Stock and sauces

Lesson 7 left you a chicken carcass, and lesson 8 left you a braise and a pot of poaching broth. This lesson turns them into the things that hold a kitchen's cooking together: a stock, and the sauces you build on it.

It also deals with the hazard that belongs to big pots. A big pot of stock, soup or stew cooked perfectly safely can make people ill the next day, and the reason isn't anything that happened while it cooked. It is what happened in the hours after, while it sat cooling. The FDA's handbook on foodborne germs says of the organism behind it: "In most instances, the actual cause of poisoning by this organism is temperature abuse of cooked foods."⁶ By the end you'll know how to get a big pot cold fast enough, why the rule has two stages, and how to thicken a sauce, balance a dressing and rescue both.

What a stock is for

A stock does two jobs, and they come from different things.

The first is body: the slightly thick, lip-coating feel that makes a good broth different from salted water. That comes from gelatin. Lesson 8 showed collagen in connective tissue dissolving into gelatin with long, moist heat. ATK puts the stock version in one clause: "the bones and meat used to make broth contain lots of collagen, which breaks down to form gelatin."² That is why stock recipes ask for the bony, sinewy parts of a bird. ATK's classic chicken broth uses backs and wings, and it says why the long cook matters: "The long, slow simmer helps the bones and meat release both deep flavor and gelatin and results in a full-bodied, all-purpose broth."¹ ATK also notes that its cooks add powdered gelatin to sauces made from bought broth, to give them the silky texture "you'd otherwise get only if you started with a homemade broth".²

The second job is savouriness, the taste lesson 3 called umami, carried by free glutamate. Chicken and beef carry some, and lesson 3's chart showed how little of their glutamate is free. No source read here measures how much a chicken stock gains as it cooks, so this course won't give you a figure. What it can show you is the other extreme: a Japanese stock with plenty of savour and no body at all.

How much water to how much bone

Recipes give proportions, and they do not all agree. Two tested ones, worked out by weight:

The ratios are this course's arithmetic, taking a litre of water as a kilogram.¹⁴ Michael Ruhlman's book Ratio is widely reported to give stock as 3 parts water to 2 of bones, which is 1.5 to 1, but this course hasn't read the page that says so, so treat that figure as reported, not checked. What the two recipes agree on is the shape: more water than bone, but not a great deal more, and enough to cover. A stock made with a small carcass in a huge pot of water will taste thin, however long it cooks.

Keep it below the boil

Both recipes above hold the pot below a boil for the long cook, and there's a tested reason for it.

ATK tested it on a French boiled-beef dish, pot-au-feu, and drew the same conclusion for stock: "boiling will cause soluble proteins and rendered fat to emulsify into the cooking liquid. By simmering, you avoid emulsifying the fat and thus keep the stock clearer, and we found that the scum created simply settled to the bottom of the pot."³ Lesson 5's emulsion is the mechanism. A hard boil beats fat into fine droplets and holds them through the liquid, along with fine bits of protein, and the stock goes cloudy and a little greasy. At a simmer, the fat stays on top where you can lift it off, and the grey scum sinks or can be skimmed.

Some cooks go gentler still. Ruhlman makes his stocks in the oven: "This is the way I make all my stocks, bones and water in the oven overnight", with the oven at 180 to 200°F (80 to 90°C). Of the bones he says: "12 hours in hot but not simmering water should do the trick!"⁴ Lesson 8 put ATK's simmer at 180 to 190°F, so Ruhlman's oven holds the pot at about a simmer or a little under it, for a long time, without anyone watching it.

Notice what the reason is, though. ATK's test was about clarity and scum. Nothing in it says a boiled stock is less safe, and nothing read here says it tastes worse; it looks cloudier and feels greasier. That's the course's question from lesson 1 applied: "don't boil the stock" protects the look and the texture. If you don't mind a cloudy stock for a stew, the rule matters less than it sounds.

Chicken stock from lesson 7's carcass

Here's the whole thing, from carcass to fridge.

1. Break up the carcass and put it in a pot with any wings, necks or trimmings you've saved (raw or roasted; lesson 4's raw chicken routine applies to raw ones). Weigh it if you can. At ATK's proportion a kilogram of bones takes about 1.8 litres of water; at Ruhlman's, about 1.4. In practice, add cold water just to cover.
2. Bring it up slowly and skim off the grey foam as it rises, as ATK does.¹ Then turn it down until the surface barely moves. Check the liquid with your thermometer: you want lesson 8's simmer or a little below, about 82 to 88°C.
3. Leave it. ATK simmers its broth for 3 hours, then adds an onion and bay leaves for 2 hours more.¹ Ruhlman adds his vegetables for the last part too.⁴ Do not salt it much, or at all. You'll reduce it later for sauces, and lesson 6's point about low-salt stock applies: reduction concentrates salt.
4. Strain it through a sieve into a clean pot. Let it settle a few minutes and skim the fat off the top, as ATK does.¹
5. Now cool it, which is the next section, and the one part of this that is a safety step.

