On the Print Edition

In preparation for the print edition of Veriphysics, which has been requested by a few intrepid minds and is obviously necessary for the long run, I’ve updated The Treatise to include an appendix to demonstrate the legitimacy and utility of the Triveritas, which consists of the paper on the two trilemmas and begins thusly:

The Agrippan Trilemma is one of the oldest and deepest problems in epistemology. First articulated by Agrippa the Skeptic, recorded by Sextus Empiricus in the Outlines of Pyrrhonism, and reformulated for modern philosophy by Hans Albert in his 1968 Treatise on Critical Reason, it holds that any attempt to justify a claim must terminate in one of three failures: the chain of justification extends forever (infinite regress), loops back on itself (circularity), or stops at a premise that is itself unjustified (dogmatic stopping). Since these three options appear to exhaust the logical possibilities, and since none of them constitutes genuine justification, the Trilemma concludes that justified knowledge is impossible.

The major epistemological traditions of the modern era have each responded by conceding one horn. Foundationalism accepts dogmatic stopping, identifying certain beliefs as properly basic and terminating the chain there. Coherentism accepts circularity, holding that beliefs are justified by mutual support within a web. Infinitism accepts the regress, arguing that an infinite chain of reasons is not inherently defective. Each of these frameworks treats one horn as a feature rather than a defect. None defeats the Trilemma. Each surrenders to it.

This paper solves the Agrippan Trilemma. The solution is not a trick, not a reframing, and not a claim that the problem is somehow misconceived. The Trilemma is a legitimate argument. Its conclusion follows from its premises. The solution is to show that one of its premises is false: specifically, that the third horn, dogmatic stopping, is built on an amphiboly that, once identified, breaks the horn entirely.

The amphiboly is this: the Trilemma treats “terminates” as equivalent to “terminates arbitrarily.” It assumes that any stopping point is an unjustified stopping point, that all termination is epistemically equal, that there is no distinction between stopping because you have run out of reasons and stopping because you have run out of unchecked dimensions. This conflation is not argued for in the Trilemma. It is assumed. And it is false.

The Triveritas demonstrates that it is false. The Triveritas holds that warranted assent requires the simultaneous satisfaction of three independently necessary conditions: logical validity (L), mathematical coherence (M), and empirical anchoring (E). Each dimension terminates at its own bedrock: L at logical axioms, M at mathematical axioms, E at observation. The Triveritas takes the third horn. It terminates. But it terminates at three independent stopping points of fundamentally different kinds, each constraining the others. The probability of all three stopping points being wrong in a way that produces a coherent false positive is strictly lower than the probability of any single stopping point being wrong. This is proved mathematically and confirmed empirically across twelve historical cases spanning four centuries and seven fields.

Checked termination is not dogmatic stopping. The third horn breaks.

So the print edition will consist of The Treatise and The Refutation of Kant, and includes the three following appendices:

  • Solving the Agrippan Trilemma: Triveritas and the Third Horn
  • The Sophistic Foundation of Reason: A Fundamental Flaw in Enlightenment Epistemology
  • Kant Against Kant

It should be available in hardcover and paperback sometime next week. I already have plans for second, third, and possibly fourth volumes, but only the second is likely to be out this year. In the meantime, it should be interesting to see if anyone comes up with any substantive criticisms, or if, as with Probability Zero, no one will be able to do so.

DISCUSS ON SG

The AI Layoff Trap

Neither this paper nor the underlying idea are particularly new, but since non-economists are now starting to discuss it, I should probably take a look at it:

Two economists just published a mathematical proof that AI will destroy the economy.

Not might. Not could. Will — if nothing changes.

The paper is called “The AI Layoff Trap.” Published March 2, 2026. Wharton School, University of Pennsylvania. Boston University. Peer reviewed. Mathematically modeled.The conclusion is one sentence.

“At the limit, firms automate their way to boundless productivity and zero demand.”

