Peter Mark Roget did not begin his thesaurus because he wanted to publish a famous book. He began because words sometimes failed him.

In 1805, the 26-year-old physician completed a small catalogue in which words were grouped by the ideas they expressed. He kept it close, consulted it while writing and added to it over the years. It was an aid for his own moments of verbal hesitation, not yet a commercial project.

Nearly half a century later, that private tool became Thesaurus of English Words and Phrases. Roget published it in 1852 at 73, an age when the Victorian world expected a distinguished physician and scientist to be winding down rather than launching the work that would carry his name into the next century.

The familiar version of the story says that retirement gave him twelve years to build the book. Broadly, that is true: Roget left medical practice in 1840 and the first edition appeared in 1852. The fuller history is more interesting. He had been building and using the underlying system since his twenties, and his public duties did not end when his patients stopped arriving.

A young doctor begins sorting the language

Roget was born in London in 1779 and studied medicine at the University of Edinburgh, receiving his degree at only 19. He worked as a physician in Manchester and later London, lectured on anatomy and physiology, investigated disease at Millbank Penitentiary and helped report on the capital’s water supply.

He was not a physician who happened to dabble in words. He had the broad, methodical curiosity common among some nineteenth-century scientific figures. He wrote on physiology, electricity, magnetism and phrenology, devised mathematical and chess problems, and presented the Royal Society with an account of an optical illusion produced by viewing a moving wheel through vertical openings.

The catalogue nevertheless had an unusually long hold on him. In the preface reproduced in a digitised 1859 edition of the thesaurus, Roget said that he had completed a small “classed catalogue of words” in 1805. He described it as scanty and imperfect, but useful in his own literary composition.

He repeatedly considered enlarging it. The size of the task, combined with what he called a “multitude of other avocations,” kept putting him off. That makes the project feel surprisingly modern: a useful private file, steadily amended, waiting for years to become the larger work its owner could already imagine.

Retirement happened in stages

Roget gave up medical practice in 1840, when he was 61. Yet this was not a clean withdrawal from professional life. He had been secretary of the Royal Society since 1827 and continued in that demanding position until 1848.

His own chronology is precise on this point. It was after retiring from the secretaryship, he wrote, that greater leisure allowed him to turn the private catalogue into a repertory for other people. The expanded work occupied him incessantly for its final three or four years and demanded far more labour than he had anticipated.

So the twelve years between medicine and publication should not be pictured as one uninterrupted stretch at a desk. They were the final span in a much longer history. The idea began in 1805, medical work ended in 1840, the Royal Society role ended in 1848, and the book arrived in 1852.

That sequence matters because it changes the meaning of Roget’s late achievement. He was neither a 73-year-old beginner nor a man who merely dusted off an old notebook. He carried a working system through several careers, then used the time opened by retirement to make it large enough for strangers.

There is a similar temptation to flatten other productive lives into a single number. ScienceBlog’s earlier look at Charles Darwin’s roughly four hours of concentrated daily study found that the famous schedule leaves out correspondence, observation and the long continuity of his work. Roget’s “twelve-year retirement project” needs the same sort of qualification.

Roget built a dictionary that runs backwards

An ordinary dictionary starts with a word. You look up retirement, for example, and learn its definitions, pronunciation and history. Roget wanted to solve the reverse problem: you have the idea but not the word.

His thesaurus therefore did not begin as the alphabetical list of synonyms many readers know today. It began as a classification of concepts. Roget divided what could be expressed into six large classes: Abstract Relations, Space, Matter, Intellect, Volition and Affections.

Those classes divided into sections and then into numbered categories. Existence sat beside non-existence. Agreement was placed against disagreement. Hope had a conceptual neighbour in hopelessness. A reader could enter through the alphabetical index, find a category number and then browse a family of nouns, verbs, adjectives, adverbs and phrases.