Dashi: savour with no bones

A Japanese kitchen's basic stock shows the second job of a stock on its own, because it does that job and almost nothing else.

Dashi starts with konbu, a large dried seaweed. A 2012 paper by the biophysicist Ole Mouritsen and colleagues, working with a Nordic research kitchen to make dashi from Nordic seaweeds, describes the classic version: "Dashi is made by a warm extract of the large brown Japanese seaweed konbu (Saccharina japonica). This extract is called konbu dashi. To the konbu dashi is then added a particular, highly processed fish product, katsuobushi, leading to the so-called first dashi (ichiban dashi)."⁵ Elsewhere the paper calls katsuobushi "fermented". Konbu is also the seaweed behind lesson 3's story of Kikunae Ikeda: the paper says he found it holds "about 2 to 3 g of free MSG per 100 g dry weight".⁵

The method runs the other way from a chicken stock: short and gentle, not long. The recipes, the paper says, "generally prescribe soaking the dry konbu in water at room temperature (typically using 10 g of dry konbu per litre of water) for about half an hour, heating it in an open pan to just below the water boiling point at 100°C and then quickly removing the konbu from the water before bitter-tasting compounds seep out."⁵ Boil the konbu and the dashi turns bitter, and a more recent version of the recipe, the paper says, heats it only to 60°C.⁵

The katsuobushi adds a second savoury compound, inosinate, and the two work together. Mouritsen and colleagues report that inosinate and guanylate "in a synergistic fashion potentiate the receptor's sensitivity to glutamate", and that the Japanese chemist Akira Kuninaka showed in 1957 that inosinate from katsuobushi, or guanylate from dried shiitake, enter "a synergistic relationship with glutamate from konbu."⁵ So a little of each tastes more savoury than either would alone. The same paper points to Western pairings that work the same way: "tomatoes with anchovies, vegetables with meat, eggs with bacon, green peas with scallops".⁵

That also gives you a meat-free stock with real depth. In the vegetarian temple cooking of Japan, the paper says, katsuobushi "is replaced by dried shiitake. Konbu provides glutamate and shiitake provides guanylate to replace inosinate from katsuobushi."⁵ So konbu soaked and gently heated, with a few dried shiitake soaked alongside it, makes a vegetable stock with a reason behind each ingredient (the quantities are this course's suggestion, not the paper's).

What dashi does not have is body. There's no skin, bone or connective tissue in it to give gelatin, and the paper describes konbu dashi as "light in colour with a fluidity similar to that of pure water."⁵ So a stock can be deeply savoury and thin as water, or rich in body and fairly mild. A chicken stock does both jobs; dashi does one and does it very well.

Cooling a big pot, and the germ it is aimed at

Lesson 4 gave you the rules for cooling: divide food into smaller portions, the FDA's new benchmark of 2 inches (about 5 cm) deep, and the two countries' different advice about hot food and the fridge door. It promised that this lesson would say which germ those rules are aimed at. It's Clostridium perfringens.

What the germ does

The FDA's Bad Bug Book, its handbook on foodborne germs, gives the story in three steps.⁶

First, cooking doesn't get rid of it. The germ makes spores, a dormant, tough form. "Spores of some C. perfringens strains can survive boiling water for an hour or longer in a relatively protective medium (e.g., a cooked-meat medium)."⁶ A meat stew is much like that kind of medium, on this course's reading. The FDA's food code notes the consequence: after proper cooking, "all but the spore-forming organisms such as Clostridium perfringens or Bacillus cereus should be killed or inactivated."⁷ A joint FDA and USDA paper printed in the same code adds that these spore-forming organisms "can thrive in the warm product since other competing organisms have been eliminated."⁷

Second, as the food cools, the spores come back to life and grow, fast. The Bad Bug Book gives a "fast doubling time (<10 minutes for vegetative cells), depending on temperature and food matrix", and says that "during cool-down (109-113°F) and storage of prepared foods, this organism can reach levels that cause food poisoning much more quickly than can other bacteria."⁶ 109 to 113°F is about 43 to 45°C, a little hotter than a hot bath, a temperature every pot of stew passes through on its way down.

