How the book was made

Why these eleven countries

The scope is the region, not a selection from it. All eleven sovereign states of Southeast Asia are covered, including the four that general Southeast Asian cookbooks most often omit: Laos, Brunei, Timor-Leste and, in any depth, Myanmar.

Recipe counts vary by chapter and are not an editorial ranking. They reflect how much distinct, well-documented material each cuisine offered within a hundred recipes. Thailand carries fifteen and Brunei and Timor-Leste three each; the three were researched to the same standard as the fifteen.

Three kinds of entry

The book contains three different kinds of thing, and they are not interchangeable.

Recipe — 100 of them

A dish. Ingredients, method, yield, timings, and the context that places it: its name in its own script, its region, and how it varies between households.

Foundation component — 24 of them

A sauce, sambal, paste or condiment that dishes are built from. Also a recipe — ingredients and a method — but a building block rather than a meal, and shared across the chapter it belongs to.

Pantry study — 27 of them

Not a recipe. An article about an ingredient: where it is used across the region, what it tastes and smells like, how to choose it, how to store it and what can honestly stand in for it. There is no ingredient list and no method, because there is nothing to cook.

The site keeps them apart for the same reason the book does. A reader looking up fish sauce should not be handed something that looks like a recipe, and a reader looking for a sambal should not be handed an essay.

Research and sources

All text is original writing created for this project. No wording is taken from other cookbooks, blogs or recipe sites. Facts, techniques and cultural context drawn from research are rewritten in the author’s own words, and the sources consulted are recorded in the project’s research log.

Reference sources are consulted for facts only, never for text. Recipes are combinations of common technique and the author’s own development; dishes that belong to a shared regional tradition are credited to that tradition and are not claimed as invented here.

Where accounts of a dish’s history or origin conflict, the book says they conflict rather than picking the better story. Where a claim could not be supported, it was left out.

What verification means here

Three different things get confused under the word “verified”, so this site keeps them separate and never uses one to imply another.

Researched

Checked against published sources and written up. Done, for all the material on this site.

Fact-checked

Names, spellings, romanisations and cultural attributions confirmed, in some cases by native speakers. In progress. The pantry studies in particular are drafts, and say so on every page.

Kitchen-tested

Actually cooked, and the quantities and timings confirmed by cooking them. Not done for any recipe in this project.

Recipe testing status

Every recipe on this site — all 100 dishes and all 24components — is an untested development draft. The quantities and timings are the author’s working text, not a result confirmed by cooking.

The rule the project works to is simple: no recipe is described as kitchen-tested unless it has actually been cooked and that test has been logged. None has, so none is. Every recipe page says so.

Pantry verification status

All 27 pantry studies are drafts. Local names and cultural notes in particular are flagged for review, and are shown under a heading that says so rather than being presented as settled.

Romanisation across the region is not standardised. The same dish may be written several ways in perfectly good sources, and some languages here have no single agreed system. Where a form is provisional the text says so instead of asserting one it cannot yet stand behind.

Images

The photographs in this book are not photographs. Every image was produced with generative AI tools and then reviewed and approved by the author, one at a time, against what the dish should actually look like.

Review was substantive rather than a rubber stamp: images were rejected and regenerated where the food was wrong. What it is not is documentary evidence that a dish was cooked. The full account is on the imagery and AI disclosure page.

Why some fields are provisional

Some entries carry a field marked pending rather than a confident answer. That is deliberate. A cookbook that fills every field looks more finished and is less honest; where the project does not yet know, the site says it does not know.

The alternative — quietly asserting a romanisation, an origin or a timing that has not been confirmed — is the failure mode this project is built to avoid.

Corrections

Errors will be corrected. Recipes, names, attributions and cultural notes are all open to correction, and a correction that is supported will be made whether it comes from a reader, a native speaker or the author’s own review.

A public route for reporting one is not yet open. See contact for the current position, which is stated plainly there rather than dressed up as a form that would not reach anyone.