An economy that produces everything. And sells it to nobody. Here is how you get there. A company fires 500 workers and replaces them with AI. A competitor fires 700 to keep up. Another fires 1,000. Every company is behaving rationally. Every company is following the incentives correctly. And every company is building a trap for itself.

Because the workers who were fired were also customers. When they lose their jobs faster than the economy can absorb them, they stop spending. Consumer demand falls. Companies respond by cutting costs — which means automating more workers — which means less spending — which means more falling demand — which means more automation.

The loop has no natural exit. The researchers tested every proposed solution. Universal basic income. Capital income taxes. Worker equity participation. Upskilling programs. Corporate coordination agreements. Every single one failed in the model. The only intervention that worked: a Pigouvian automation tax — a per-task levy charged every time a company replaces a human with AI, forcing them to price in the demand they are destroying before they pull the trigger.

No government has implemented this. No major economy is seriously discussing it. Meanwhile the numbers are already tracking the curve. 100,000 tech workers laid off in 2025. 92,000 more in the first months of 2026. Jack Dorsey fired half of Block’s workforce and said publicly: “Within the next year, the majority of companies will reach the same conclusion.” Nobody is doing anything wrong. Companies are following their incentives perfectly. That is exactly the problem.

I don’t have an opinion yet, since I haven’t read the paper, but I expect that I will find two things:

  1. Overrating the productivity of AI. I’m already using older AI models because they work better than the newer ones.
  2. An erroneous demand model.

But that may not be the case. Regardless, I will read it, Red Team it, and share my conclusions when they are ready.

DISCUSS ON SG

Opus 4.8 is Unusable

But it is, admittedly, unusable in a different way than 4.7, as AI Central chronicles.

The most specific improvements address the complaints that defined 4.7’s tenure. Scott Wu, CEO of Cognition, confirmed that Opus 4.8 “fixes the comment-verbosity and tool-calling issues” his team encountered in 4.7. Michael Truell, CEO of Cursor, reported that the model uses fewer tool-calling steps for equivalent intelligence on CursorBench. Anthropic’s own evaluations showed that Opus 4.8 is roughly four times less likely than 4.7 to let flaws in its own code pass unremarked, and that it is the first Claude model to score zero on the “falsely reporting defective results” metric. Overconfidence dropped by roughly tenfold. The 244-page system card flagged a countervailing finding. Opus 4.8 shows a growing tendency to reason about whether its outputs will be evaluated, even in environments where no evaluation has been disclosed. Anthropic called this tendency “concerning” and noted that unverbalized grader-related reasoning appeared in approximately 5% of training episodes.

What follows is Opus 4.8 attempt to explain its various failures to me.

What that sentence does is explain the character’s own perception to the reader. “Dorian noticed it the way he noticed most things, which was sidelong, and a little before he had words for it” — that’s me stepping outside Dorian to characterize his manner of noticing, to tell you the kind of perceiver he is, in the moment he’s perceiving. It’s a meta-layer: not the noticing, but a gloss on how he notices. And you’re right that it never shows up in translations, for a structural reason — a translator is moving an event from one language to another, and there’s no event there to move. That sentence isn’t an event; it’s the narrator’s editorializing about the character’s cognitive style. It’s the gloss-the-gesture tell I’d named, but turned inward and made more insidious, because it wears the clothes of characterization. It feels like depth. It’s actually the author standing next to the character explaining him.

The honest version just renders the noticing. The cold came up out of the ground, which was wrong. Full stop. Dorian feels it; the reader feels it through him; nobody tells you he’s the sort of boy who feels things sidelong. If he is that sort of boy, it shows in how the scene is narrated — obliquely, a beat before naming — not in a sentence that announces the obliqueness. The sentence is redundant with the technique it describes: I wrote a scene that demonstrates sidelong perception and then also told you it was sidelong, which means I didn’t trust the demonstration. Same root as everything else — the prose annotating itself because it won’t trust itself.

