabulary and grammar.

Re:!(!confirmed)

[Golygydd Max]

About six years ago, the Royal Shakespeare Company presented a performance of Edward III and attributed it to Shakespeare. It's accepted that Shakespeare didn't write every word of every plays in his canon (for example, he didn't write most of Pericles and Henry VIII) but there was obviously enough evidence for most Shakespeare scholars to accept that he wrote a substantial part of it. This latest piece of research is just a further piece of evidence, but it's nothing radically new.

Re:Or...

[jipn4]

Or complete sentences. If this is plagiarized it is at least seriously rewritten.

Yes. People actually rewrote things while copying back then; no cut-and-paste.

Shakespeare being famous is not necessarily the one being plagiarised. Maybe he is the one plagiarising.

There was no plagiarism in the modern sense back then. Authors, artists, and scientists copied each others works; that's why we got such a rich cultural heritage. Today, you can get in trouble for a single sentence.

Imagine how backwards computers would be if you had to write a new kernel, window system, and libraries every time you wanted to write an application.

Re:Now Try This

[radtea]

I did this test using three different programs. They said that at least two people wrote the work.

This is interesting, but have you validated this method of analysis by applying it to works of known authorship, say on fanfic sites or alt.politics newsgroups, which would be reasonable control sources--unedited outpourings of interested amateurs? That would tell you that works of the same author don't get flagged as different simply due to your reading-level split.

Ideally I'd like to see a p-value for your claim that "the work was written by at least two people" against the null hypothesis "only one person wrote the work". Without a p-value you really aren't saying anything. Presumably the plagiarism detection software produces a probability of works being by the same author. What you need to do is apply your reading-level split to a bunch of works and generate a distribution (histogram) of the probabilities that the two parts of each work are from different authors. Then ask the question: what are the odds that the probability I get from applying this analysis of Kaczynski was drawn from this distribution? That is your p-value. If it is very small, it is implausible that Kaczynski's work was written by one author.

There are still problems with your approach, but doing this would bring you into the realm of discourse where people could argue about your method, but not dispute the objectivity of your result given your assumptions.

Re:Divine inspiration

[gtbritishskull]

Some religious sects do encourage questioning of the bible. Some don't believe the (King James version of the, lol) bible is the actual word of god, but instead a document written by humans. While it is still the basis of the faith, it is understood that it is written for people in the 1st century BC and therefore should be interpreted through that lens. You shouldn't stereotype over such a diverse range of people like that. It just makes you sound ignorant and reactionary.

Re:!(!confirmed)

[BrokenHalo]

for example, he didn't write most of Pericles and Henry VIII...

As to the latter, he might not have wanted to claim too much ownership to that play, given its first performance only 10 years after the death of Elizabeth, Henry's daughter. Dangerous ground indeed, given the treatment meted out by the Queen's secret police to other playwrights of the time.

In fact, the first performance of that play happened to be the same night the Globe Theatre burnt down. Good fodder for conspiracy theories there...