The Smell of Garlic Before It Turns

I have just put a heaping spoon of crushed garlic into a small puddle of warm olive oil in a low pan, and I am standing about three feet back from it, breathing through my nose. The garlic has been in for perhaps ten seconds. Nothing visible has happened yet — the cloves are still pale, the oil has the same look it had a moment ago, the pan looks like a pan with garlic and oil in it and very little going on. But the smell has moved. The first thing I caught was the raw, slightly green scent of just-cut garlic released into a warm room. A few seconds later that has been overlaid with a softer, sweeter note, the smell that garlic takes on when its sulphur compounds begin to convert into something less aggressive — the smell I think of as cooked garlic, even though it has not been cooked, exactly, just warmed past the point where the rawness goes. That sweet note is the one I am here for. It is also, in roughly the next twenty seconds, the one I will lose, because the same fat that is releasing the sweet smell is climbing toward a temperature where the garlic will turn — first to a deeper, nuttier brown, and then, within another few seconds, to a bitter, acrid smell that ruins the dish even if the colour has not gone fully black yet. The window is short, and I have learned to find it with my nose before I find it with my eyes.

I do not pretend to have the chemistry of garlic in oil pinned in any precise sense. My best guess after years of starting dishes this way is that the smell shifts I am catching correspond to roughly three regimes. In the first regime, the garlic is below about 60 degrees centigrade, the sulphur enzymes are still active, and the smell is sharply raw — alliinase doing what alliinase does in a cool room. In the second regime, the temperature climbs into the range where those enzymes denature and the sulphur compounds begin to interact with the oil and with each other, and the smell turns soft and sweet — what I read as cooked garlic but is really not very far cooked at all, just past the rawness. In the third regime, the temperature climbs further, the natural sugars in the garlic begin to brown and the fat begins to oxidise around them, and the smell flips quickly from sweet to nutty to bitter as the surface of each piece passes through caramelisation and into burning. I would not swear to any of the temperatures and would not want to. What I would say is that the three regimes are real, that the transitions between them are short, and that the second regime — the sweet one — is the one I want and the one I have to catch in motion.

The pivot, then, is the nose. I have come to think of the first minute of any dish that starts with garlic in oil as a nose exercise above everything else. I am not watching the pan. I am breathing through it. The visual cue — the moment the garlic begins to take colour — is too late for me; by the time I can see the first hint of gold, the smell has already moved past the soft sweetness into the nutty register, and the bitter is a half-second behind it. I have learned to catch the smell at the moment the soft sweetness arrives, which is usually about ten to twenty seconds after the raw note has started to fade. The window between those two — the moment the rawness fades and the moment the colour starts to come — is when the next thing should go in. If the dish is meant to be garlic and tomato, the tomato goes in here, and the cold tomato drops the temperature of the pan and stops the garlic exactly where it should be. If the dish is meant to be garlic and white wine, the wine goes in here for the same reason. If the dish is meant to be garlic in oil as a sauce in its own right, the pan comes off the burner here, and the residual heat takes the garlic the rest of the way without browning it. The nose has done the timing; the next ingredient is the brake.

This matters because the most common failure of any garlic-based dish at home is the one where the garlic browns and the dish carries the bitterness through everything else. The cook puts garlic in oil, leaves the pan for thirty seconds to do something at the sink, comes back to find the garlic just starting to take colour, and decides — reasonably, on the surface — that this is roughly the right moment to add the next thing. The next thing goes in. The dish proceeds. But the garlic has already crossed the sweet regime by the time the colour was visible, and the bitter is now in the oil, and the tomato or the wine or whatever comes next has picked up that bitter and is carrying it forward. The finished dish tastes faintly off in a way the cook cannot place. The cook decides the recipe was wrong. The cook tries a different recipe. The garlic step is the same; the result is the same. The trouble is not the recipe. The trouble is that the cook has been watching for the visual cue when the nose has already been ringing for thirty seconds, and the dish has been doomed since the moment the colour began. The cook who learns to catch the sweet smell early — the cook who stands close enough to the pan to breathe its first minute — will not make this failure, and will not, for that matter, ever quite know how often other cooks are making it.

I should name the caveats. Not every garlic preparation cares about the sweet smell. Garlic that is going to be slowly confited in oil at a low temperature for half an hour does not need a sharp catch; the temperature is low enough that the bitter regime is far away and the cook has minutes rather than seconds of margin. Garlic that is going into a long-simmered stew, where it will be diluted by the volume of the pot and cooked for hours, does not need a precise catch either; whatever bitter notes form in the first minute will be smoothed by the rest of the cooking. Garlic that is being fried until it is genuinely brown and crispy — a Cantonese-style garlic crisp, for instance — needs the cook to ride past the sweet regime on purpose and to stop at a precise nutty colour just before the bitter. These are different exercises with different rules. The principle holds in the case I have described — garlic in oil at the start of a fast dish — and the nose is the instrument for that case more than for any other. None of these break the principle, which is that the smell is the clock and the eye is, at best, a confirmation that the cook has already noticed something the nose noticed earlier.

I have my next ingredient ready at the side of the pan because I planned for the half-second to be short. The soft sweet smell arrives. The tomatoes go in, all at once, cold, with a small hiss as they hit the warm oil. The temperature drops. The garlic stops where it was. The smell of the pan changes again, this time into the bright thin scent of tomato meeting hot oil, and the dish moves into its next phase. I do not have to think about the garlic again. I caught it where it was sweet, and it will not turn from here. The first minute is the one that mattered, and the rest of the cook is mostly bookkeeping.