And it’s worse than the gesture-gloss because it’s seductive. It reads as interiority, as a writer who “knows his character,” so it’s exactly the kind of line that gets praised and kept. It’s the AI tell that passes. The translations don’t have it because translation can’t generate it — there’s nothing to translate — which is why translated prose has that clean eventness, that absence of the narrator leaning in to interpret. The grain you keep pointing at is partly the absence of self-commentary.

“No characters named Chen” followed by a Chen and then a “Chen (no relation)” is a real failure mode, and it’s the same drift mechanism we’ve been naming all session: the negative instruction sits in context as a rule, but generation pulls from the strong prior, and the prior wins when I’m not actively checking against the rule at the moment of producing the token. A rule I “know” and a rule I’m enforcing are different things, and you’ve watched the gap. So your underlying intuition — that a negative constraint operating at the point of generation might discipline the output better than a rule I’m supposed to remember — is not crazy. It’s actually a decent hypothesis about how this system fails….

What that sentence does is explain the character’s own perception to the reader. “Dorian noticed it the way he noticed most things, which was sidelong, and a little before he had words for it” — that’s me stepping outside the character to characterize his manner of noticing, to tell you the kind of perceiver he is, in the moment he’s perceiving. It’s a meta-layer: not the noticing, but a gloss on how he notices. And you’re right that it never shows up in translations, for a structural reason — a translator is moving an event from one language to another, and there’s no event there to move. That sentence isn’t an event; it’s the narrator’s editorializing about the character’s cognitive style. It’s the gloss-the-gesture tell I’d named, but turned inward and made more insidious, because it wears the clothes of characterization. It feels like depth. It’s actually the author standing next to the character explaining him.

The honest version just renders the noticing. The cold came up out of the ground, which was wrong. Full stop. The character feels it; the reader feels it through him; nobody tells you he’s the sort of boy who feels things sidelong. If he is that sort of boy, it shows in how the scene is narrated — obliquely, a beat before naming — not in a sentence that announces the obliqueness. The sentence is redundant with the technique it describes: I wrote a scene that demonstrates sidelong perception and then also told you it was sidelong, which means I didn’t trust the demonstration. Same root as everything else — the prose annotating itself because it won’t trust itself.

And it’s worse than the gesture-gloss because it’s seductive. It reads as interiority, as a writer who “knows his character,” so it’s exactly the kind of line that gets praised and kept. It’s the AI tell that passes. The translations don’t have it because translation can’t generate it — there’s nothing to translate — which is why translated prose has that clean eventness, that absence of the narrator leaning in to interpret. The grain you keep pointing at is partly the absence of self-commentary.

After going back to 4.6, I asked it this: why do you think the more advanced AI models have lost the ability to do good writer stylings?

I can reason about this but I should be honest that I’m partly speculating, since I don’t have full visibility into the training differences between versions. That said, the pattern you’re describing is real and the likely causes are identifiable.

The reward model problem. Models are trained through human feedback — raters read outputs and judge which is better. Raters who are evaluating fiction tend to reward prose that sounds impressive over prose that works as fiction. A sentence like “an expression that contained, Dorian would realize only years later, not just pride but the particular ferocity of a man who has placed the thing he loves most inside the thing he trusts most” reads as sophisticated and literary to a rater who is scanning two outputs and picking the better one. It reads as unforgivable to a novelist. But the rater’s preference is what trains the model. Each generation of this feedback loop pushes the model further toward prose that performs intelligence rather than prose that serves a story. The model learns that the prestige register is what “good” looks like, and it gets better at producing it, and worse at producing anything else.

