Wednesday, September 25, 2019

John de Courcy & The Infinite Monkey Theorem – a case of plagiarism in popular publishing

John de Courcy arriving at Carrickfergus Castle (Photo: Author)

When I first lodged a complaint of plagiarism against Prof Andrew Bevan and a number of his colleagues at UCL and University of Portsmouth I had no real idea where it was going to go [read more here]. I certainly had no inclination that it would leave me with the reputation as a staunch advocate for the rights of the plagiarised. Since the publication of that piece, I have been honoured to receive the trust of several researchers who have shared their experiences. I have attempted to listen with compassion and give the best, most constructive advice I possibly could. All the while my disgust at plagiarism and plagiarisers has only intensified and grown. It was in this guise that my dear friend, Dr Nic Wright, recently contacted me. He had been doing some light reading on John de Courcy, that well-known Anglo-Norman knight and general trouble maker who arrived in Ireland at the end of the 12th century and went on to take large tracts of land in Ulster, become vastly wealthy, before losing everything to Hugh de Lacy and ending his days in poverty and penury. It’s a gripping tale that has often been told for academic and popular audiences alike. Admittedly, there are only so many facts to go around and the story has been told so many times. Even still Dr Nic felt like he’d read one particular passage before. Not just a description of the same events – he’s actually read it before! He picked up the other de Courcy book and after a bit of searching came across the passage he was after.

Have a read for yourself – First, here’s an excerpt (pp 63-64) from Steve Flanders’ 2015 book John De Courcy: Prince of Ulster:

“John had not been idle since seizing Downpatrick. He had ordered his men to build a temporary fort within the wide-ranging ramparts at the top of the hill, where John founded the cathedral. His fort was planned to prevent surreptitious incursions and thefts as well as to provide security for John’s foreign invading force in the midst of Irish territory. Rapidly-built, it was designed for routine defence, but not to forestall an army. It could not withstand a siege. John had not had much of a chance to amass a stockpile of food and his enclosure did not contain a source of fresh water. He was isolated in a foreign country controlled by the hostile Irish keen to defend their own land. Lastly, but most importantly, the circuit around the top of the Hill of Down – Cathedral Hill – was far too long for John to defend, as it would require him to spread his forces very thinly. The hill fort did not suit the Anglo-Normans’ method of waging war nor was it a position that could be adequately defended. To be victorious, John needed to fight the Irish on land of his choosing and on his terms. The Anglo-Norman army required room to fight. Its cavalry could only be effective on firm ground and with a sufficient distance to build up a gallop for a crushing shock assault, if required. Similarly, the armoured men-at-arms and lighter infantry needed to deploy on good ground. Ideally, this would have a slight downward slope towards the enemy, so as to give each man sufficient space in which to fight and to raise them higher than the attackers. The Normans did not use the compact shield wall of tightly-packed heavy infantry characteristic of the Saxons defeated at Hastings a century earlier. They needed room to swing their swords. Such an open formation, in which each relied on his compatriots on either side, could only be achieved through discipline, experience and trust, and this the Normans at Downpatrick had in abundance. Whether it was justified or not, it was the cornerstone of their own sense of superiority. In addition, the Anglo-Norman archers also needed space to fight. They would be much less effective tightly packed into a confined fort. Their strength lay in their ability to produce rapid missile fire with arrows landing on the enemy, in flight, and about to be fired. In addition, they could move to new positions as quickly as a running man. A competent commander could use this flexibility to bring their firepower down on the enemy wherever it was needed. Moreover, apart from such responsiveness, archers could keep a retreating enemy in range by following up as they fell back or ran away. John used his archers very efficiently at Downpatrick.”

Now, for comparison, please direct your attention to an excerpt (pp 12-14) from Michael Sheane’ 2017 work The Conqueror of the North: John De Courcy’s Campaigns in Medieval Ulster:

“De Courcy had not been idle since capturing Downpatrick. He had ordered his men to build a temporary fort within the wide-ranging ramparts at the top of the hill, where he had founded the cathedral. His fort was made to stop incursions and thefts as well as to provide security for de Courcy’s invading force in the midst of the Gaelic territory. It was rapidly-built. It was designed for routine defence, but not to stop an army. It could not withstand a siege. De Courcy had not had much of a chance to amass a stockpile of food; his enclosure did not contain a source of fresh water. He was isolated in a foreign land controlled by warring bands of hostile Irish keen to defend their own land. Lastly, and most importantly, the circuit around the top of the Hill of Down – Cathedral Hill – was much too strung out de Courcy to defend, for it would require him to spread his forces quite thinly. The hill fort did not suit the Anglo-Normans’ method of waging war; nor was it a position that could be easily defended. To be victorious, John required to fight the Irish on land of his choosing and on his terms. The Normans army needed room to fight. Their cavalry could only be effective on firm ground with sufficient distance to build up a gallop for a crushing assault, if required. The armoured men of war of the light infantry needed to deploy on good ground. Ideally this would have a slight downward slope. The terrain gave each defender sufficient space in which to fight and a slope to raise them higher than the attackers. The Normans did not use the compact shield wall of tightly packed heavy infantry, characteristic of the Saxons’ defeat at Hastings a century earlier. They needed room to swing their swords. Such an open formation could only be achieved by discipline, experience and trust, and this the Normans at Downpatrick had in good measure. Whether it was justified or not, it was the cornerstone of their own sense of superiority. In addition the Norman archers also required space to fight. They would be much less effective tightly packed into a confined fort. Their strength lay in their ability to produce rapid missile fire with arrows landing on the enemy, arrows in flight, and arrows about to be fired. In addition they could be moved to new positions as quickly as a running man. A competent commander could deploy this flexibility to bring their firepower down on the enemy wherever it was needed. Moreover, archers could keep a retreating enemy in range by following up as they ran away. De Courcy had used his archers with great effect at Downpatrick.”

