Case Study 1 — Aroon's Grandmother and the Right Color

A Conversation in a Toronto Kitchen

This case study is built from a long evening I spent in early 2025 at Mae Som ("The Mother's Kitchen"), the Thai restaurant Chef Aroon Sornprasit owns in Toronto's Annex neighborhood. He had just come off a service shift, the dining room was emptying out, and we sat at a small table near the kitchen pass, drinking tea. The conversation wandered, as it often does, but the heart of it was about his grandmother, his mother, and a teaching that had taken him thirty years to translate.

Aroon's grandmother was Pikul Sornprasit. She was born in 1923 in a village outside Chiang Mai, in northern Thailand. She did not have formal schooling past primary level. She ran a kitchen at home and, for fourteen years, helped her sister run a small noodle stand in the local market. She cooked a kaeng — a curry — almost every day for sixty-five years. She died in 2003.

When Aroon was about eight years old, he remembers his grandmother in her kitchen one afternoon, slowly cooking a curry paste. The ingredients were dried red chilies, lemongrass, galangal, garlic, shallot, kaffir lime, coriander root, white pepper, and shrimp paste. She had pounded them all in a large stone mortar. Then she had heated coconut cream in a wok until it cracked — that is, until the water in the cream evaporated and the fat separated out as small bubbles. She added the pounded paste and began to slowly fry it.

Aroon remembers her stirring constantly. He remembers her watching the paste change color — from the bright red of the fresh chili paste to a deeper red, then to a darker reddish-brown, then to a mahogany so dark it was almost black at the edges. He remembers her saying — in Thai, which I have translated here approximately — "You wait for the right color."

"What color?" young Aroon asked.

"This color. Now."

She pointed at the paste. He looked at it. He had no language for the color. To him it was just brown — like other brown things in his life, like roof tiles or his uncle's coat or wet earth. She didn't have language for it either. She had only the act of pointing to this color, right now.

Then she added more coconut cream and the broth ingredients, and the curry came together. Aroon, who would be a chef thirty years later, did not consciously remember the lesson at the time. But it sat in him.

The Lesson Repeated

When Aroon was fourteen, his mother was teaching him to make a different curry — gaeng massaman, a southern Thai curry with influences from Persian and Indian traders. Different paste, different aromatics, but the same step: cook the paste until si ti mai — "the right color" — appeared.

Aroon's mother stood next to him at the wok and watched as he stirred. She would say, "More." He would stir. "More." He would stir. "Stop. Now."

He pulled the paste off the heat. She nodded.

"How did you know?" he asked.

"You see it. You will see it."

He did not see it. But he kept practicing.

By the time he was nineteen, immigrating to Canada and starting to work in restaurant kitchens, he could almost see it. Sometimes he would pull the paste off too early; the resulting curry would taste flat, with a sharp raw chili note that masked the other flavors. Sometimes he would pull too late; the paste would have a bitter edge, with charred notes that did not belong. The window for "the right color" was narrow, and his eye was not yet calibrated.

By the time he was twenty-five, working at a fine-dining French restaurant during his classical training, he understood "the right color" without thinking about it. He could cook a curry paste, an onion soubise, a duck reduction, a beurre noisette, a brunoise of shallot for a stew base, and know — in his hand, in his eye, in the smell coming up from the pan — when each had reached its endpoint.

He didn't know the word "Maillard" until he was thirty-two.

The Discovery

In 2010, Aroon was reading Harold McGee's On Food and Cooking, which a culinary friend had given him for his birthday. He encountered the chapter on browning reactions. He read about Louis-Camille Maillard. He read about amino acids and reducing sugars and Schiff bases and Amadori products and Strecker degradation and melanoidins. He read about pH effects on rate, and about the importance of dry surfaces, and about the temperature thresholds.

He told me he stopped reading three times to put down the book and walk around the kitchen.

What he had been doing for fifteen years, with no name for it, was running the Maillard reaction. The "right color" his grandmother had taught him was the density of melanoidin polymer in the curry paste. The "right color" his mother had refined in him was the same. He had been internalizing the chemistry, in his eyes and his hand, without ever using the word.

This is, in fact, exactly how cooking traditions have transmitted Maillard expertise for as long as there has been cooking. There were no chemists at Aroon's grandmother's village stand. There was just a woman who had learned, from her own mother, what a properly-cooked curry paste looks like. The chemistry had a name in 1912. The color had a name in northern Thailand long before that. Both names referred to the same molecules.

Aroon told me: "I read the science and I think — okay, this is what I have been doing. I am happy because someone has named it. I am also a little sad, because I think people will sometimes confuse having the name for the thing with knowing how to do the thing. My grandmother could not tell you the chemistry. But she could cook a better curry than most chemists."

He thought for a while.

"And then I think — actually, both are good. The chemistry is real. Her cooking is real. I am lucky to know both."

What Aroon's Grandmother Was Doing

Let's translate Aroon's grandmother's curry-paste cooking into the chemistry of this chapter.

When she pounded the paste in the mortar, she was creating massive surface area on the chilies, aromatics, and shrimp paste — exposing amino acids and sugars from cellular structures that had been bound up in plant tissue and protein. The pounding was, in effect, mechanical preparation for the chemical reactions to come.