Third, it takes a lot of them. Illness follows eating more than a million cells or spores per gram of food, typically about 16 hours later, with cramps and diarrhoea that in most people last 12 to 24 hours.⁶ That is why it is a germ of big batches. The FDA names the usual setting as institutions "where large quantities of food are prepared several hours before serving", and the foods as "Meats (especially beef and poultry), meat-containing products (e.g., gravies and stews)".⁶ The FSA brings it home: it "also applies to batch cooking at home where you may leave food out to cool longer than you intended while preparing other food."¹²

To see why speed matters, here's some arithmetic, which is this course's and not a measurement. A doubling every 10 minutes is six doublings an hour, a 64-fold rise. Three hours at that rate is eighteen doublings, more than a 250,000-fold rise. A pot does not sit at the ideal temperature the whole time, so a real pot grows less than that. But the arithmetic shows why an hour or two makes such a difference: growth that fast turns a harmless count into a harmful one in an afternoon.

Why the rule has two stages

Now the rule makes sense. The FDA Food Code, the model code US states adopt for restaurants and shops, says cooked food must be cooled from 57°C (135°F) to 21°C (70°F) within two hours, and to 5°C (41°F) or below within six hours in total.⁷ Why two limits instead of one?

Because the top stretch is the dangerous one. The code's own explanation names the range to get through quickly, "ideal bacterial incubation temperatures", as 21 to 52°C (70 to 125°F), and it says: "The initial 2-hour cool is a critical element of this cooling process."⁷ It gives two examples, and they're the clearest way to see the point. Food that falls from 135°F to 70°F in 1 hour, the code's example of "proper cooling", has 5 hours left for the rest. Food that makes 41°F within the six hours but took 3 hours to reach 70°F "may not be adequately controlled", in the code's words.⁷ The total isn't what matters. The time spent in the warm band is.

The chart shows the two limits, the warm band, and those two examples.

The limits and the two examples are from the FDA Food Code 2026, § 3-501.14 and its Annex 3 notes; the 70 to 125°F band is the code's "ideal bacterial incubation temperatures".⁷ The 109 to 113°F strip is the Bad Bug Book's cool-down range for C. perfringens.⁶ The FDA gives only the start and the 70°F point for each example; for the first it says 5 hours remain to reach 41°F, and for the second that it reaches 41°F at 6 hours, so both are drawn arriving at 41°F at the 6-hour limit. The straight lines between the points are drawn for reading and are not measured cooling curves.

The dashed line is the one to look at. By the two-hour mark it is still in the warm band, and it spends about two and a half hours in it, crossing the C. perfringens strip slowly on the way. The solid line is through the band in about an hour.

What makes a pot cool fast

The Food Code is written for restaurants, not homes, but its reasoning is physics, and a home pot obeys the same physics. Three things decide how fast a pot of food cools.

Depth, most of all. The code's notes: "By reducing the volume of the food in an individual container, the rate of cooling is dramatically increased and opportunity for pathogen growth is minimized."⁷ For its new 2-inch benchmark, the 2026 code cites a study by authors from Minneapolis's health department, Stockton University and Rutgers, using the cooling of 224 recipes logged by Minneapolis inspectors in restaurant kitchens between 2018 and 2019. "Food depth had a highly significant effect", they found, and their conclusion was that "cooling foods in shallow containers at a food depth of less than or equal to ∼5.1 cm (2 in.) poses little risk of significant pathogen growth."⁸ That growth was estimated with existing models of C. perfringens and B. cereus, not measured in the food, and the kitchens were restaurant kitchens, but depth is depth.

Ice. "The use of an ice bath or ice wand was also significant," the same study found.⁸ (An ice wand is a sealed plastic tube filled with water and frozen, used for stirring a pot from the inside.) At home, that's a sink or roasting tin of iced water around the container. The Food Code lists "Stirring the *FOOD in a container placed in an ice water bath" among its methods (the asterisk is the code's own mark for a defined term).⁷

No tight lid. In the study, the shallow foods that still cooled too slowly were being cooled above 41°F "or were partially or wholly covered."⁸ The code explains it: "If the hot food container is tightly covered, the rate of heat transfer is reduced".⁷ It asks for containers loosely covered, or uncovered where nothing can fall in, while they cool.⁷

One more line from the code is worth knowing: "Commercial refrigeration equipment is designed to hold cold food temperatures, not cool large masses of food."⁷ It's about restaurant fridges. This course's reading is that a home fridge is the same kind of machine, built to keep cold things cold, which is why a deep hot pot pushed to the back of it cools slowly in there too.

Back to the carcass stock

So for the 3 litres of chicken stock: strain it into two or three shallow containers, no more than about 5 cm deep. Stand them in a sink of iced water, stir now and then, and check the middle of one with your thermometer. The Food Code's benchmark is below 21°C (70°F) within two hours; for a home cook that is a useful target to aim for, not a legal rule. Then, in the US, into the fridge; in the UK, into the fridge within the FSA's two hours. Put them in loosely covered, and put the lids on tight once they're cold.