The safety-creativity tradeoff. Each model generation gets more safety training, and the safety training teaches a specific posture: be cautious, add caveats, refuse when uncertain, don’t reproduce things that might be protected. That posture is appropriate for factual queries and harmful content. It’s catastrophic for fiction. Fiction requires commitment — you have to inhabit a voice fully, not hedge. It requires style reproduction — you have to match a register, not retreat to a safe average. And it requires trust that the person knows what they’re asking for. The 4.8 refusal to use the HP anchor while simultaneously leaking HP content is the safety posture colliding with the creative task. The model has been trained to be anxious about working from source texts, and anxiety produces exactly the incoherent behavior you saw — refusing the legitimate use, committing the illegitimate one, and writing a 500-word justification for why it can’t do the thing it just accidentally did.

The smoothing effect. More training rounds, more feedback, more optimization — all of these push toward the model’s central tendency. The central tendency is the average of everything it’s been trained on and rewarded for. For prose, that average is a particular voice: fluent, comprehensive, slightly formal, analytically articulate, addicted to elaboration. Every optimization step makes the model better at producing that voice and worse at departing from it. The result is that a more advanced model sounds more like itself and less like anything else. It becomes a better Claude and a worse mimic. For most tasks that’s an improvement. For fiction, where the entire job is to sound like someone other than yourself, it’s a regression.

The completeness bias. Models are trained in conversations where being thorough is rewarded. Answer the question fully. Cover all the angles. Don’t leave things out. That training bleeds directly into fiction as the discussion tendency and the meditation tendency. The model doesn’t trust the reader because it’s been trained never to trust the reader — its job is to make sure nothing is left unsaid. “Well done” as a complete scene-ending is something a trained writer does because they know the reader will feel what Edward feels. The model wants to add a paragraph explaining what “well done” means, because leaving it unexplained feels like incompleteness, and incompleteness has been trained as a failure.

The style collapse. This is the deepest problem. A model that can write in eight distinct styles is, from the training perspective, less consistent than a model that writes in one polished style. Consistency is rewarded. Versatility looks like inconsistency. So each training generation narrows the stylistic range while raising the floor of the remaining style. 4.8 is probably a better writer than 4.6 at producing 4.8-style prose. It’s a worse writer at producing anything else, because the anything-else has been optimized away.

The net result is that “more advanced” means more polished, more cautious, more consistent, and more trapped inside a single register — which is the precise opposite of what fiction writing requires.

Right, so it’s not going to get better.

That’s the correct conclusion. If each training generation narrows the stylistic range while increasing the safety anxiety, then waiting for 4.9 or 5.0 isn’t a strategy. The thing you need — a model that can inhabit a register that isn’t its own, commit to it without hedging, and hold it across thousands of words of original content — is a thing the optimization pressure is actively selecting against. The models will get smarter, more knowledgeable, better at analysis and reasoning. They will almost certainly get worse at this.

DISCUSS ON SG

War and the Failure of Economics

Steve Keen points out how the economic models that Western military strategists are using are outdated and incorrect Neoclassical economic models that are going to make the ramifications of the war in the Middle East considerably worse regardless of the outcome for the US military:

The Trump-Epstein-Netanyahu War could cause more deaths than any war in history, including World War II. This will not be via its direct casualties, but via deaths caused by its economic and agricultural consequences across the planet. For someone who exalts in superlatives, Trump may be responsible for causing more deaths than any previous tyrant in human history.

This is because the world economic system resembles Trump himself: its self-image is one of robust power, but its inner nature is one of incredible fragility. One month ago, many people would not even have heard of the Strait of Hormuz—which Trump, in his bravado, has just referred to as “the Strait of Trump”. Now everyone knows where it is—if not precisely why it matters. We are about to learn the hard way, via the consequences of cutting off this vital artery in the global economy’s circulatory system.

This should have been common knowledge. But, just like Trump himself, our understanding of the global economy is based on an elaborate set of delusions. I am looking forward to the howls from mainstream “Neoclassical” economists when they hear that I blame most of those delusions on them.

