A Cup of Flour Is Not a Unit, and That Is Why Your Baking Is Inconsistent
The same recipe, the same oven, the same person, three weeks apart. One time the cake is tender and one time it is dry and tight, and nothing identifiable was done differently.
Something was. A cup is a measure of space, and flour is a substance that occupies wildly different amounts of space depending on how it got into the cup. The recipe was not inconsistent. The measurement was.
The size of the error is larger than people expect
Scoop a measuring cup directly into the bag and you compress the flour as you go. Spoon it in loosely and level the top and you do not.
The difference between those two is roughly twenty percent by weight, and it can be more if the bag has been sitting undisturbed and the flour has settled. Twenty percent more flour in a cake batter is not a subtle change; it is the difference between tender and dry, and between a batter that pours and one that has to be spread.
It gets worse across recipes, because recipe writers do not agree on what a cup weighs. Some test at 120 grams, some at 125, some closer to 142. A recipe written at the low end and made at the high end is carrying an error before anything is mixed.
Brown sugar is the same problem with the packing made explicit. Cocoa, powdered sugar and anything else that clumps and aerates all behave the same way.
Which is why the errors compound in baking and not in cooking
A stew with slightly more carrot in it is a stew. A cake with twenty percent more flour is a different formula, because everything in a batter is in a relationship with everything else.
More flour means more protein to hydrate, so the structure develops differently. It means less water available per gram of starch, so the crumb sets drier. It means the ratio of leavening to flour has quietly dropped, so the rise is weaker. And it means less sugar per gram of flour, so there is less tenderizing and less moisture retention.
One measurement, four consequences. A cake that failed is frequently a cake whose flour was measured on a different day, in a different mood.
Salt has the same problem and it is worse, because the food tastes of it
A tablespoon of one brand of kosher salt weighs roughly half what a tablespoon of table salt does, because the crystals are a different shape and pack differently.
So a recipe calling for a tablespoon of salt is meaningless without knowing which box the writer had. Four different volumes, one weight is the whole of it, and it is why a recipe that worked for the person who wrote it can be inedible in another kitchen for no visible reason.
Salting by percentage of the food’s weight removes the brand problem, the country problem and the crystal-shape problem in one move.
What volume is actually fine for
Liquids. A cup of water, milk or stock is a cup of water, milk or stock, because liquids do not compress and cannot be packed. Oil likewise.
Anything where a twenty percent error does not matter: a handful of herbs, a splash of vinegar in a dressing you are going to taste anyway, the amount of onion in a soup.
And small quantities of potent things are the opposite case, where volume is worst rather than best, because a quarter teaspoon measured carelessly is a large proportional error in something with a narrow margin. Baking soda and yeast are the two that punish this hardest.
The scale does three things and only one of them is accuracy
It is accurate, which is the obvious one.
It removes washing up. Put the bowl on, press tare, add flour to the number, press tare, add sugar to the number. One bowl, no cups, no spoons, no scraping a sticky measure.
And it makes scaling arithmetic rather than guesswork. Two-thirds of 250 grams is a number. Two-thirds of a cup and a half is a negotiation with a drawer full of measures, and it is the specific thing that makes cutting a recipe down feel harder than it should.
A scale that reads to one gram costs less than a decent saucepan and is the single highest-return object in a baking kitchen.
Percentages are what the weights are for
Once everything is in grams, the relationships become visible. Bakers describe a dough as percentages of the flour weight — seventy percent water, two percent salt, one percent yeast — and that notation is not professional affectation. It is what lets you see at a glance that one bread recipe is wetter than another, and to move a formula between batch sizes without redoing any arithmetic.
The same idea runs outside baking. One percent salt by weight for a pot of water food will be lifted out of, a similar figure for meat, a fixed ratio for a vinaigrette. A sauce is a ratio rather than a recipe, and ratios only exist if the things being related are measured the same way.
The case for putting the scale away
Which is the other half, and it is not a contradiction.
Weigh the things where the relationship is fixed and the margin is narrow: flour, salt, water, leavening, sugar, anything baked, anything brined, anything you intend to repeat exactly.
Do not weigh the things that are meant to be judged. The amount of lemon in a dressing, the quantity of herbs, how much cheese goes on, whether that needs more salt — those are decisions made by tasting, and a scale cannot help with any of them. A cook who weighs the vinaigrette is measuring the wrong thing.
The scale is for the arithmetic you cannot taste your way to. Everything you can taste your way to should be tasted.
If you keep the cups
Spoon the flour into the cup and level it with a straight edge. Never scoop, and never tap or shake it down.
Expect roughly 120 to 125 grams per cup of all-purpose flour, which is what most modern recipes are written to. If a recipe gives weights as well as volumes, use the weights, because the weights are what the writer actually tested with.
And pack brown sugar firmly every time, since that is what every recipe assumes and the assumption is never stated. Consistency is worth more than correctness here: the same wrong measure every time still produces the same cake.
Part of Recipes & Technique.