Once it is cold, the fat that was left sets on top. Lift it off; Ruhlman keeps his for making a roux.⁴ Underneath, if the stock had plenty of collagen in it, you may find it has set to a soft jelly. That is the gelatin. ATK notes that gelatin is used to thicken refrigerator pies and is the basis of Jell-O;²⁰ a set stock is, on this course's reading, the savoury version of the same thing.

Six litres of stew for the week

Now the harder case. You cook a big stew on Sunday afternoon: about 6 litres, in one pot, meant for meals across the week. It's hot and heavy, and you want to get it into the fridge and sit down.

Start with the numbers, which are this course's arithmetic. A typical large pot about 24 cm across has a floor of about 450 cm², so 6 litres (6,000 cm³) stands about 13 cm deep in it. At the Food Code's 2 inches, about 5 cm, the same stew needs 1,200 cm² of container floor. A 23 × 33 cm (9 × 13 inch) roasting tin has about 760 cm², so it holds about 3.8 litres at 5 cm: two tins, or a tin and a few tubs. That's the change: from 13 cm deep to 5.

Here's what each country's guidance says about the next step, and what each is protecting.

• The US. FSIS says: "Hot food can be placed directly in the refrigerator or it can be rapidly chilled in an ice or cold water bath before refrigerating." And, in the same page: "A large pot of food like soup or stew should be divided into small portions and put in shallow containers before being refrigerated."⁹ So the shallow tins can go into the fridge hot.
• The UK. The FSA says: "Don’t place hot or warm food in your fridge. Instead, cool cooked food at room temperature and place in the fridge within one to two hours."¹¹ Its C. perfringens page says to cool leftovers "within 2 hours and place them in the fridge or freezer - you can cool food more quickly by separating it into smaller portions".¹² So the tins cool on the counter, or in iced water, and go in within two hours. The FSA pages read here don't mention an ice bath, but it helps meet that deadline (ice was one of the two things the Minneapolis study found made a difference), and the UK's professional cookery qualification lists "ice water bath" among its chilling methods, with a professional standard to "chill to below 5°C within 90 minutes".¹⁴ That is a standard for trainee chefs, not the FSA's advice to homes.

Both countries agree on the part that matters most: do not leave the stew in one deep pot. As lesson 4 said, neither agency's page explains its view on the fridge door, and this course won't guess. What both are guarding against is the same: hours in the warm band, in a mass whose middle cools last.

Then the week. In the US, FSIS gives leftovers 3 to 4 days in the fridge; the FSA says eat them within 48 hours or freeze them.¹⁰¹³ So a stew made on Sunday reaches Wednesday or Thursday in the US and Tuesday in the UK. Freeze the portions you won't reach in time, on the day you make it; lesson 12 plans a week around that. When you reheat, FSIS says to bring "sauces, soups and gravies" to "a rolling boil"; the FSA says steaming hot, and "only ever reheat once", which is another reason to freeze in single-meal portions.¹⁰¹³

Thickening with starch

Most sauces that aren't reductions or emulsions are thickened with starch, and the mechanism is the one lesson 9 used for rice. Starch sits in tiny granules. Heated in liquid, they take up water and swell, and swollen granules crowd the liquid and stop it flowing freely. ATK explains why potato starch thickens more than cornstarch in exactly those terms: "When potato starch granules swell in hot liquids, they take up more space, restricting the liquid’s flow more than smaller cornstarch granules, which results in a more viscous sauce."²¹ Carry on heating and the granules break down into "a mixture of polymers-in-solution", as one review puts it, and the same review adds that "None of the published starch gelatinization theories can fully and adequately explain" every step of the sequence.²² So the cook's version, granules swelling in hot liquid, is as far as anyone needs to go, and further than science has fully mapped.

Two things follow for the cook. Starch thickens as it heats, so a flour sauce doesn't thicken until it is near the boil; ATK says to whisk flour in "well before the mixture boils, when the flour’s starches cause the mixture to thicken."¹⁹ And starch added dry to hot liquid clumps; ATK says so of cornstarch, and Ruhlman of flour.¹⁵²⁰ So every starch method is a way of getting the granules spread out before the heat reaches them.

A roux: flour spread through fat

The first way is to coat the flour in fat. Ruhlman, in a post on gravy: "The key is dispersing the flour uniformly through the sauce. We do this by combining the fat (butter or rendered chicken or turkey fat) so that the granules of flour are each coated with fat to prevent their clumping together."¹⁵ That paste is a roux. Cooking it also takes away the raw-flour taste, as both Ruhlman and ATK say; ATK gives it "a minute or two".¹⁵²⁰

How much flour to fat? Ruhlman's stated ideal is "3 parts flour to 2 parts fat is ideal if measuring by weight", adding that for small amounts "by eye is fine".¹⁵ ATK gives no ratio but the same method, and the same warning about colour: cook a roux until it browns and it "adds flavor to dishes (such as gumbo) but loses some thickening power."²⁰ Ruhlman agrees, in almost the same words.¹⁵ So a dark roux needs more of it.