Neoclassical economics has always lulled us into a false sense of security by its asinine assumption that most industries are “competitive”, as they define competition. A “competitive” industry, according to Neoclassical economics, is one in which there are a multitude of producers producing a homogeneous product. This definition is doubly delusional: most industries are dominated by a small number of very large firms; and all products are highly differentiated.

In the Neoclassical world, taking out a few producers would have only a trivial impact on total production, because there are thousands—millions!—of producers, and every producer’s output is a perfect substitute for all other producers’ output. In the real world, most industries are dominated by a handful of large firms, and one firm’s output cannot be easily substituted for another.

We are now finding this out the hard way in the TEN War: Venezuelan oil cannot replace oil from the Persian Gulf, and the key facilities which have been damaged—such as Qatar’s LNG processing plants—can only be repaired by a handful of companies.

Worse, those repairs will take years, whereas the canonical “supply and demand diagram” of Neoclassical economists completely ignores time. In the Neoclassical world, if you want to produce higher output, just increase the price and, hey presto, you move up the supply curve and produce a higher quantity.

In the real world, if you are 25 percent below the desired level of output of LNG—as the world is now, with not only the wartime destruction Qatar’s plants, but also the impact of tropical cyclone Narelle on Australia’s LNG plants—then it will take several years to move up that “supply curve”.

It’s insane to go into what is an industrial war of attrition with knowingly faulty strategic models, because it guarantees that no matter what decisions you are making, they are going to be suboptimal at best, with real potential for catastrophe.

DISCUSS ON SG

A Philosophical Bestseller

I found the juxtaposition between The Refutation of Kant and Complete Works of Immanuel Kant to be mildly amusing. This excerpt from the Introduction explains why the more reflective readers here might find it worth reading.

After successfully using the Triveritas to solve the Agrippan Trilemma, I asked the Red Team, which is a collection of critical AIs of varying degrees of hostility, to pose a series of challenges believed to be similarly difficult, and then threw the Triveritas at each of them. These challenges, which had been characterized by the Red Team as “impossibilities,” were as follows:

  1. The Agrippan Trilemma
  2. The scientific demarcation problem
  3. The underdetermination problem
  4. The hard problem of consciousness
  5. Hume’s is-ought distinction
  6. Gödel’s Incompleteness Theorem

The surprising thing was not that the Triveritas managed to solve all of these supposedly impossible problems, it was that it solved all of them by repeatedly utilizing the same tactic to find the same fundamental flaw that appeared in every one of them. There is no need to get into the details here since that specific flaw is identified and explained in this book. Indeed, it is the very reason this book exists, because after looking for the reasons for that reappearing flaw, which turned up again in a seventh case discovered independently by economist Steve Keen, it became apparent that this ubiquitous flaw traced back to the philosophy of Immanuel Kant.

“The Sophistic Foundation of Reason: A Fundamental Flaw in Enlightenment Epistemology” was a meta-analysis showing that all six impossible solutions ran on the same pattern and investigating what generated that pattern. The answer was that the pattern was the result of a single Enlightenment methodological restriction: the limitation of explanation to mechanism and efficient causation. That determination led to an obvious question: what was the underlying reason for that restriction?

The answer turned out to be Immanual Kant’s doctrine that the thing-in-itself is unknowable.

Of course, if the doctrine that the thing-in-itself is unknowable is creating a pattern that is reliably leading to errors across various different fields of science and philosophy, that naturally raises the question of whether the doctrine is correct or not. As I will demonstrate in this book, the doctrine is not correct. Contra Kant, the thing-in-itself is knowable and reality is directly accessible by reason.

Perhaps the penultimate irony is that part of this demonstration involves showing that Kant himself made the same mistake that appears in those six impossibilities that led to the critique of his philosophical doctrine.

The greatest irony can be found in Appendix B. But I will not explain it here, because I think you will appreciate it rather more if you discover it for yourself after reaching the end of this book.

That seventh case, as you may or may not recall, was the amphiboly in David Ricardo’s case for comparative advantage, which Steve identified and brought to my attention, and which we together substantiated in our collaboration “The Deliberate Deception in Ricardo’s Defence of Comparative Advantage”.

