Case Study 2 — Aroon's Curry Paste, Twelve Years of Iteration

A signature, slowly arrived at

Mae Som — The Mother's Kitchen — is a Thai restaurant in the Annex neighborhood of Toronto. Aroon Sornprasit opened it in his late thirties, after seventeen years of cooking in Bangkok, Hong Kong, Vancouver, and Toronto. The restaurant is twenty-two seats. The menu is short. The kitchen is a galley wide enough for two cooks who like each other.

The signature dish at Mae Som is, depending on whom you ask, either Aroon's grandmother's gaeng kiew wan (green curry) or a plate Aroon invented in his thirties that is not in any Thai tradition. The grandmother's curry is the answer most journalists get when they ask. The other dish is the answer Aroon gives when the conversation has gone on long enough that he stops giving the journalist answer.

This case study is about the green curry. It is about how a recipe handed down from his grandmother became, over twelve years of running his own kitchen, a recipe that is still legibly his grandmother's curry but has been adjusted, in small and specific ways, to the kitchen, the ingredients, the dining room, and the cook he has become. It is, in a different register, the same story as Maya's jollof — a recipe held in another cook's body that has been transcribed into the cook's own kitchen by sustained iteration.

The case study is told, with Aroon's permission, with details he was willing to share. Some things — the grandmother's specific proportions, the small private flourishes — he asked be left out, and they have been. The point is the method, not the recipe.

The grandmother's curry, as it came to him

Aroon's grandmother was named Khun Yai Bua. She lived in Chiang Mai. She fed three generations of Sornprasits and an indeterminate number of neighbors out of a courtyard kitchen that was, Aroon says, no larger than the back of a transit van. She made green curry once or twice a week. Her recipe was held entirely in her hands and her tongue. Aroon learned the curry by watching her from age eight to eighteen, and by making it under her supervision from age sixteen to nineteen, when he left for culinary work in Bangkok.

The curry, as Aroon learned it, had several elements:

• The paste: green chilies (a specific Thai variety, sometimes substituted), shallots, garlic, lemongrass, galangal, kaffir lime zest, kaffir lime leaves, cilantro root, white peppercorns, shrimp paste, salt — pounded for about thirty minutes in a granite mortar.
• The cook: coconut cream cracked first to extract the oil; paste fried in the cracked oil; coconut milk added; meat or fish or vegetable; fish sauce, palm sugar; Thai basil at the end.
• The serve: with jasmine rice, never the brown variety, always slightly al dente.

Khun Yai Bua's specific proportions — how much chili, how green, how much shrimp paste — were sometimes different from week to week, depending on what was at the market and what kind of week she was having. The dish was never exactly the same twice. It was recognizably the same dish each time, because the structural moves were the same. The variation was in the small dial-adjustments.

Aroon, at nineteen, could make the curry to his grandmother's satisfaction. He could not yet make it to his own satisfaction in someone else's kitchen, with someone else's ingredients, for someone else's diners. That part took the next twenty years.

The twenty-eight ingredients problem

When Aroon opened Mae Som at thirty-eight, the green curry was the dish he most wanted to put on the menu. He could not, at first, get it right. The Toronto kitchen was different from the Chiang Mai courtyard in dozens of small ways:

1. The chilies were different. The Thai chilies he could source in Toronto were either frozen (a different texture, less aromatic), dried (the wrong dish entirely), or fresh from a small handful of importers who carried them inconsistently. Most of what he could get reliably was a Mexican-grown serrano, which approximated the heat but had a different flavor profile.
2. The galangal was different. The galangal sold in Toronto Asian markets was sometimes fresh, sometimes frozen, sometimes ginger mislabeled. The freshest galangal he could find was fibrous in a way that affected the paste's texture.
3. The kaffir lime zest and leaves were sometimes from kaffir limes (correct), sometimes from regular limes (wrong species, missing aromatic compounds).
4. The cilantro root — the part Thai cooking specifies and Western cooking discards — was almost never available in Toronto. He could sometimes get bunches of cilantro with the roots intact from the Vietnamese market on Spadina; otherwise he was substituting cilantro stem.
5. The shrimp paste was variable in quality. Aroon eventually settled on one specific Thai brand and would buy six months at a time when it came in.
6. The coconut milk was the biggest difference. Aroon's grandmother used fresh coconuts, cracked the morning of cooking, with the cream skimmed and the second pressing for the milk. The closest Aroon could get in Toronto was specific brands of canned coconut cream and milk. Some brands cracked properly when fried; some did not. Some had stabilizers (carrageenan, gellan gum) that prevented the cream-cracking step entirely. He had tested at least eighteen brands.
7. The palm sugar — the jaggery-like unrefined sugar from coconut palms — was available in Toronto but in different forms (puck, granulated, paste). The puck was closest to his grandmother's. He would weigh and grate it.
8. The fish sauce — same brand-test discipline. He had settled on a specific Vietnamese fish sauce (against his grandmother's Thai brand, which was hard to source reliably) because it had a more consistent flavor across batches. He tested every new shipment.