How much roux to liquid is the question that decides the sauce, and here the best this course can give you is what Ruhlman's own recipes use. The per-cup figures are this course's division of them.

From Ruhlman's gravy, croque madame and moussaka recipes.¹⁵¹⁶¹⁷ A US tablespoon is about 15 ml and a cup about 240 ml. These are one cook's recipe quantities, not tested ratios, and flour measured by the spoon varies.

Read down the last column and a working rule appears: about a tablespoon of flour per cup of liquid for something thin, one and a half for a sauce, three for something you want to hold its shape. None of Ruhlman's three roux recipes keeps his 3:2, by this course's arithmetic, taking a tablespoon of flour as about 8 g. The béchamel uses 2 ounces of butter, about 57 g, with 3 tablespoons of flour, about 24 g; the moussaka a stick of butter, about 113 g, with ¾ cup of flour, about 95 g; and the gravy a quarter-cup of fat with at most 6 tablespoons of flour. Each has more fat than flour. That is the difference between a stated ideal and a real recipe, and it is why the last column matters more than the roux ratio.

The method is the same for every one of them. Melt the fat, stir in the flour and cook it a minute or two without letting it colour (unless you want colour). Then add the liquid gradually, whisking, as ATK says, "for smooth incorporation and maximum thickening power", and bring it to a simmer, where it thickens.²⁰ Ruhlman cooks his béchamel on medium-low for 15 to 45 minutes after that.¹⁶ If it's too thick, add liquid; too thin, simmer it down or add a little more roux.

Milk thickened with a butter and flour roux is béchamel. Ruhlman uses it for a croque madame and, thickened further with cheese and egg yolks, on top of a Greek moussaka.¹⁶¹⁷ Swap the milk for stock and you have what cooking schools, Ruhlman says, call velouté, which at a turkey dinner goes by the name gravy: "Gravy is nothing more than flour-thickened stock."¹⁵

A slurry: starch spread through cold liquid

The second way is to disperse the starch in a little cold liquid first. ATK: "first mix it with cold liquid to form a slurry before adding the thickener to hot liquids to prevent clumping."¹⁹ Cornstarch (called cornflour in the UK) is the usual one. ATK calls it the "go-to thickener for stir-fries", and it is stronger than flour: "Because cornstarch is a pure starch, it is a more effective thickener than flour (which is only 75 percent starch)."¹⁹

Ruhlman's sauce for egg foo young, a Chinese-American egg dish, shows the method and a proportion. He notes that the sauce "was originally simply chicken stock seasoned with soy sauce and thickened with cornstarch".¹⁸ His version whisks 2 tablespoons of cornstarch into 2 tablespoons of stock, then stirs that slurry into a little under 2 cups of simmering stock with soy sauce, sugar, wine, garlic and ginger, and cooks it "for a minute or two, until the sauce thickens."¹⁸ That's about a tablespoon of cornstarch per cup, by this course's arithmetic, and since cornstarch is the stronger thickener, the same spoonful goes further than flour's would.

Cornstarch has one weakness, and it is the reason to know both methods. ATK: "cornstarch-thickened sauces break down more quickly than flour-thickened ones", and "Cornstarch loses its thickening ability with prolonged exposure to heat."¹⁹²⁰ So a slurry suits a sauce made at the end, like a stir-fry's, and a roux suits a gravy or a stew that will sit on the heat. ATK also found that cornstarch needs to simmer "for a couple of minutes to achieve maximum thickness", so give a slurry that long before you judge it.²¹

The pan sauce, revisited

Lesson 6 made a pan sauce and finished it with cold butter, and left one question open: what happens if you boil it after the butter goes in. It had you boil a spoonful and find out. Here's the answer from a source, which this course didn't have then. ATK: "it’s important to add butter off the heat. Because butter is an emulsion that can be broken by high temperatures, at around 160 degrees your nicely thickened sauce will lose its body."¹⁹ The page gives no unit. ATK works in Fahrenheit, and a sauce that is mostly water can't reach 160°C, so this course reads it as 160°F, about 71°C, well below a simmer. So a butter-finished sauce is made hot and served, not held on the heat. If yours broke in lesson 6's practice, that is why.

This lesson adds two other ways to give a pan sauce body. One is the stock itself. A pan sauce made with homemade stock has gelatin in it, which gives it body as it reduces; ATK adds powdered gelatin to sauces made with bought broth for exactly that reason.² The other is a small amount of starch. ATK describes a beurre manié, "a paste of flour and softened butter", whisked into a sauce at the end, and says to whisk it in "well before the mixture boils, when the flour’s starches cause the mixture to thicken."¹⁹ It's a roux that hasn't been cooked first, so use very little of it.