The arrangement reveals how ambitious the project was. Roget was not merely collecting substitute words. He was trying to draw a navigable map of ideas expressible in English.

He explained the difference neatly. A dictionary takes a given word and finds its meaning. His work would take a given idea and find the words “most fitly and aptly” able to express it. The alphabetical index was the doorway; the real book was the conceptual structure behind it.

That distinction also guards against a common misuse. The words grouped under one heading are not necessarily interchangeable. They may differ in strength, tone, grammar or implication. Some are associations rather than strict synonyms, and Roget deliberately placed opposing concepts near one another. The book offers candidates. Judgment still belongs to the writer.

The thesaurus was a tool for thinking, not just decorating prose

Roget understood word choice as more than ornament. He believed language helped organise thought itself. Reviewing related expressions, he argued, could bring collateral ideas into view and let a writer see a subject from a new angle.

That claim should not be inflated into the idea that a larger vocabulary automatically produces deeper thinking. ScienceBlog has previously examined research showing that broad vocabulary captures only one part of what intelligence tests measure. Knowing an obscure word is not the same as having a sound argument.

Still, Roget had identified a real cognitive nuisance. A thought can be present before its label is available. We may know the emotional temperature of a sentence, the direction of an argument or the shape of an object while the precise term remains just beyond reach.

A thesaurus externalises part of that search. Instead of repeatedly demanding that memory produce the missing word, the reader moves through a neighbourhood of meaning. One entry sparks another until something fits.

The system also reflects a fact explored in ScienceBlog’s recent coverage of information density across roughly 1,000 languages: language structure influences how meaning is packaged and navigated. Roget was not studying language comparatively in that modern sense, but his book rests on the same broad recognition that meaning has structure.

At 73, the first edition was only another beginning

The title page of the 1852 edition presented the book as a practical aid, “classified and arranged so as to facilitate the expression of ideas and assist in literary composition.” Readers evidently agreed that the problem needed solving.

A Royal College of Physicians biography, originally published in 1878, recorded that the first edition appeared in 1852 and a 33rd edition had been issued by 1875. Roget continued revising the work during his lifetime. He died in 1869 at 90.

The book did not remain fixed after him. His son, John Lewis Roget, revised and expanded it, and his grandson Samuel Romilly Roget later continued the work. Editors added newer language, removed terms that had faded and adapted the organisation for different readers.

This is what the familiar statement that Roget’s Thesaurus has never gone out of print should mean. It is not one untouched Victorian text travelling intact through 170 years. It is a continuing editorial lineage. The title, principle and family name survived through repeated alteration.

There have also been competing branches: international editions, concise editions, student editions and versions rearranged into dictionary form. “Roget’s” became so closely associated with the thesaurus as a type of book that the name can now feel almost generic.

That history makes its endurance more impressive, not less. A reference book survives only by changing because language does. The 1852 vocabulary could not simply be sealed behind glass if the work was to remain useful to writers living amid technologies, institutions and social relationships Roget never knew.

The achievement was late; the practice was lifelong

Roget’s story is often told as an encouraging tale about retirement, and it is one. A person can publish the work for which they are best remembered at 73. A long professional identity does not exhaust the possibilities of later life.

But there is a second lesson that may be more honest. The publication was late; the practice was not. Roget began arranging words at 26, returned to the catalogue through decades of medical and scientific work, and knew from personal experience why the system might help another writer.

He did not spend 47 years following a smooth master plan. By his account, he repeatedly postponed the larger undertaking because it looked too big. Only after his obligations narrowed did he give the expansion his sustained attention.

That slow path explains something about the finished book. Its organising idea had been tested not only by scholarship but by use. Roget had reached for the catalogue when he was stuck, noticed what it lacked and kept adding what he needed.

More than 170 years later, the underlying problem has not gone away. We still find ourselves holding an idea without its word. Roget’s lasting achievement was to understand that the gap could be mapped, and to spend a lifetime drawing the map.