How retranslating an 1881 French classic revealed 140 years of hidden distortion
By Emma-Rose Elizabeth
Some scientific laws become so fundamental that we forget someone had to discover them. Ribot’s Law is one of those laws. Formulated in 1881 by French psychologist Théodule Ribot, it states that memory dissolves in a predictable order: recent memories are lost first, older memories last; complex memories before simple ones; voluntary before automatic. It is one of the oldest findings in psychology that has never been overturned. In 145 years, no evidence has contradicted it. No competing law has replaced it.
When a doctor explains why an Alzheimer’s patient can sing childhood songs but cannot remember this morning, that is Ribot’s Law. When a neurologist assesses traumatic brain injury by mapping which memories survived and which didn’t, that is Ribot’s Law. When researchers at MIT or DeepMind study why artificial neural networks catastrophically forget recent training while retaining earlier patterns, they are describing the same hierarchical dissolution that Ribot documented in clinical patients before the invention of the telephone. The standard model of memory consolidation, proposed by Larry Squire in 1984 and still the dominant framework in neuroscience, was built specifically to explain the mechanism behind Ribot’s observation. The law came first. The theory followed a century later.
No other law of memory has this reach. Ebbinghaus’s Forgetting Curve of 1885, the only comparable principle from the same era, describes how normal memories fade over time. Ribot described how pathological memory breaks apart, and in doing so revealed something profound about how memory is built: in layers, from simple to complex, from automatic to voluntary. Understanding how it dissolves tells us how it was constructed. That insight, 145 years old, remains the foundation of clinical memory science.
And yet, for all its influence, Ribot’s original text has been largely inaccessible to English-speaking readers. The only English translation, produced in 1882 by William Huntington Smith, went out of print generations ago, and as I discovered over three years of retranslating the book from the original French, it was never accurate to begin with.
What I Found
The problems fell into patterns. The most damaging involved terminology. Ribot used precise French terms that carried specific theoretical weight, and Smith replaced them with English words that meant something different.
Take conscience. In Ribot’s materialist framework, this means consciousness: a biological epiphenomenon, a shadow that accompanies nervous processes. Ribot says it plainly: la conscience n’en est que l’accessoire (consciousness is only an accessory). Smith rendered conscience as ‘intuitive perception’. That’s not a synonym. ‘Intuitive perception’ implies innate knowledge, non-empirical awareness, exactly the kind of faculty psychology that Ribot spent his career dismantling. The English reader meets a different philosopher from the one the French reader meets.
Or take reconnaissance. Ribot uses this to mean recognition: the automatic act of identifying something as previously encountered. Smith translated it as ‘recollection’. But recognition and recollection are different cognitive processes. Recognition is simpler, more automatic, and more resistant to memory dissolution. Recollection is higher, more complex, and dissolves earlier. That hierarchy is Ribot’s Law. By collapsing the two terms into one, Smith erased the mechanism at the heart of Ribot’s theory.
Then there was the rhetoric. Ribot writes like a scientist who knows he’s right. His prose is declarative, assertive, and unapologetic. The very first line of Chapter 1 announces: “La mémoire est essentiellement un fait biologique, et accidentellement un fait psychologique.” Memory is essentially a biological fact, and only accidentally a psychological one. That’s radical materialism, stated without hedging. Smith softened it throughout: adding ‘perhaps,’ ‘it may be,’ ‘it would seem,’ and swapping active constructions for passive ones. The bold French scientist became a cautious Victorian gentleman.
Why This Matters for Translators
Every translator knows that fidelity is a spectrum, not a switch. But what I found in the Smith translation goes beyond the normal compromises of moving between languages. The distortions were systematic. Translation theorist Antoine Berman identified twelve ‘deforming tendencies’ that translators unconsciously impose on prose: rationalization, ennoblement, qualitative impoverishment, and the destruction of terminological networks. Smith’s translation exhibits nearly all of them. He wasn’t a bad translator. He was a product of his era. Victorian publishing norms valued fluency and elegance over fidelity. The translator’s job was to make the foreign text read as if it had been written in English. Lawrence Venuti calls this ‘domestication’, and Smith’s Ribot is a textbook case.
The result is that for 140 years, Francophone and Anglophone scholars reading the ‘same’ book were actually reading different books. Researchers who cited the English Ribot were, in important respects, citing a text Ribot did not write. The two scholarly communities drifted apart not because they disagreed but because they were, unknowingly, reading different books.
The 2026 Edition
My retranslation, Diseases of Memory: The Forgotten Classic That Launched a Science (Lafayette Publishing, 2026), was produced directly from the 1881 French first edition. I worked sentence by sentence against both the French original and Smith’s edition, classifying every departure. The retranslation renders conscience as ‘consciousness,’ reconnaissance as ‘recognition,’ and désagrégation consistently as ‘dissolution’ throughout, where Smith had fragmented it across multiple inconsistent equivalents. It includes approximately 12,000 words of original annotations connecting Ribot’s nineteenth-century observations to contemporary neuroscience, so that a modern reader can see how his clinical insights map onto what we now know about Alzheimer’s disease, procedural memory, and even catastrophic forgetting in artificial neural networks.
By my assessment, Smith’s edition achieves approximately 65% fidelity to the French original. The 2026 retranslation achieves approximately 97%. That thirty-point gap represents a century and a half of lost meaning.
A Lesson in What We Miss
Ribot’s case is probably not unique. Many foundational texts in psychology, neuroscience, and philosophy were first translated into English during the Victorian period under norms that prioritized elegance over accuracy. Scholars like Claudia Wassmann and Thomas Dixon have shown that these translation problems were systemic rather than isolated. How many other foundational texts are sitting on our shelves in English editions that quietly distort the original?
For translators, the lesson is both humbling and empowering. Humbling because it shows how deeply cultural norms can shape our work without our knowing it. Empowering because it shows that retranslation matters. It’s not redundant. It’s not merely updating the language. Sometimes it’s the only way to give a community of readers access to the book the author actually wrote.
Emma-Rose Elizabeth is a trilingual scholar (English, French, Mandarin) and Lead Assistant Researcher at the University of Texas at Austin. Her translation of Ribot’s Les Maladies de la Mémoire is the first modern scholarly English edition from the original French.
She has presented her work at TEDxUTAustin and Harvard Business School’s Harvard@SXSW, and has been accepted to present at the Cheiron Society’s 2026 annual meeting. She is a distinguished recipient and Ambassador of the Harvard Summer School program.