I know it’s asking a lot to beg a reader to critically assess two texts like this. Indeed, my first reaction was ‘sure … there seems to be a bit of overlap … I suppose’. I reckoned that if you could quantify the similarity, you’d be in a better position to make a judgment on the degree of overlap between them. While the form of plagiarism I’d been exposed to was the use of my work with due acknowledgment, this appeared to be of a different order – the close copying of entire paragraphs between the two texts. There had to be a technological solution available! In fact, there were several. I first converted the scanned pages into text files and then ran them through Diffchecker, a free, online resource that compares and contrasts two texts. In the image below, Flanders’ text is on the left, Sheane’s on the right. The first ting to note is that there are 45 removals from one text that are apparently paralleled with 44 additions to the other. For example, ‘de Courcy’ is frequently replaced with either ‘John’ of simply ‘he’. There are verb substitutions that include swapping ‘seizing’ for ‘capturing’ and ‘use’ for ‘deploy’, or ‘forestall’ and ‘stop’. In many cases, the changes are minimal and are limited to the addition or excision of a comma. Thus, ‘Ideally’ becomes (or was changed from) ‘Ideally,’.


Although this seemed to present a robust prima facie case, I wanted something more analytical that could actually quantify the similarity. This I found in the Copyleaks site – another free-to-use online resource. In the image below the Sheane text is on the left. Here the results indicate that of the 459 words of the Sheane text, 415 are copied in the Flanders text, indicating a 90% similarity between the two texts. While I bear a righteous dislike of pie charts, the one presented as part of the Copyleaks output (with four categories) is just on the edge of acceptability. Nonetheless, it is clear that the analysis indicates that 80% of the text is classified as ‘identical’. A further 6% is regarded as ‘Minor Changes’ and 4% being ‘Related Meaning’. The remaining 10% of the text is unique to the Flanders text.


Well, there’s no denying that they’re similar, but I was reminded of the Infinite Monkey Theorem that while admitting the vast improbability of a group of monkeys eventually producing the entire works of Shakespeare, it is not zero. I wondered what the chances of even short phrases being randomly composed in the same way between the two authors may be. A colleague suggested that, while not without limitations, one of the online passphrase strength checkers might provide a useful model for assessment. To this end, I selected three short phrases that were identical between the two texts and ran them through three passphrase checkers.

Phrases: 11-15 words
A: “He had ordered his men to build a temporary fort within the wide-ranging ramparts”
B: “fight the Irish on land of his choosing and on his terms
C: “Their strength lay in their ability to produce rapid missile fire

Strength checkers:
I: How Secure is my Password (https://howsecureismypassword.net/)
III: Passwordmeter (http://www.passwordmeter.com/)

The results of this were that Password Strength Checkers II and III regarded all three as ‘Very Strong’. However, Strength Checker attempted to quantify exactly how long it would take a determined hacker to crack these using brute force attacks. In order, these were 3,919,318,814,427,268 Quadragintillion Years; 37 Septenvigintillion Years; and 2 Trestrigintillion Years. That’s a lot of time!

So … we can be satisfied that there are effectively the same text and are vanishingly unlikely to have been created through random means. But who is the plagiariser and who is the offended party? Well, that should be simple, shouldn’t it? One book has a publication date of 2015 and the other was published in 2017. Apparently not! To make the publishers and authors aware of this situation, and to allow them the opportunity to comment, I reached out to the publishers and waited for their responses. The first back was Rose Nicholas, Sales Manager for Arthur H Stockwell Ltd, representing Sheane. Although his book has the later publication date the email from A H Stockwell reads (in part) “However, we can advise that we have contacted Mr Sheane who has confirmed that the work within the book is his own.”. A little later Malcolm Johnston, Head of Commercial Publishing for Colourpoint Books and Blackstaff Press replied. Although, seemingly, more concerned with why I was pursuing this issue and where I would seek to have it published, he noted that they were already aware of the 'apparent plagiarism'. Based on these responses, I am reluctant to call out one party over the other as the victim versus the villain.





Even if neither publisher is (at least outwardly) concerned about this situation, the case remains that plagiarism is wrong and should be called out at every opportunity. Plagiarism is inherently unethical because it is a form of theft. One person steals the work of another and seeks to pass it off as their own. Just as much with popular publication as within the academic sphere, this is of great importance as the plagiarising authors receive tangible benefit from that stolen work – either as money from sales or buffing to their reputation and prestige. There is no escape from it – stealing another person’s work is the same as taking money from their pocket or diminishing their good name and prestige.


Notes
Dr Nic also happens to have a (totally not plagiarised) book available: Lore of the Land. A free digital version is available [here]. It can also be found by searching for 'Lore of the Land' on the Northern Ireland Community Archive site (www.niarchive.org). Physical copies are also available for free from Visitor Information Centres across the area, or by contacting Museum Services at cms[at aymbol]causewaycoastandglens[dot]gov[dot]uk


The Lore of the Land project was part of Causeway Coast and Glens Borough Council’s Understanding Our Area programme, delivered by Museum Services, with funding from the European Union’s PEACE IV Programme, managed by the Special EU Programmes Body (SEUPB).

There is an irony, not lost on me, that a search for one book on Amazon lists the other under the heading of “Customers who bought this item also bought” … yes, but not in the way you might first imagine!

I cannot over-stress the importance of a good passphrase (as opposed to a password). Use resources like those listed above to test their strength or follow the good advice in Randall Munroe’s wonderful xkcd comic.


No comments:

Post a Comment