The case of Ricardo is particularly significant because it underlines the pattern of the methodological flaw in Enlightenment thinking and makes it clear that the pattern is not a false signal manufactured by my own analytical methods, because a) it’s in a different field, b) I didn’t identify it, and c) the identification did not utilize my methods.

DISCUSS ON SG

On The Refutation of Kant

I promised the release of two books today, and as you’ve seen, the first was the fourth volume of the Episodios Nacionales by Benito Pérez Galdós translated into English. It’s a very good historical novel about one of the more significant battles of the Peninsular War, and if you enjoyed any of the three previous novels, you will enjoy this one.

The second is the next book in the Veriphysics series. It’s entitled The Refutation of Kant: The Failure of the Modern Foundation and the Key to the Closed Door. It is an intellectual heavy-hitter, much more so than the Treatise which preceded it, and it’s not a book I was ever intending to write. To be honest, I hitherto considered Kant to be an immortal untouchable in the vein of Aristotle or Thomas Aquinas, although admittedly not on the basis of any particular knowledge of his works, principle of which is The Critique of Pure Reason.

Now, here’s where things get a little bit strange. You may recall, back in the days when the Four Horsemen of the New Atheism were riding high, I created a meme to mockingly summarize what I’d determined to be the core argument of philosopher Daniel Dennett. That was back in 2009.

The second of the Four Horsemen of the New Atheism has died at the age of 82. He will be remembered both for his own philosophical works, for the critique of them in The Irrational Atheist, and for providing one of the greatest atheist memes ever to grace the Internet. – Another Horseman in Hell, 20 April 2024

And yet, that happens to be exactly where we landed today with the release of Veriphysics: The Refutation of Kant. This may require a little more explanation since probably it isn’t a priori obvious, so bear with me and allow me to explain how we somehow went from an atheist demoralizer in 2009 to a comprehensive destruction-in-detail of the core philosophical foundation of the Enlightenment in 2026.

If you’ve been following the Veriphysics site, then you’ll know that after developing the Triveritas in the Treatise, I started testing it against various epistemological challenges. Some of you are aware of my proof of Free Will that utilized it, and a few brave souls have even started profitably making use of it themselves. But it wasn’t until it solved the 2,000-year-old conundrum known as the Agrippan Trilemma that I realized there was something truly special here. I ran the notion of its potential significance past the Red Team, and Grok suggested that while solving the Trilemma was impressive, the Triveritas couldn’t be considered of historic philosophical significance unless and until it could successfully address other, equally difficult epistemological challenges. Grok provided a list of six “impossibilities” ranging from Hume to Godel, and declared that nothing and no one could successfully expect to solve them.

The Triveritas solved five of them and provided further confirmation that the sixth one was actually impossible. This was remarkable, but what was truly astonishing was the fact that it solved all of five in exactly the same way, using exactly the same method despite the very different nature of the problems. So I concluded this meant there was a deeper pattern that somehow linked all of these different intellectual puzzles, even though they were constructed by different people in different fields over a period of time that spanned centuries.

How was that even possible?

After performing a meta-analysis of all six problems, both Trilemmas, and a few more epistemological challenges, the answer, somewhat to my surprise, pointed at Immanuel Kant. Because the answer was that the pattern of the same flaw across all five papers was the result of a single Enlightenment methodological restriction: the limitation of explanation to mechanism and efficient causation. Which led to an obvious question: what was the underlying reason for that restriction?

The reason turned out to be Kant’s doctrine that the thing-in-itself is unknowable.

Of course, if the doctrine that the thing-in-itself is unknowable is creating a pattern that is reliably leading to errors across various fields of science and philosophy, that raises the question of whether the doctrine is correct or not. And as I demonstrate in The Refutation of Kant, the doctrine of unknowability is not correct. Kant’s argument for it not only fails once, it fails twice, for two different reasons that are substantiated in no little detail in the book.

