Thoreau's Journal Was the Real Work, Not Walden

2026-10-02 · Yesterself

Most people meet Henry Thoreau through Walden, a book of eighteen chapters that takes a few evenings to read. Behind it sits a manuscript perhaps forty times its length, which he worked at for twenty-four years and which almost nobody read while he was alive. He called it his journal. It is fair to say the book is the by-product and the journal is the work.

Twenty-four years of it

The first entry is dated October 22, 1837. He had just finished at Harvard, was twenty years old, and opened with a question someone had put to him, usually credited to Emerson: was he keeping a journal? So he began one that day.

The last entry is dated November 3, 1861. He was ill by then with the tuberculosis that killed him the following May, and the entry is about wind, and about the way a storm had left gravel and vines lying in a particular direction along the railroad causeway. It is not a farewell. It is a note on how sand behaves. That is characteristic.

In between he wrote on most days. The whole thing runs to something like two million words. When it was finally printed in 1906 it filled fourteen volumes. Most of the surviving manuscript notebooks are now at the Morgan Library in New York, and Princeton has been publishing a careful scholarly edition of them, volume by volume, for decades.

The turn toward observation

The early journal reads the way you would expect from a young man with literary ambitions. Extracts from his reading, arguments with himself, sketches for lectures, the occasional set piece written to be admired.

Around 1850 it changes. The entries get more particular and much more dated. He starts recording when each plant came into flower, when the ice went out of Walden and Fair Haven Bay, when the first bluebird arrived, how high the river stood, how deep the snow lay, the temperature of springs and brooks. He walked most afternoons over the same few square miles of Concord and wrote down what that ground was doing that day.

He also worked in two stages, and this is the part worth stealing. In the field he carried a small notebook and a pencil and made brief notes, because a notebook in a pocket survives rain and stiles and a desk does not travel. In the evening he wrote the fuller entry from those notes, while the day was still close enough to reconstruct.

Later he began pulling the years together into charts, month by month, trying to compose a portrait of an average Concord year out of his own accumulated dates. He left that project unfinished. Some of the late manuscripts that grew out of it, on seed dispersal and on wild fruits, were not published in full until the 1990s.

Walden was cut out of it

He lived at the pond from July 4, 1845 to September 6, 1847, two years and two months. Walden was published in 1854, seven years after he moved out. The book was not written at the cabin. It was assembled and rewritten afterward, largely at a desk in his family's house in town, and a great deal of its raw material came out of the journal.

J. Lyndon Shanley went through the surviving manuscript drafts in the 1950s and sorted them into a series of distinct versions written across those years. The picture that emerges is not of a book that arrived whole. It is of a writer returning again and again to a body of notes he had kept for other reasons, lifting out of it what the current draft needed.

You can watch the process at the level of a single subject. The thawing sandbank in the deep railroad cut near the pond appears in the journal across several different springs, each time in raw form, the sand doing its particular thing on a particular date. In the book those separate springs are compressed into one long passage in the chapter called "Spring," and it becomes one of the most worked-over pieces of prose he ever wrote.

What migrates from journal to book is rarely a finished paragraph. It is usually a dated observation, an exact detail, a verb he found on the day, a phrase that turned out to be better than he knew when he wrote it down. The journal was where he had the luck. The book was where he noticed it.

Working the same way

The lesson is not that you should write two million words. It is that the useful journal and the polished piece are two different acts, and confusing them ruins both.

Repeat the place, not the topic. Thoreau's material came from walking the same ground for twenty years. A single visit gives you a description. Forty visits to one pond, one park, one street of trees, gives you a record, and a record can be questioned.

Write particulars, and number them where you can. Not "the pond is thawing" but the date, the extent, how far out the open water reached. Numbers and dates are what you cannot reconstruct later from memory. Conclusions you can always add.

Keep a firsts list. First frost, first swallow, first cherry blossom, first morning you needed a coat. It costs one line each and it is the entry you will most want in four years.

Take the notes twice. Pencil in the field, longer entry that evening. Both halves matter. Field notes alone stay cryptic. Evening entries alone drift toward mood and away from fact.

Let the entry be raw. He did not write the journal to be read. That is exactly why it was worth quarrying. If every entry is already performing for an audience there is nothing left in it to discover later.

Reread on the anniversary. He indexed his own volumes and went back to the same date in earlier years. This is the habit most people skip, and it is where the compounding happens: an observation is only data once there is a second one to set beside it.

That last habit is the hard one, because rereading requires you to go and get the old page. It is also the habit a journal on a phone can carry for you. Yesterself is built around exactly that motion. Each morning it brings you one letter out of your own past writing, an entry from this day in an earlier year, a note from yesterday, or a sentence you postmarked forward to today, and your reply that evening becomes tomorrow's mail. It is not released yet, and the only thing to sign up for is the waiting list.

What it was for

He published two books in his lifetime and neither sold well. The journal, the thing he actually did every day for twenty-four years, sat in its notebooks and was read properly only long after he was dead.

That should be encouraging rather than sad. The journal earned its keep while he was writing it, as the place he did his looking. The book was one thing he happened to make from it. Ecologists reading his dates a century and a half later were another. Whatever you take out of your own record later is a third, and you do not have to know what it is in order to start keeping one.

The first entry is always the same size. His was a few lines long.

Common questions

How long did Thoreau keep his journal?

The first entry is dated October 22, 1837, when he was twenty, and the last is dated November 3, 1861, about six months before he died. He wrote in it on most days across those twenty-four years, and the whole runs to roughly two million words.

Is the journal just an early draft of Walden?

No. Walden was made partly out of journal material, but the journal came first, kept running long after the book was published, and became something the book never was, a dated natural history record of one town.

What did Thoreau actually write down?

Early on, reading notes, arguments and drafts of lectures. After about 1850, more and more dated observation, when a plant flowered, when the ice went out of the pond, when a bird arrived, river levels, measurements, along with the thinking those things set off.

Where can I read it?

The fourteen-volume 1906 printed edition is out of copyright and easy to find online. Princeton has been issuing a scholarly edition for decades. Most of the surviving manuscript notebooks are at the Morgan Library in New York.

Do I need to write every day for a nature journal to be worth anything?

No. Thoreau's own record has gaps in it and survived them. What makes an observation journal useful is returning to the same place and dating what you see, not an unbroken run of days.