The weight does not change, but the volume does. A sieve aerates flour and breaks up clumps; the same flour takes up more space after sifting. That is why “one 200 ml measure of sifted flour” and “one 200 ml measure of flour, sifted” are not the same amount: in the first case, the measure is taken after sifting; in the second, before. Weighing removes this distinction entirely.
Sifting is a process that breaks up clumps in flour and separates the particles, letting air in between them. It adds nothing to the flour’s mass and takes nothing away; it only increases the space taken up by the same amount. The result is that sifted flour fills the same measure with less weight. Weight stays constant, volume increases, and all the confusion comes from this one sentence. The same applies to fine powders such as cocoa, powdered sugar, and starch. The difference is even more pronounced with cocoa, because it clumps much more easily than flour.
That is why where the adjective appears in a recipe directly changes the amount. “one 200 ml measure of sifted flour” means that the flour is sifted first and then put into the measure; the measure fills with aerated flour, so its weight is lower. “one 200 ml measure of flour, sifted” means that the flour is measured first and sifted afterward; the measure fills with packed flour, so its weight is higher. The two phrases look very similar and are often switched in recipe translations. If the adjective’s position is not preserved in translation, the recipe quietly turns into a different one.
The difference compounds an uncertainty that already exists with flour. According to the corpus measurements, the way flour is added to a measure by itself can create a weight difference of 20–30%: if the measure is dipped directly into a bag of flour, its sides compact the flour; if it is filled loosely with a spoon and leveled, the flour stays aerated. Sifting is the third variable in this picture. When these three uncertainties stack up, “a measure of flour” no longer describes one amount but a wide range.
The corpus’s own measurement equivalents are therefore given as a range. For one 200 ml measure of flour, values from 110 to 150 g are cited, and an average of about 120 g is given for sifted flour. This range is an admission of uncertainty, not a flaw; no chart can give a single number without knowing how the flour went into the measure. The lower end of the range describes aerated, sifted flour; the upper end describes packed flour, and both are correct. A chart that tries to narrow the range promises readers a level of precision that does not exist.
Sifting also has a main purpose that has nothing to do with measuring. A sieve breaks up clumps, blends dry ingredients evenly, and distributes the leavener throughout the flour. That is why the sifting step should not be skipped; if it is skipped, at least aerate the flour with a whisk. The answer to a measurement problem is not to stop sifting, but to weigh the amount. When flour is specified in grams, the order of sifting makes no difference and the recipe means the same thing to everyone reading it. Clumps of leavener can leave bitter-tasting spots in the batter.