Much Ado about Marlowe
I set out to assess the probability that Christopher Marlowe wrote any of the plays of William Shakespeare; anticipating that it would become evident quite quickly that each of their works had been created by their accredited authors. The results of my simple statistics test proved interesting and probably controversial.
The idea that I might be able to spend time analysing the semantics involved in detail and be able to decide who, if anyone, copied whom was beyond my literary skill. So I resorted to employing the statistical techniques I had used in industry.
I examined the works of each author hoping to find some property that could be assessed. It soon became apparent that the frequency of use of some small every day words might be used to discriminate between the two writers.
One word; the pronoun, I, would be indicative of a preference for first person declamatory speech. Another word that seemed likely to yield some benefit was, and.
I wondered how conscious the writers of that century would have been about how often they had used the word. I realised myself that there are many words that can easily be thoughtlessly repeated and have to consciously guard against such offences.
The word, but, was another candidate. This was interesting because Marlowe sometimes used the word, yet, in a similar way.
The employment of optical character recognition was very useful when it came to reading each play and counting the number of repetitions of my chosen key words.
Unfortunately I only had the works of Christopher Marlowe on my Kindle but happily accepted the onus of careful counting the words using the mark one eyeball. Fortunately, Marlowe produced fewer plays.
The data was analysed individually for each word and did show some marked differentiation. Marlowe's plays revealed a very tight distribution for the repetition of the key words. Shakespeare's on the other hand had a double distribution.
I resorted to creating a ratio of the frequency of the use of the words and and I.
It might be debatable whether or not this exaggerated any differences, but it certainly highlighted a double distribution for the Shakespearian results.
William Shakespeare's repetitions fell into two groups: one with a very high comparative ratio and another that had a tight distribution similar to that of Christopher Marlowe.
The notable high differences occurred in the following Shakespeare plays: Trilus and Cressida; Much ado about Nothing; Richard III; Hamlet; and Henry V. The remaining ten plays included in the study had the Marlowe-like tight distribution.
This study may be flawed for many reasons: my mark one eye ball might have erred; the sample sizes might have been too small (I did randomly sample each play, taking into account both act and scene).
If nothing else, I may have stimulated others to make more extensive studies using statistical techniques. However, I must add that the results of my simple test do show a striking difference between the two authors.
Richard Labram
I began writing poetry and fiction after I retired. Never far from rhythm or lyrics, I was the winner of the first European Piano Teachers• Composition Challenge in 1987; later working on musical arrangements for amateur stage shows.
During my career in the electronics industry, I published articles on Astronomy and wrote a number of walking commentaries. In addition, I have written articles for house magazines and was for a few years the editor of an in-house publication.