When she cracked the coconut cream — heating it until the water evaporated and the fat separated out — she was creating a hot oil environment. Coconut cream is mostly fat (about 65–70%) with water and proteins. When the water evaporates, the local temperature in the wok can climb above 100°C, into the Maillard regime. Coconut oil itself has a smoke point around 175°C, plenty of headroom.

When she added the paste to the hot fat, she was bringing amino acids and reducing sugars (from the chilies, garlic, shallot, lemongrass, and other ingredients — all of which contain free amino acids and natural sugars) into contact with surfaces above 140°C. The Maillard reaction began running immediately.

She stirred constantly because: 1. Stirring distributes heat evenly — every part of the paste gets the right thermal exposure. 2. Stirring prevents local hot spots from charring (entering the pyrolysis regime, which produces bitterness). 3. Stirring exposes new surface to the hot fat, allowing more of the paste's volume to participate in Maillard.

She watched the color because the color was her readout of how far the reaction had progressed. The progression went:

• Bright red (raw chili paste, no Maillard) →
• Slightly duller red (early Maillard, some color change) →
• Deep red-brown (substantial Maillard, melanoidins forming) →
• Mahogany (advanced Maillard, deep flavor compounds developed, complex character) →
• Almost black at edges (warning sign — pyrolysis beginning, time to pull off the heat).

She was looking for the moment between mahogany and almost-black — the deepest possible Maillard short of the burning regime. This is also, in modern terminology, what we might call "the optimal endpoint of the Maillard cascade for this particular paste." She didn't have those words. She had the eye.

Si ti mai — "the right color" — is, chemically, the Maillard endpoint where melanoidins are dense, Strecker products are abundant, and pyrolysis has not yet begun.

What Maillard Gives a Curry

Here is what changes in a Thai curry between the under-cooked and properly-cooked paste stages:

Under-cooked paste: Chili harshness dominates. The other aromatics (lemongrass, galangal, lime leaf) are present but bright and sharp. The shrimp paste is fishy and slightly raw. The curry tastes flat — like a list of ingredients, not a unified flavor.

Properly-cooked paste: All the harsh notes have been mellowed. The chili is now warm and complex. The aromatics have melded with each other through Maillard products that bridge them. The shrimp paste contributes a deep umami foundation. The whole thing tastes like a curry, not like a list of ingredients. The flavor is layered, with high notes (citrus from lime leaf), middle notes (warm spice), and low notes (deep brown, savory).

Over-cooked paste: Bitterness emerges. The aromatics taste burnt. The chili tastes harsh in a different way (charred, not fresh). The flavor has lost some of its complexity because the late-stage Maillard products have given way to pyrolysis products.

The window between under-cooked and over-cooked is, in Aroon's experience, about 30–60 seconds. His grandmother knew this from forty years of practice. He learned it from twenty years of practice plus eventually reading the science. The chemistry tells us why the window is narrow (the Maillard cascade transitions through its peak quickly at high temperatures, and pyrolysis follows close behind). It does not change the practical question of how to find the window in a particular wok with a particular paste on a particular day.

For that, you need to cook many, many curries.

The Generalization

What Aroon's grandmother was doing with curry paste is what cooks all over the world have been doing forever:

• A French cook making a roux watches it go from pale to ivory to peanut to brown to dark, pulling it at the desired endpoint.
• A Mexican cook making a recaudo roasts chilies, tomatoes, and onions on a comal (griddle) until they're charred at the edges and deeply roasted.
• An Indian cook preparing a bhuna base cooks onions, ginger, garlic, and spices in oil for 30+ minutes until deeply colored.
• An Italian cook making soffritto slowly browns onion, carrot, celery, watching for the right caramelization on the edges.
• A Japanese miso-maker (over months, not minutes) lets the paste develop deeper color through slow Maillard at room temperature.
• A coffee roaster watches color, smell, and "first crack" timing to pull beans at the right roast level.
• A Korean cook making doenjang jjigae uses doenjang (fermented soybean paste) that has spent months developing its color and complexity.

Different ingredients. Different temperatures. Different time scales. The same chemistry, the same skill: an internalized model of the Maillard reaction's progression, calibrated against an outcome that has been refined over generations.

⚖️ Analyze This

1. Aroon's grandmother and his mother both refused to give him a time for cooking the curry paste. They said only "watch the color." Why might cooking time be a poor instruction even though we now know the chemistry depends on temperature and time?

2. If a beginning cook were given a thermometer and instructed to "hold the paste at 165°C until it turns mahogany," would they make a better or worse curry than someone using only the eye? Argue both sides.

3. The narrow window between "right color" and "burnt" — say 30–60 seconds — has a specific chemical explanation. What is it?

4. A chef new to making this kind of paste might be tempted to cook at higher heat to save time. What would the trade-offs be in terms of flavor profile and risk of burning?

5. Aroon's grandmother could not, with the language she had, distinguish Maillard products from caramelization products. Does it matter for her cooking that she didn't make this distinction? Does it matter for Aroon's cooking that he can?

6. Translate Aroon's grandmother's lesson into a teaching script for a new line cook in Aroon's restaurant. What sensory cues, what decision rules, what examples would you give? Compare your script to a script that uses thermometers and timers.