The catalyst to this thread was my expressing regret at attempting to read Peter Frankopan's new book, "The Earth Transformed: An Untold History". Given the relevance of the stated topic of the book, I expected that many ASIF members would be tempted to buy it, and so perhaps some sort of review could warn them off (or at least help them prepare) for any attempted reading of the book. Since nobody has expressed strong revulsion to this I have decided to attempt a review as follows:
Peter Frankopan is a Professor of History at the University of Oxford, apparently specializing in Byzantine history. My previous encounter with him was through the quite decent Popular History books on the Silk Roads ("
The Silk Roads: A New History of the World (2016)" and "
The New Silk Roads: The Present and the Future of the World (2018)")
The subtitles are certainly hyperbolic, but the books were entertaining and readable, although I did disagree with the central thesis of the first book, i.e. that the land-route above the Himalayas (what most people think of as the "Silk Road") was ever of any significance except as a result of the pacification and unification of the entire route following the Mongol Conquest in the 13th century, to fade away until the Russian Siberian tea trade took off in the 1700s. The trade of silk to the west in exchange for silver (as lamented by Cicero in our earliest known documented evidence, around 70 BCE) was through the Indian subcontinent. The much shorter and easier landroute from China to the Bay of Bengal was also the most lucrative, with the immensely rich principalities of the Ganges and Brahmaputra valleys being the main buyers. But some of this silk found its way to Trabopane ( Ceylon / Sri Lanka) and the west coast of India, into the hands of eager Roman buyers. Central Asian trade networks, on the other hand, were primarily north-to-south, with the traditional taiga products of skins, pelts and honey (and probably also slaves) being exchanged for luxury items.
The second book was more of a "let us make money on the back of the popularity of the first one", and was sorely lacking in any decent discussion of the Belt and Road initiative, which I in my naivity thought might be of central interest in any discussion of the "New Silk Roads". Unfortunately, the book had nothing new to say and nothing very memorable.
I have not had the change to read some of his other books that are closer to his field of expertise. But there is something about this newest book that makes me less likely to do so.
"
The Earth Transformed: An Untold History (2023)" claims to be about the interaction between humans and human civilisation on the one hand, and climatic change on the other. A very topical topic, and a tempting bait to all who worry about the future of mankind, particulary ASIF members who foresee climatic catastrophe in the near future (and not without some gleeful schadenfreude at times).
One of the first thing one notices is that the book is very long - 694 pages in large format, an estimated 300.000 words (which would clock in at well over 1000 pages in standard Paperback format - even the most prolific of modern-day novelist dare not publish books of that size).
Now don't get me wrong - I love long books, being a voracious and very eclectic reader of fiction and non-fiction. Last month I re-read, for the fourth time, Neil Stephenson's Baroque Cycle, all 2700 pages, while waiting for my latest Amazon order to arrive here at the source of the Blue Nile. My basic requirements are decent writing and enganging content, with a preference for history, religious history, philosophy, modern fiction, classic fiction, fantasy and the occasional science fiction. But my preferred category of books are those that tackle the whole world, and the entirity of human history. Jared Diamond's "Guns, Germs and Steel" and Yuvel Noah Harari's "A Short History of Human Kind" spring to mind, and if anybody wants a fuller list I could dig down into my library and see what I can find.
So imagine my excitement at recieving this door-stopping tome of global enviromental history. I dug in at full speed, but with rapidly lowering enthusiasm. The fist day I managed to finish the introduction and the first chapter and half before throwing the book away in exasperation. I decided to give it another go, and then a third, and all in all I managed to finish almost a third of the book, although I must admit that towards the end I was simply skimming, to see whether any nuggets of gold might be hidden in the dross.
So what is wrong with this book? Firstly, the writing style. It is extremely formulaic, and is based on what appears to be the following:
- Use Google to find X papers on a given subject.
- Use ChatGPT or similar to rephrase the introduction and conclusion of each paper into 100 words or less.
- Glue these AI generated factoids together with a bit of meaningless verbiage.
Now I may be wrong about the use of LLM such as ChatGPT, but giving the timing, it is certainly possible. Basically the book can be described by the formula (<verbiage>factoid</verbiage>)
nThis makes for very tedious reading, with seemingly random jumps from one part of the world to another, with claims being made and even redacted within the same paragraph only to be totally ignored in the next, with no coherence or overall arch of storytelling or at least theorizing.
My second problem with the book is that the author absolutely misses out on several very important points.
The author seems totally unaware of the glacial / interglacial rythm of the last couple of million years, and how this could tie in with the evolution of modern humans, and the waves of spread of Neanderthals and Homo Erectus before that. He does mention Milankovitch at one point, but doesn't seem to understand the implication!
Although he does mention the very exciting (although not generally accepted) theory of pre-industrial human-induced GHG releases as an explanation for climatic variation during the Holocene, he then totally neglects to follow up on this. Here is a very rich seam of exploration for the overall topic of the book, only to be almost totally dismissed. Deforestation due to population increase and reforestation due to population collapse are very interesting climatic feed-back mechanism with very interesting explanatory capabilities that could have ben investigated and synthesised into a coherent narrative. But not in this man's book!
He (almost) totally ignores the deforestation of the Mediterranean seaboard during pre-classical times, with the concomitant silting of previously decent harbours, which led to repeated failure of Greek and Punic colonisation. Also whether the massive deforestation of Northern Africa during Roman times may have caused the Roman Climatic Optimum, or been a factor in the consequential drying up of the North-African littoral and the collapse of it's agricultural potential.
I did say almost in that last paragraph. He does mention it but then dismissed it without any real justification, flying in the head of generations of research.
A minor quibble are the strange errors that litter the book, mostly apparent in very strange dating errors. In one place he claims that the Indian subcontinent broke away from Africa 1 million years ago. Another has the minor climatic optimum of the last glacial period, and the consequential cooling, at 450.000 BP in stead of 45.000 BP - and blamed on volcanism, which seems to be his preferred explanation for anything that happens.
I decided to stop reading the book because I was sure I would be increasingly exasperated by a growing tide of errors, and a total lack of context, coherence and historiographical analysis.
All in all, I think I should stick to my AI generated hypotheses. This is quite posibly one of the first examples of LLM generated popular history books - at least it bears all the hallmarks of "artificial intelligence" which is total lack of contextual and conceptual understanding, total disregard for facts and a verbal spaghetti style, grammatically correct and without spelling errors, but devoid of interest.
(Disclaimer: All the above was written by a meat brain, and not an LLM).
