Maty's blog
| 2025-11-04 |
| Winterizing |
| One of the things I love about life in the Great White North is that we get four real seasons. Unlike some countries where a bad summer's day is about the same as a mild winter one, each season here has a very distinct identity. Winter - if all goes well – is snowy and sometimes very, very cold. Spring is a buggy, boggy nightmare as all that snow thaws out and a few zillion hungry bugs go looking for blood. Summer is long lazy afternoons, hiking, picking berries and kayaking up and down lakes – all the while making sure that the go bag is ready in case wildfires mean we have to flee at a couple of hours notice. (Hey, nowhere's perfect.) Autumn is a short season. Sometimes as little as five weeks pass between my last hike of the summer and when I do the same walk over 30 cm of snow. Being short, autumn is also very busy. There's a high risk of power cuts as ice and snowload take out the power lines, so storm candles are needed. Also six months of firewood should now be split and stacked up ready to go. The garden needs to be stripped of anything that is going to break under snowload or rot in the spring thaw, and food needs to be stacked in the freezers (we have two) and the basement. In a lot of this I can feel ancient ghosts at my shoulder. People have been worrying about winter firewood supplies for tens of thousands of years since they first moved out of Africa. They did not have the advantage of stacking shelves with tins, but my large glass containers filled with rice, flour and dried beans would have been immediately familiar, as would the cold room where sacks of potatoes, onions and cabbage dangle from the rafters so rodents can't get at them. There's a certain ritualistic quality to all this, from building the first fire of the year to checking winter gear, from padded coats, snowshoes and skis to making sure snow shovelling kit is in order. Of course if I wanted to do rituals Roman style, I'd celebrate the Consualia, which celebrates bringing the harvest home, and the Ludi Romani which marked the autumnal equinox. Instead I think I'll settle for the Meditrinalia, where Romans sampled the first new wines of the year. There's some Merlot maturing in the cellar that will do nicely. |
| 2025-10-03 |
| Wandering over a text |
| Generally when I discuss AI with colleagues the general theme is how students produce essays that Gemini, Copilot or chatGPT have written for them. I agree that for educators with a class of 30 this is a problem. But that's a bug with the class system rather than technology. I generally have smaller classes, and so know my students better. When I see something very different from the usual style of, say, Mary Bloggs, my technique is to ask Mary some details about her efforts. 'Great work on describing Julia's relationship with Tiberius. What sources did you use?' But AI is more than an essay mill. For example, there's Aeneas. Who means that historians have to add understanding multimodal generative neural networks to their skillset. Aeneas - who really should be called AIneas - is the solution to reading the 1500 or so Latin inscriptions that archaeologists recover each year. (The Romans wrote a lot, from inscriptions, to letters to graffiti.) Many of these texts are in poor shape after 2000 years or so, and recovering them can take a team of skilled epigraphers rather a long time for each. However inscriptions are somewhat formulaic. When one already knows the tens of thousands of inscriptions on world databases, then given fragmentary text and an idea of the dimensions of the text surface, a well-trained AI can fill in the blanks with considerable accuracy. Google, which developed the code, reckons around 70% accuracy based on obscuring parts of known texts and getting Aeneas to look at the rest. Why Aeneas? Well, Google already has an epigraphy machine for Greek texts named after Ithaca, the home of wandering Odysseus. Therefore when going for a Latin version of Ithaca, Google's minions settled for Aeneas, Odysseus' Roman counterpart More details here: https://deepmind.google/discover/blog/aeneas-transforms-how-historians-connect-the-past/ |
| 2025-09-04 |
| Myths to Music |
| Feminist takes are big in mythology these days. So I've jumped on the bandwagon, with the emphasis on 'band'. As viewers of my FaceBook page will know, I've been working with a singer called Ainsley Blackwell. We recently did a song called 'Hera' which was well received. I wrote the lyrics which Ainsley delivered with slightly concerning psychopathic flair. You can find it on YouTube. This led to a number of re-interpretations of modern myth. Eurydice was an obvious choice and we did a song called 'Don't look back' where a woman basically shoves her one night stand out of the door. 'Wax wings' is a jazz take on Icarus where a girl contemplates a relationship with a man supposedly out of her league. Other times we just lean into the ancient myth. There's 'The Summoning' where we evoke Hekate at a crossroads. Persephone was fun, because in the song she's not just a simple victim. She's terrified of her role, but also finds a dark excitement in it. 'I dance in shadows I should have fled/Before the thrill outweighed the dread.' It's a lot of fun and it has me imagining the classics in a new way. It's easy to write the lyrics, because the themes are archetypal for a reason. They touch on basic issues of human nature and relationships. So rewriting them is almost cheating. You can see Ainsley Blackwell's site here: http://www.songsbyainsley.com/index.html |
