About this translation

These are early Buddhist discourses in clear, everyday English. What makes this edition unusual is not any single sentence. It is the system underneath.

One word, one meaning

Most translators work one passage at a time. They meet a Pali word, choose an English word that fits that sentence, and move on. Over a few hundred pages, one Pali term can end up with four or five English faces. None is wrong, but none is quite the same.

This project fixes the rendering first. Every important term has a record. The record gives the preferred English, the allowed alternatives and where they apply, the renderings we rule out, and the reason. When a text uses the term, the translation uses that rendering. If a passage truly needs something different, we write down the exception and the reason.

There are now more than a thousand of these records.

Formulas stay formulas

These texts repeat themselves, and the repetition is not padding. People memorised and recited them before anyone wrote them down. A phrase that comes back usually does structural work: it is the same move, made again, and the text wants you to notice.

So we translate each recurring formula once and reuse it. When a stock line appears in eight discourses, it reads the same way in all eight.

We also keep the repetition inside each text. It can feel strange if you usually read translations that trim it. After a few pages, most readers find that the rhythm is part of how the teaching works.

The reasons are written down

Every translation on this site has a companion document in the repository. It records what we decided and why: which renderings we chose, which we rejected, what was ambiguous, and what we left open on purpose.

That last part matters. Sometimes the Pali is truly unclear, or a later tradition settled a question that the text leaves open. Then the translation does not quietly pick a side. It renders what is there and records the question.

Checks, not memory

Nobody can hold the consistency of a collection this size in their head. So scripts check it automatically. They confirm that renderings match their records and that shared formulas have not drifted. They confirm that each source we cite contains the passage we cite it for. They also check that the English has not slipped back into the stiff, ceremonial style these translations avoid.

These checks have caught real mistakes many times. In several cases the translation was right and the record behind it was wrong.

The reader is generated, not hand-copied

The pages you read here are not separate copies of the translations. A program builds each page from the governed source text. So when an editor corrects the source, the correction appears here without anyone copying it by hand. The newcomer guides live apart from the translations, and the program keeps them when it rebuilds the pages.

We do this on purpose. A reading edition that drifts away from its own source is worse than no reading edition.

Source and corrections

We check the Pali against the Chaṭṭha Saṅgāyana Tipiṭaka of the Vipassana Research Institute. We use it, with thanks, under its terms for non-commercial use. Every sutta page names the book its text comes from. It links to that edition and to the translation notes in the repository.

We drafted the translations published before 10 November 2026 against SuttaCentral's Mahāsaṅgīti edition. The project no longer uses SuttaCentral's texts, in line with the use requests in their publication records.

Every published page has passed the automated checks for terms, sources, structure, and readability. We record human read-aloud and newcomer reviews when they happen. They are optional, and the pages do not show them as a status.

Each page shows when its text last changed and a short version code. If something looks wrong, use the page's Report a problem link, so your report names the exact sutta.

The translations on this site carry the CC BY-NC-SA 4.0 license. You may share and adapt them for non-commercial use, with credit, under the same license. The site's code is under the MIT License. See the repository's content license for details.

Plain English, on purpose

The English aims to be direct and easy to take in. Not "one abides contemplating the body as a body, ardent, clearly comprehending", but something you can read at ordinary speed.

That does not simplify the content. Several of these texts are truly difficult, and the translation does not pretend otherwise. It only refuses to add a second difficulty on top of the first.

When a word does technical work, we explain it instead of smoothing it away. Open the words-used panel on any sutta, or see the glossary. You do not need to hover or use a mouse to read these definitions.

Reader feedback

On some pages you can tell the editors where the reading got hard. A small "Give feedback" control follows each passage. The words-used panel asks whether an explanation helped. A short, optional review at the end of the page asks what you took from it. You do not need an account, any Pali, or a better wording to suggest.

What happens to your feedback:

• It goes to a private queue that only the translation's editors read.
• The editors use it to find wording and explanations that readers stumble on, and to decide what to revise. When readers stumble on the same word in several texts, that shapes the shared term decisions.
• The system stores it with the exact passage you saw and the version of the text at that time. That lets the editors check a later revision against it.
• Nobody publishes it or posts it to the public repository, and it never changes a translation by itself. An editor reads it, decides, and records why.
• Nobody asks for your name, email address, or an account. The system sends what you type, the page and passage, and the time. It goes through a Google Form that the editors run, and Google's own rules for form responses apply. The editors copy it into their private queue and delete it from the form after review.
• If you take part in a session with a facilitator, your answers carry the anonymous label the facilitator gave you. That keeps one person's answers apart from another's.

Public feedback does not count as a formal review. Formal newcomer reviews happen in sessions with a facilitator, and the editors record them separately.

Reading it yourself

The editorial layer is all public: the term records, the translation notes, the checks, and the reasoning.

github.com/timedrapery/shiny-adventure

The repository is the working system. This site is what it produces.