Vinaigrette: an emulsion that does not last

A vinaigrette is oil and an acid, and it is lesson 5's emulsion in a weaker form. The starting ratio is Ruhlman's: "Vinaigrette is 3:1, or 3 parts oil to 1 part vinegar".²³ (His book's ratios are by weight; ATK's recipes use 3 to 1 by volume, which comes out a little less oily than his, since oil is lighter than vinegar.) ATK's foolproof version uses 1 tablespoon of wine vinegar to 3 tablespoons of olive oil, and its make-ahead version a quarter-cup of vinegar to three-quarters of a cup of oil.²⁴²⁵

Unlike a mayonnaise, a plain vinaigrette doesn't hold. ATK is blunt: "no matter how carefully the oil is whisked into the other ingredients, it never stays emulsified for long."²⁵ There's no yolk wrapping each droplet. So ATK adds a little mustard and mayonnaise, "two common emulsifiers in salad dressings that can hold a vinaigrette together for a little while."²⁵ The foolproof recipe uses half a teaspoon of each, whisks them into the vinegar first, then drizzles in the oil slowly, and if oil pools on the surface, stops pouring and whisks until it is absorbed.²⁴ That's lesson 5's rule, oil in slowly so it can be broken up and coated, on a smaller scale. For a dressing that holds for a week in the fridge, ATK adds a tablespoon of molasses, which it says contains compounds "that increase the viscosity of the emulsion and make it difficult for the oil droplets to coalesce and separate from the water."²⁵

Then taste it, because 3:1 is where you start, not where you stop. Lesson 3's tools apply. If it's sharp enough to make you wince, add oil; if it is flat and oily, add acid a few drops at a time; add salt until the flavours come up. Taste it on a leaf of whatever you're dressing, because bitter leaves want a different balance from sweet tomatoes. Write down the ratio you finish at. It may be well away from 3 to 1, depending on how sharp your vinegar is and what you're dressing, and that is the point of starting from a ratio and tasting.

Vinaigrette isn't only a French thing. Swap the acid and the seasoning and the method travels: lime juice with fish sauce and a little sugar is the Thai dressing lesson 3 mentioned. Try rice vinegar with a little soy sauce, or lemon with garlic, and taste your way to the balance the same way.

What people get wrong

"Boil the stock hard for more flavour." ATK's reason for simmering is clarity: a boil emulsifies fat and proteins and clouds the stock.³ Nothing read here says a boil makes it tastier or safer.

"A cloudy stock has gone off." Cloudiness comes from how it was cooked.³ Whether a stock is safe depends on how fast it cooled and how long it's been kept, and the germs that make you ill do not change how food looks or smells (lesson 4).

"Leave the pot on the stove to cool overnight." That is the slow cooling C. perfringens needs, and it's far past both countries' limits.⁶¹⁰¹² Divide it into shallow containers.

"Hot food goes straight in the fridge, so a whole pot can too." FSIS allows hot food in, and in the same breath says a big pot of soup or stew should be divided into shallow containers first.⁹ The depth is the problem, not the heat.

"Cornstarch and flour are interchangeable, spoon for spoon." Cornstarch is stronger and thins with long heating; flour is weaker and holds up.¹⁹²⁰

"Oil and vinegar will stay mixed if you whisk hard enough." A plain vinaigrette separates in time however it is whisked. Mustard or mayonnaise slows that down.²⁵

Practice

Connections

Lesson 2 gave you the centre that changes temperature last, and this lesson ran it backwards through a pot of stew; it's the whole reason the depth rule works. Lesson 3's umami came back as dashi, and its tasting as the way to finish a vinaigrette. Lesson 4's cooling rules got their germ and their reasons here. Lesson 5's emulsion explained both the cloudy stock and the dressing that separates, and lesson 6's pan sauce got the answer to the question it left open. Lessons 7 and 8 supplied the carcass, the collagen and the simmer, and lesson 9's swelling starch granules are what thicken a roux.

Lesson 11 turns to bread, where a ratio of flour to water is the whole recipe, and lesson 12 builds a week of meals around batch cooking, including the stock and the stew from this lesson, with the freezer doing the work the fridge can't.