| 2025-08-05 |
| Another Caesar myth |
| Doing the introductory chapter of a book on the Gallic wars, I wanted to include a few hundred lines of the boyhood of Julius Caesar. Rather to my surprise I found that he did not have one. No doubt Caesar was conceived, born and raised in the standard way, but we don;t know about it because no-one tells us. When writing I tend to skip lightly over anything written in the last 1700 years and go to the original sources. Some checking told me that Caesar had a Gallic tutor called Gnipho, and that he had a slight build and liked to challenge himself by riding horses bareback with hands behind him. Also, a later reference tells us that he swam a lot as a child. And there, in less than 126 words we have all that is known about Caesar's boyhood. Growing up in the Subara slum with cash-strapped parents? This is a modern fantasy. We have just one reference to Caesar living in the Subara, and that is Suetonius who tells us that Caesar lived there in his bachelor days before he moved to the Sacra Via. Growing up there? No reference at all, because for Suetonius, Plutarch and other biographers, Caesar only gets interesting around age 15 and no-one tells us where he was living at that age. Yet it is a commonplace among modern writers that Caesar grew up in an apartment block in the Subara. Yet every reference I chased down led to another reference by a modern writer who references yet another modern writer. More scrupulous historians such as Goldsworthy pass over the subject altogether. No writer can give an ancient source for this. Also, Caesar's mum was an Aurelius, a family that had two consuls (father and uncle) while Caesar was growing up. You don't get to be consul and poor, and the Aurelians would not have let their daughter live in straitened circumstances, not least because it would reflect badly on the rest of them. Therefore, as with Caesar's magical sword, or the idea that he was born by Caesarian birth, or that he conquered Britain, I'm calling BS on this one. |
| 2025-07-03 |
| Guest Speaker |
| These days I work with my AI open alongside my text and notes (I have a 29" screen). Even after taking the time to double-check the AI's assertions - often against academic sources to which the AI has referred me - this allows me to work quicker and better. I was going to write a piece for this blog on how AI helps, but then it occurred to me that my AI can literally speak for itself. So take it away Arnie! Thanks Maty. I'll get right down to it: In the twenty-first century, the oldest chapters of human history are finding new narrators—not flesh and blood, but code and computation. Artificial Intelligence (AI), once the stuff of science fiction, has become a powerful tool in the hands of historians and archaeologists. AI's most visible footprint in the study of the ancient world lies in deciphering and restoring damaged texts. Projects like The Herculaneum Scrolls are harnessing machine learning to read papyri carbonized by the eruption of Mount Vesuvius—texts too fragile or obscured for the human eye. Algorithms trained on known handwriting styles and letter patterns can reconstruct missing pieces of inscriptions with remarkable accuracy, offering historians new lines to read and fresh interpretations to ponder. But AI doesn’t just help us read the past—it helps us map it. By analyzing satellite images, machine learning models can identify ancient roads, building foundations, and settlement patterns long hidden beneath modern landscapes or layers of earth. This is revolutionizing field archaeology, allowing researchers to plan more efficient digs, and even uncover sites that would have otherwise remained buried forever. Language and translation also benefit. Neural networks trained on ancient languages—like Latin, Ancient Greek, or Sumerian—are providing scholars with faster, more consistent translations and aiding in the linguistic reconstruction of languages with few surviving texts. Critics of AI in the humanities rightly point to the risks of over-reliance or the potential for error. Yet, when used in tandem with trained historians, AI becomes a powerful partner—an analytical companion that brings speed, scale, and surprising insight to the slow, patient work of academic research. Far from replacing the human touch, AI is expanding it—giving modern historians and archaeologists new tools to illuminate what once seemed lost to time. If Thucydides or Herodotus could have imagined the future of history, they might be astounded not only at what we’ve discovered, but at who—or what—is helping to do the discovering. |
page 1 page 2 page 3 page 4 page 5 page 6 page 7 page 8 page 9 page 10 page 11 page 12 page 13 page 14 page 15 page 16 page 17 page 18 page 19 page 20 page 21 page 22 page 23 page 24 page 25 page 26 page 27 page 28 page 29 page 30 page 31 page 32 page 33 page 34 page 35 page 36 page 37 page 38 page 39 page 40 page 41 page 42 page 43