After losing the Gorton and Denton parliamentary by-election to a leftwing plumber, Reform’s Matthew Goodwin has published a new book: Suicide of a Nation: Immigration, Islam, Identity.
It’s clear that Goodwin was trying to emulate Douglas Murray’s The Strange Death of Europe: Immigration, Identity, Islam, which was published in 2017. It has almost the same tagline and rough thesis. But while we can be sure that Murray did not use AI to write his book, I cannot extend the same confidence to Goodwin’s attempt, which opens with:
‘There are moments in the life of a nation when everything changes – not with a bang, not even with a conscious decision, but with a quiet, creeping loss of confidence so profound that a people start to forget who they are.’
Anybody who has used Chat-GPT will recognise the telltale signs of possible AI writing, such as the ‘it’s not X, it’s Y’ comparisons and the strange obsession with things being quiet or silent.
‘The sloppiness does not end here.’
Coincidence? Perhaps. Goodwin has responded to my criticisms saying that they are ‘spurious and false’. He has also said, elsewhere, that his critics on the left have clearly not read his book, and that he has not used as many references as an academic book would have as he is seeking to reach a mass audience. But we can see, at the very least, that ChatGPT helped with parts of his book, because he left the ChatGPT URL in some of the few references used to justify his arguments. Goodwin says that he only used AI to obtain datasets and cross-checked them.
Much of his analysis appears to refer to events, places and people which do not exist – another indication that a writer has relied on large-language models (LLMs), which frequently ‘hallucinate’ facts.
Take Goodwin’s claims about British schools. He cites reports that in one Bradford classroom, only four out of 28 pupils spoke English as a first language, with teachers reduced to mediating ‘dozens of languages.’ I can find no reporting that backs up this claim and Goodwin provides no source for it in his book. The case sounds suspiciously like the response when you type, ‘find me an alarming case of no English in a primary school’ into ChatGPT and hit enter.
Goodwin also cites a ‘BBC West Midlands’ report from 2019 about a school where 30 languages are spoken by its pupils. I also cannot find any record of this report. What’s strange is that there are plenty of stories, including in Birmingham, which Goodwin could have cited. In 2021, the Metro and Birmingham Mail wrote about Water Mill primary school in Selly Oak, for example, where 31 languages were spoken. But if Goodwin had read these reports he would have found that Water Mill was at the time one of the top two per cent rated schools in the country.
Instead, Goodwin quotes an ‘inspection report’ which says: ‘Many pupils join the school with extremely limited English.’ Again, I cannot find any reference to this quote after reading multiple Ofsted reports about schools where multiple languages are spoken. What the Ofsted reports do say, repeatedly, is that these schools are generally good at their jobs, that their pupils make strong progress, and that support for English as an additional language is effective. But perhaps this doesn’t fit into Goodwin’s narrative.
The sloppiness does not end here. Goodwin seems to have created quotes by Cicero, Hayek, Roger Scruton, Livy, Noah Webster, James Burnham and Walker Connor – an impressive feat, in a sense. ‘The most dangerous experiments are those conducted on entire societies’, is a quote that Goodwin attributes to Hayek, despite there being no record of it elsewhere. It seems the most dangerous experiment is publishing a book without any fact-checking.
Reading the book at some point you have to ask: did Goodwin verify any of his claims? Did he open a single book writing his own? Or did he just accept whatever an AI chatbot spat out because it would make him sound vaguely informed? After I posted about the errors in his book, Goodwin suggested I was a ‘left-wing troll’ and thanked me for boosting sales. He did not address his use of the seemingly fake quotes or produce a source for them. He has, in fairness, invited to debate him about the veracity of his book on GB News next week, something I am greatly looking forward to.
Goodwin has no real excuse for inaccuracies in his book. In the past he was a well-regarded academic at the University of Kent. If one of his students had handed in a dissertation with hallucinated quotes and references that don’t exist, he would obviously have failed them. Unfortunately, it appears that any academic rigour Goodwin once had has long since dissipated.