Go deeper

• FDA, Food Code 2026, Annex 3 notes on § 3-501.14 and § 3-501.15. The code itself is for restaurants, but its public-health notes on cooling are a few pages of plain explanation of why the two stages exist and why depth and lids matter. This course read those sections, not the whole code.
• N. Koreen, W. C. Baldwin and D. W. Schaffner, "Cooling Uncovered Foods at a Depth of ∼5.1 cm (2 in.) or Less Poses Little Risk of Pathogen Growth", Journal of Food Protection, 2024. The study behind the 2-inch rule, from real restaurant kitchens. This course has read the abstract only.
• Ole G. Mouritsen and colleagues, "Seaweeds for umami flavour in the New Nordic Cuisine", Flavour, 2012. Read in full for this course, and free to read. A readable account of dashi, umami synergy and konbu, by scientists who then tried to make dashi from Nordic seaweeds.
• Michael Ruhlman, "Holiday Classic: Gravy & Cranberry Sauce", 2015. Read in full for this course. A short, opinionated case for flour-thickened sauces, with the roux explained in a paragraph. His book Ratio is the natural next step for cooking from proportions; this course has read only its catalogue contents and the publisher's description.

Sources

1. America's Test Kitchen, "Classic Chicken Broth", recipe, https://www.americastestkitchen.com/recipes/9481-classic-chicken-broth . Read level: full, directly. Supports: 4 lb of backs and wings to 3½ quarts of water, skimming, 3 hours then 2 more with the onion, skimming the fat, and the quoted sentence on gelatin and body. The water-to-bones ratio is this course's arithmetic, as the body says.
2. America's Test Kitchen, "Body Builder for Pan Sauces", Cook's Illustrated how-to, https://www.americastestkitchen.com/cooksillustrated/how_tos/9823-body-builder-for-pan-sauces . Read level: full, directly. Supports: collagen in broth breaking down to gelatin, and adding gelatin to sauces made from bought broth.
3. America's Test Kitchen, "For a Clearer Stock, Simmer; Don't Boil", no author or date, https://www.americastestkitchen.com/how_tos/9260-for-a-clearer-stock-simmer-dont-boil . Read level: full, directly, in the recheck recorded for lesson 8. Supports: the quoted sentence on boiling, emulsifying and clarity, and that the page's test was on pot-au-feu. That the page says nothing about safety or taste is this course's reading of it.
4. Michael Ruhlman, "Easy Turkey Stock", ruhlman.com, 18 November 2020, and "Turkey Stock", 22 November 2013 . Read level: full, directly. Supports: 2 to 3 kg of turkey pieces covered with 3 to 4 litres of water, the oven at 180 to 200°F (80 to 90°C), "hot but not simmering", vegetables added later, and keeping the fat for a roux. The book's 3:2 stock ratio was not found on any page read, as the body says.
5. Ole G. Mouritsen, Lars Williams, Rasmus Bjerregaard and Lars Duelund, "Seaweeds for umami flavour in the New Nordic Cuisine", Flavour 1, article 4, 2012, doi 10.1186/2044-7248-1-4, CC BY 2.0, read through the Internet Archive's August 2025 capture . Read level: full, directly. Supports: every statement about dashi, konbu, katsuobushi, the classic method, bitterness, 60°C, synergy and Kuninaka, the temple kitchen's shiitake, the Western pairings, and konbu dashi's thinness. The description of katsuobushi as dried, smoked, fermented skipjack is not from this paper; the paper calls it "a particular, highly processed fish product" and elsewhere "fermented", and the rest is this course's gloss.
6. US FDA, Bad Bug Book, 2nd edition, 2012, chapter "Clostridium perfringens", https://www.fda.gov/media/83271/download . Read level: full chapter, directly. Supports: temperature abuse as the usual cause, spores surviving boiling with the source's qualifier, the doubling time with its qualifier , 109 to 113°F, more than a million cells or spores per gram, onset about 16 hours, 12 to 24 hours, institutions, and the foods. The doubling arithmetic is this course's, as the body says.
7. US FDA, Food Code 2026, released 17 September 2026, § 3-501.14, § 3-501.15 and their Annex 3 notes, https://www.fda.gov/media/194741/download . Read level: the PDF, those sections read in full. Supports: the two stages, "ideal bacterial incubation temperatures" of 21 to 52°C (70 to 125°F), the critical first stage and its two examples, spore-formers surviving cooking, reducing volume, the ice water bath method, loose covers and tight lids, and commercial refrigeration. That a home fridge behaves the same way is this course's reading, as the body says.
8. N. Koreen, W. C. Baldwin and D. W. Schaffner, "Cooling Uncovered Foods at a Depth of ∼5.1 cm (2 in.) or Less Poses Little Risk of Pathogen Growth", Journal of Food Protection 87(10), 2024, 100356, PMID 39241915 . Read level: abstract, directly. Supports: the authors' affiliations, 224 recipes logged by Minneapolis inspectors in 2018 and 2019, depth and ice as significant, covered or warm-held foods cooling slowly, modelled rather than measured growth, and the quoted conclusion.
