# A garden log that argues with itself

> Forty years of sowing dates, corrected in pencil against what actually came up — a record that only becomes useful once somebody disagrees with it.

August 20, 2026 · https://justog.club/stories/the-sowing-diary

A pale sage tee, pinned flat to a wall. A page from a ruled garden notebook, deep green, two narrow
columns in a steady old hand: a date on the left, a crop on the right. 3 Mar — broad beans. 17 Mar —
onions. 2 Apr — carrots. 21 Apr — runner beans. 6 May — sweetcorn. Beside two of the entries, in a
later and thinner pencil, three words: TOO EARLY.

The first hand wrote down what was planned. The second hand wrote down what actually happened, and
that's the entry doing the real work.

## Why is the pencil note worth more than the date it's correcting?

Because the date was a plan and the note is a result, and a plan on its own can't be checked against
anything.

3 Mar — broad beans is a decision made in advance, based on habit or hope. TOO EARLY, added later in
a different, thinner hand, is what the garden actually said back. A diary with only the first kind of
entry is a schedule. A diary with both is an instrument, because it now holds two things that can be
set against each other — what was tried, and what worked — and that comparison is the entire function
a record like this exists to perform.

## What makes forty years of entries different from four?

Range — a single year's result could be down to the season, not the date, and there's no way to tell
the difference from one data point alone.

Forty years of the same columns, kept the same way, turns "this year was late" into a claim that can
actually be checked against a run of other years rather than against a feeling. The pencil corrections
cluster in some years and not others, which is itself information — a season that needed correcting
everywhere was a bad season, not a bad plan. That's the same logic behind [the price written inside a
lid, checked against what things actually cost rather than trusted from a
sticker](/stories/the-price-written-inside-the-lid) — the record only earns its keep once it's
been checked against reality more than once.

## Does copying a page like this from an actual notebook need anyone's permission?

Yes, if it's someone else's real handwriting and real garden being reproduced — a personal notebook
is exactly the kind of thing worth asking about before it goes on a shirt, the way you'd ask before
photographing anyone's private papers.

An invented version, in the same ruled format and the same hand, carries none of that obligation,
because nobody owns the format of two columns and a date. Most gardeners keeping a record this long
are glad to see the method noticed rather than assumed to be nothing.

## Does the idea work for a record other than a garden?

Yes — anything repeated on a yearly cycle benefits from the same two-hand structure: a planned date
in one column, a pencil correction in another, once the year has had its say.

[Planting something for a reason that only pays off well after the planting is
finished](/stories/the-tree-he-planted-for-later) sits at the other end of the same patience —
one entry waits a season to be proven right or wrong, the other waits decades. Both only work if
somebody actually goes back and writes down what happened.

## Isn't this just nostalgia, since the weather's different every year anyway?

No — the weather changing every year is exactly why the record has to exist, not a reason to give up
on it.

A memory of "when things usually go in" gets rewritten by whichever year was most recent and most
vivid, which is the least reliable kind of evidence available. A written record with forty years of
corrections in it isn't claiming the weather repeats — it's holding enough actual years that this year
can be placed among them fairly, rather than compared to whatever's easiest to remember.
[A balance that compares two masses instead of reading off a dial](/stories/the-brass-weights)
runs on the same principle: trust the comparison, not the impression.

## Does it print well?

Yes, and the two-hand contrast is the entire point, so the pencil correction needs to sit visibly
lighter and thinner than the ink columns rather than blending into them.

Keep the ruled lines running off both edges rather than boxed in — a hard border around a page like
this reads as a picture glued on, where letting the rules bleed off the fabric keeps it reading as a
page rather than a frame.

## How do I get one made?

Describe the record at [JustOG](/) — what's tracked, how the correction reads, how many years are
behind it. Pick a direction, drag the crop frame, see it composited on the real garment, and it's
made to order and shipped.

Designs other people have published are in the [shop](/shop).

The next entry goes in pencil, in case this year has something to add too.