For each of these, Aroon had to decide: substitute and adjust, or hold out for the original ingredient at higher cost?

Each substitution rippled through the recipe. The Mexican serrano produced a different heat profile, so the chili quantity adjusted up. The fibrous galangal gave a coarser paste, so he adjusted the pounding time. The substituted coconut milk required a different cracking protocol — some brands needed to be reduced first to break the stabilizer; some needed not to be touched at all. He documented each substitution and the adjustment that made the substitution work. The result, after about three years of iteration, was a green curry that tasted like his grandmother's curry — if you had been at his grandmother's table, you would recognize it — but was technically a different recipe by the count of substitutions.

The recipe was, in his words, a translation. Not a forgery. Not an invention. A translation, with the same fidelity to the source you would expect from a translator of poetry — choosing words that preserve the meaning, the music, and the intent, even when the words on the page are not the same words.

The Toronto palate adjustment

The second adjustment was harder, because it was about the diner, not the ingredients.

Bangkok diners eat green curry with Thai chilies at full strength. Toronto diners do not. This is not a value judgment about Toronto diners; it is a fact about palate-acclimatization. Capsaicin tolerance develops over years of exposure. A diner who grew up eating Thai food at Thai heat levels has a different sensory threshold than a diner who has eaten chili-forward food only occasionally. A green curry served to a Bangkok regular at Aroon's grandmother's heat level is a perfectly enjoyable curry. A green curry served to a Toronto regular at the same heat level is, for many of those diners, painfully hot — the heat overwhelms the rest of the dish, and the diner cannot taste the lemongrass-galangal-lime-leaf aromatic complex Aroon has spent thirty minutes in the mortar building.

Aroon's choice was: serve the dish at full heat (his grandmother's authenticity), and accept that many Toronto diners would not be able to taste the rest of it; or serve at a lower heat that let the rest of the dish come through.

He chose the lower heat. About 60% of his grandmother's chili load.

He made the choice deliberately, after about two years of testing dishes at multiple heat levels with various diners. He landed at 60% because at that level, the heat was present and assertive — Toronto diners reading the menu's "spicy" indicator did not feel cheated — but the lemongrass, galangal, kaffir lime, basil, fish sauce, and palm sugar layers were all distinctly tasteable. The dish was, by his lights, his grandmother's intent served to a different audience. He included a note at the bottom of the menu offering the dish at full Thai heat for diners who wanted it, and about 5% of the dining room ordered that version.

This is, he will say if you ask, the most contested decision he has made about the dish. Some Thai diners have taken him to task for it — "you are a Thai chef cooking for Thais," one regular told him; "do not water down our food." Some white food critics have praised him for it — "approachable Thai," one wrote, in a review Aroon does not love. Aroon thinks about the criticism. He has not changed the choice. His view is that the intent of the dish — to deliver the full aromatic complex of his grandmother's curry to the diner's palate — is better served at the lower heat for the audience he has, and that pretending otherwise would be a different kind of inauthenticity than the one he is being accused of. He made the call. He stands by it. He keeps the dish at full heat available for those who want it.

The fish-sauce-and-acid late move

The third adjustment was the late acid.

Aroon's grandmother finished the curry with Thai basil and served it. Aroon, after some years, started adding a small splash of lime juice at the very end, off the heat — about a teaspoon for a single-portion curry. His grandmother did not do this. Aroon, at first, felt slightly guilty about the addition. He was changing his grandmother's dish.

He thought about it for a long time. He concluded the following: his grandmother's coconut milk was fresher than his. The fresh coconut had a brightness and a slight tang that the canned coconut lacked. The fish sauce in his grandmother's kitchen was made by a neighbor and had a different acid profile than the brand he could buy. The lemongrass and kaffir lime he was using had been sitting in cold storage longer than his grandmother's would have. All of these factors meant his curry started slightly less bright than his grandmother's. The teaspoon of lime juice at the end was the adjustment that brought his curry's brightness back up to where his grandmother's started.

In the framework of this chapter, Aroon was applying the late acid fix. A long-cooked dish loses volatile aromatics over time; a finishing acid restores brightness. The fix is documented in the chapter (and in McGee, and in many practicing chefs' notebooks). Aroon was doing the fix because, technically, his curry needed more of it than his grandmother's did, because his ingredients had a longer cold-storage history than hers had.

He still feels slightly guilty about the addition. He still does it. He thinks his grandmother, if she were watching, would taste the curry, taste the version without the lime, taste the version with, and approve of the lime. He cannot ask her. She passed in 2018.