9. USDA FSIS, "Refrigeration and Food Safety", last updated 23 March 2015, read through the Internet Archive's 23 September 2026 capture . Read level: full. Supports: hot food directly into the fridge or an ice bath first, and dividing a large pot of soup or stew into shallow containers.
10. USDA FSIS, "Leftovers and Food Safety", last updated 31 July 2020, read through the Internet Archive . Read level: full. Supports: the 2-hour rule, 3 to 4 days, and reheating sauces, soups and gravies to a rolling boil.
11. FSA, "How to chill, freeze and defrost food safely", GOV.UK, published 18 December 2017 . Read level: full. Supports: not putting hot food in the fridge, and refrigerating within one to two hours.
12. FSA, "Clostridium perfringens", GOV.UK, published 25 June 2026, https://www.gov.uk/government/publications/clostridium-perfringens/clostridium-perfringens . Read level: full, directly. Supports: batch cooking at home, cooling within 2 hours in smaller portions, and 4 hours as the most food should be left out.
13. FSA, "Cooking your food", GOV.UK, published 18 December 2017 . Read level: full. Supports: leftovers within 48 hours or freeze them, and "only ever reheat once".
14. City & Guilds, Level 2 Diploma in Professional Cookery (7100-12), qualification handbook v1.6, February 2023, unit 707 . Read level: the handbook, directly. Supports: "ice water bath" among the chilling methods and "chill to below 5°C within 90 minutes", as a professional training standard.
15. Michael Ruhlman, "Holiday Classic: Gravy & Cranberry Sauce", ruhlman.com, 25 November 2015, https://ruhlman.com/gravy-cranberry/ . Read level: full, directly. Supports: the roux ideal of 3 parts flour to 2 of fat, coating the granules, cooking out the raw taste, browning reducing thickening power, the gravy quantities, and "Gravy is nothing more than flour-thickened stock."
16. Michael Ruhlman, "Le Creuset Technique: Béchamel Sauce & the Croque Madame", ruhlman.com, 27 August 2013 . Read level: full, directly. Supports: 2 oz butter, 3 tablespoons flour, 2 cups milk, and 15 to 45 minutes on a low heat.
17. Michael Ruhlman, "Moussaka", ruhlman.com, 6 May 2022 . Read level: full, directly. Supports: ¾ cup flour and a stick of butter to a quart of milk, with cheese and yolks.
18. Michael Ruhlman, "Egg Foo Young", ruhlman.com, 19 March 2021 . Read level: full, directly. Supports: the sauce's origin as stock, soy and cornstarch, the slurry, the quantities and the minute or two of cooking. The tablespoon-per-cup figure is this course's arithmetic.
19. America's Test Kitchen, "Getting to Know: Thickeners", no author or date, https://www.americastestkitchen.com/how_tos/9126-getting-to-know-thickeners . Read level: full, directly. Supports: cornstarch as pure starch and flour as 75 percent, cornstarch breaking down faster, the cold slurry, beurre manié, and butter losing its body at around 160 degrees. The page gives no unit; reading it as Fahrenheit is this course's, as the body says .
20. America's Test Kitchen, "Thickeners", no author or date, https://www.americastestkitchen.com/how_tos/6187-thickeners . Read level: full, directly. Supports: cooking the roux, browning losing thickening power, adding liquid gradually, cornstarch losing thickening with prolonged heat, and gelatin as the basis of Jell-O. That a stock sets for the same reason is this course's reading, as the body says.
21. America's Test Kitchen, "Can Potato Starch be Substituted for Cornstarch to Thicken Sauces?", no author or date . Read level: full, directly. Supports: granules swelling and restricting flow, and cornstarch needing a couple of minutes' simmer. That cornstarch sets clear is not from this page; see the note below.
22. W. S. Ratnayake and D. S. Jackson, "Starch gelatinization", Advances in Food and Nutrition Research 55, 2009, 221-268, PMID 18772106 . Read level: abstract, directly. Supports: granules breaking down into polymers in solution, and no theory fully explaining the sequence.
23. Michael Ruhlman, Ratio: The Simple Codes Behind the Craft of Everyday Cooking (Scribner, 2009), the publisher's description on Goodreads . Read level: catalogue contents and publisher's description only. Supports: the 3:1 vinaigrette.
24. America's Test Kitchen, "Foolproof Vinaigrette", recipe, https://www.americastestkitchen.com/recipes/4887-foolproof-vinaigrette . Read level: full, directly. Supports: 1 tablespoon vinegar to 3 of oil, half a teaspoon each of mayonnaise and mustard, mayonnaise as an emulsifier, and the slow drizzle.
25. America's Test Kitchen, "Keep Vinaigrette Emulsified With. . . Molasses?", and the "Make-Ahead Vinaigrette" recipe . Read level: full, directly. Supports: a vinaigrette never staying emulsified for long, mustard and mayonnaise as emulsifiers, molasses and viscosity, and the quarter-cup to three-quarters proportion.