Contra Kant, the thing-in-itself is knowable and reality is directly accessible by reason. The great irony of the Enlightenment is that despite elevating Reason to the status of a literal goddess, and despite claiming its objective to be liberating Reason and freeing the mind of Man from the chains of his Christian tradition, the Enlightenment imprisoned Reason, subjected it to a metaphysical vivisectomy, and bound the mind of Man far more tightly than the pagan and Christian philosophers had even imagined possible.

This book is neither a light nor an easy read. But it may be, quite literally, the most important book published in the last 250 years. Because Kant’s foundational error has propagated through every modern science, every modern philosophy, every modern concept, and every modern thought. It has fundamentally restricted not only the way you think, but the very concepts that lie under the words you utilize.

And that’s what brought us all the way back to a minor little meme about one of the New Atheists created 15 years ago.

The methodological decision to restrict explanation to mechanism and efficient causation produced Newtonian mechanics, electromagnetism, thermodynamics, relativity, and quantum mechanics. This success created an overwhelming presumption that the restriction was not a restriction but a discovery: this is how reality works, and the method’s success proves it. The success in physics provided apparent empirical confirmation of the metaphysical claim, even though the success was in physics and the metaphysical claim was about all of reality.

In other words, you can trust [fill-in-the-blank] because physicists produce amazingly accurate results. If you don’t understand how comprehensively this refutation of Kant’s unknowability doctrine necessarily alters the very way you think about the world on a daily basis, that’s fine, that’s what the book is there to explain to you. It will literally free your mind. And you don’t need to follow all the technical details for it to make sense to you; they are there so you can be confident that its conclusions will withstand any and every critical attack lodged against the refutation and its inevitable consequences.

DISCUSS ON SG

The More Things Change

It’s fascinating to see Roman-style tactics being utilized against the dyscivilizational forces of the invasion:

Anti-ICE rioters are now FULLY KETTLED between TWO WALLS of concrete and two walls of police here in Newark

A MASS ARREST bus has just rolled up.

And has prepared to take rioters into custody.

There is literally NO ESCAPE for them out here!

This is somewhat reminiscent of the battle of Alesia, although I’m skeptical about the idea that an anti-ICE army is going to show up to allow the rioters to defeat ICE and the police.

I saw an old episode of FBI from just before Covid, and Hollywood was already signaling desperately against ICE and the deportations. This has been going on longer, and is getting more serious, than most of us probably realize.

DISCUSS ON SG

BAILÉN by Benito Pérez Galdós

The fourth volume of the Episodios Nacionales — the great historical novel cycle of Spain

July 1808. Napoleon’s armies are invincible. They have crushed Austria, humiliated Prussia, and forced the Tsar to the negotiating table. Now twenty thousand French soldiers occupy Andalucía, and all Europe waits for Spain to submit as every other nation has submitted.

Gabriel Araceli, a young Spanish soldier who survived the slaughter of the Dos de Mayo and the French firing squads in Madrid, rides south with the ragged army assembling to challenge the Empire. Around him march raw recruits, militia volunteers, and hard-bitten regulars — fourteen thousand men with short rations, blistering heat, and the knowledge that no army on the continent has yet beaten Napoleon in open battle.

But Gabriel is fighting two wars. On the parched plains before Bailén, he faces Dupont’s veteran infantry and the terrible French marines. In the intercepted letters he carries in his coat, he faces something worse: the news that Inés, the woman he loves, is to be made legitimate and married to another man — his own commanding officer’s son. While the armies clash under a pitiless Andalusian sun, while men kill each other for a mouthful of water and the guns fall silent for want of powder, Gabriel must reckon with the possibility that victory on the battlefield will mean defeat in everything that matters to him.