What the case study illustrates

Aroon's twelve-year iteration of his grandmother's green curry is the chef's version of Maya's two-year iteration of her mother's jollof. The pattern is the same:

1. Receive a beloved recipe in another cook's body.
2. Reverse-engineer it through observation, repetition, and asking.
3. Substitute ingredients you cannot source, with adjustments to make the substitution work.
4. Adjust for context — the diner, the kitchen, the cook you are.
5. Make late-stage corrections for ingredient drift (the lime move).
6. Hold the original cook's intent as the touchstone, not the original ingredient list.
7. Document each change, with reasoning.
8. Be honest, in citation and in conversation, about what you are doing — translation, not forgery, not invention.

The case study illustrates, in addition, the working ethic Aroon articulated to Maya: the recipe belongs to where it came from; the technique is yours. Aroon's curry, in 2026, is still legibly his grandmother's. The substitutions and adjustments are visible only to the cook. To the diner, the dish is gaeng kiew wan. The menu calls it gaeng kiew wan, after Khun Yai Bua, Chiang Mai. The citation is on the menu. The lineage is acknowledged. The translation is made with care.

This is the working version of recipe design at the chef level — slower, harder, more accountable than home cooking, but the same fundamental loop. Hypothesis, experiment, observation, revision. Salt-fat-acid-heat plus umami-sweet-bitter. Substitution by function. Adjustment by context. Documentation. Citation.

What separates Aroon's twelve years from a journalist's "innovative chef" framing is that Aroon is, by his own lights, not innovating. He is translating. The distinction matters. Innovation, when applied to another tradition's dish, is often a quiet appropriation. Translation, done with humility and citation, is one of the deepest acts of respect a cook can perform on another cook's recipe.

Aroon's other dish — the one he invented

There is a second dish at Mae Som, a plate Aroon developed in his early forties that is on the menu under his own name. The dish is a slow-cooked beef cheek with a tamarind glaze, served on top of a black-rice porridge with a handful of crispy shallots and a drizzle of an oil he infuses with kaffir lime peel and white peppercorn. The dish uses Thai techniques (the tamarind reduction, the infused oil, the rice porridge form) but is not from any Thai tradition Aroon knows. He invented it.

The menu lists this dish as beef cheek, tamarind, black rice — Aroon Sornprasit, Toronto, 2019. The citation line is short and explicit. The dish is his. He created it, using techniques that came from his training, in a kitchen in a city that is not his hometown. He is willing to call it his own.

The clarity of the second dish makes the first dish's citation work even more visible. The two are presented side by side on the menu — the grandmother's curry with her name, his own dish with his name — and the diner can read, from the menu alone, the working ethic. The thing that came from somewhere is named where it came from. The thing he made is named where he made it.

This is, perhaps, the most useful thing any recipe-design framework can do: not just produce good food, but produce good food with the citation work intact.

Analyze this

1. The 60% heat decision. Aroon's choice to serve his grandmother's curry at 60% of her chili level for Toronto diners is presented as deliberate, considered, and contested. (a) Is it a defensible choice? (b) Where would you draw the line — is there a reduction (40%? 20%? 0%?) that would feel like a betrayal of the dish, even with the option to order the full version? (c) How is this different from an Italian-American restaurant serving a "Marinara" that has been sweetened for American palates?

2. The lime adjustment. Aroon adds a finishing lime that his grandmother did not, with the reasoning that his ingredients arrive less bright. (a) Is this faithful to his grandmother's dish or a departure? (b) The chapter argued that "the recipe is a hypothesis." Does the late-lime move qualify as iterating the hypothesis or as departing from the original? (c) Where is the line between faithfully translating a recipe and rewriting it?

3. The substitution rippling. Each ingredient substitution Aroon made required adjustments elsewhere — different chilies meant a different chili quantity, different coconut milk meant a different cracking protocol. (a) For your own kitchen, identify three substitutions you make routinely and trace the ripple. What other adjustments are you making (consciously or not) to compensate? (b) What does this say about how portable a printed recipe is across kitchens?

4. The two-dish menu. Aroon's menu has both his grandmother's curry (cited to her, by name) and his own invented dish (cited to him, by name). The juxtaposition is, in itself, a position-taking on the question of authorship. (a) What does the menu's citation work do for diners? (b) Could a less-careful chef use the same citation form to legitimize a dish that is not really translated, just lifted? How would you tell the difference?

5. The grandmother who is gone. Aroon's grandmother passed in 2018. He cannot ask her about the lime. He cannot ask her about the chili reduction. He has to make the calls himself, with her cooking-in-his-body as his only consultation. (a) What does it mean for a chef to be the carrier of a tradition whose source has died? (b) What is the carrier's responsibility — to preserve, to translate, to evolve, all three? (c) Is there a parallel from your own life — a recipe carried by you that was given by someone now gone — and what is your felt responsibility toward it?