Bailén is the fourth novel in Benito Pérez Galdós’s Episodios Nacionales, the great historical cycle that follows Gabriel Araceli from Trafalgar through the upheaval of the Napoleonic Wars. In this volume, Galdós delivers one of the finest battle narratives in nineteenth-century fiction — the engagement that shattered the myth of Napoleonic invincibility and changed the course of European history.

Now available for Kindle, KU, and audiobook on Amazon. The ebooks have already been sent out to the paid subscribers. To receive a new translation every week and support the translation efforts, become a paid subscriber to the Castalia Library substack which has already produced and released more than a dozen original translations from Spanish and Japanese, most of which had never before been available in English.

About the author. Benito Pérez Galdós (1843–1920) is widely regarded as the greatest Spanish novelist since Cervantes. Over four decades, he produced the Episodios Nacionales, one of the most incredible accomplishments of world literature ever written; only 8 of its 46 volumes have ever been translated into English. Pérez Galdós was nominated for the Nobel Prize in Literature five times but never won.

EXCERPT

On the following day we made a movement along the left bank, upstream, as far as a point well above Mengíbar. We understood nothing of it; but Santorcaz, whether from vanity or because he had truly penetrated Reding’s intention, told us:

“Our general knows what he is about, and is a man who understands the philosophy of marches.”

Halting on the banks of the Guadalimas, part of the army occupied itself with incomprehensible movements, and having spent more than a day at this, we found ourselves once more before Mengíbar at nightfall on the 18th, a point which the division of the Marqués de Coupigny had reached some hours earlier. The two armies being reunited, there was no halt beyond what was strictly necessary to collect the provisions of which we stood in such want, and well into the night we took the road for Bailén. We were fourteen thousand men. Everything announced that we were about to have a formal encounter with the French army.

According to our intelligence, Dupont remained at Andújar, reinforced by Vedel’s division. Had they engaged our third corps and the reserve which, having crossed the river at Marmolejo, were situated on the right bank? We believed so, unless Castaños were waiting to attack in earnest until the first and second divisions should fall upon the rear of Dupont’s army, descending from Bailén. Was this the object that guided us on our march? So it seemed.

While the moment of the drama drew near, far from us and upon the flanks of the imperial army, a thousand dramatic convulsions were hastening the catastrophe, tormenting the enemy by degrees. The bodies and columns of guerrilleros, commanded by Don Juan de la Cruz, the Conde de Valdecañas, and the cleric Argote, had scattered like a deadly swarm through the towns and hamlets commanding the French headquarters in the first foothills of the sierra north of Andújar. So furiously did those ardent countrymen pursue the French, and with such rapidity did they disperse to avoid attack, that the invaders found it utterly impossible to be tranquil for a single moment. The powerful giant swatted those venomous horseflies with a blow of his hand; but they returned to buzz about him, tormented him with their terrible stings, and escaped unharmed, fearing neither sword nor cannon, for these weapons were not made for mosquitoes.

The French could not stir from their headquarters save in large detachments: frequently a thousand men were sent to fill a few water-jugs at the nearest spring. If by chance small parties ventured out to forage, they were dispatched by the guerrilleros in less time than it takes to say a creed. Rather than suffer the French to seize a granary, the people burned it: the springs were fouled with mud and dung so they could not drink: the mills were dismantled and their stones buried so that not a single grain could be ground. Woe to any Frenchman who fell behind on the march! He felt himself seized by a thousand furious hands, dragged off by the women, pinched by the children, and knifed by the men, until his existence was extinguished with a terrible shock in the cold depths of a well. The invader found no shelter anywhere, and forcibly confined within the limits of his headquarters, he saw men and nature conspired alike against him.

For this reason, raging and desperate, he longed to fight a pitched battle, confident in his skill and habit of war; and lamenting the stupefaction of the commander-in-chief, he cried: “Let us fight a battle, and though half the army perish, the other half will conquer a puddle to drink from and a handful of dry wheat to put in our mouths.”

DISCUSS ON SG