Salt in a sweet bake is doing four jobs and only one is taste

What salt does to the perception of sweetness

Sodium ions interact with the taste system in ways that are not confined to the salty channel. At low concentrations sodium suppresses the perception of bitterness quite markedly, and because bitterness and sweetness partly mask one another, removing some bitterness leaves the sweetness reading as stronger. That is the whole trick behind salt in a caramel or a dark chocolate dessert. The salt is not adding sweetness. It is removing something that was competing with it.

There is an aroma effect running alongside. A great deal of what we call flavour arrives through the nose rather than the tongue, and salt influences how readily volatile compounds leave the food and reach the receptors. A sweet bake made without any salt tends to be described as flat or one-dimensional even by people who cannot say what is missing, and the description is accurate — the sweetness is intact but everything arranged around it has receded.

In a yeasted dough it is a structural control

Salt tightens gluten, and it does so by a mechanism worth knowing. Gluten proteins carry charged groups that repel one another, holding the network slightly loose and open. Dissolved salt shields those charges, letting the strands pack more closely and bond more readily, which is why a salted dough feels firmer and less sticky and holds its shape better on the bench. A dough mixed without salt is noticeably slack, tears more easily, and spreads instead of standing up during its final proof.

The second effect is on the yeast itself. Salt is osmotically active, so it draws water away from the yeast cells and slows their metabolism, moderating the rate of gas production. That slowing is usually welcome, because it stretches fermentation out and gives flavour compounds time to accumulate. Push it too far and the dough sulks: a heavily salted enriched dough can take twice as long to prove, which is one reason salt and sugar are usually dispersed rather than tipped onto the yeast directly.

Foams and custards feel it in the other direction

Salt is a mild destabiliser of egg white foams. It interferes with the protein interactions that build the film around each bubble, so a meringue made with salt in the whites takes longer to reach volume and is a little more prone to weeping. This is why older recipes that call for a pinch of salt with the whites have quietly disappeared from careful modern ones, and why cream of tartar took the job over. The two additions were never doing the same thing.

In custards and sauces the picture reverses, because there is no foam to protect and plenty of flat dairy flavour to lift. A small quantity of salt in a crème anglaise or a pastry cream sharpens the vanilla and stops the whole thing reading as sweetened milk. The quantities involved are genuinely small, well below any threshold at which a taster would identify salt as present, and the difference between the salted and unsalted versions is obvious when the two are set side by side.

Which salt, and why the measurement is a trap

Salt is salt chemically, but crystal size changes both how much of it fits in a spoon and how fast it dissolves. Fine table salt packs densely, so a teaspoon of it can carry close to twice the sodium of a teaspoon of a coarse flaked salt. A recipe written for one and executed with the other is not slightly off, it is badly wrong in a way that is easy to blame on something else. Weighing salt solves the problem completely and takes no longer.

Dissolution speed matters where the salt has to disappear. Anything folded into a batter at the last moment, or scattered into a dough near the end of mixing, should be fine enough to dissolve in the water available. Coarse flakes on top of a finished bake are a different device altogether: they are meant not to dissolve, so that the eater meets brief points of salinity against a sweet background rather than an evenly seasoned whole.

Where salt is the point rather than the support

Salted caramel works because caramelisation generates genuinely bitter compounds alongside the sweet and nutty ones, and salt takes the edge off that bitterness while the residual saltiness contrasts with the sugar. Salted butter in a pastry does something similar and quieter. Neither is a modern invention — coastal butter has been salted for as long as it has needed keeping, and the pairing of salt with caramel is old in several regional traditions well before it became a menu fixture.

How much is a matter of taste, and taste here is cultural rather than technical. Palates accustomed to salt in sweet things read a lightly salted dessert as complete, while palates that are not read the same dessert as odd. That is worth saying plainly instead of prescribing a ratio. The defensible technical claim is narrower: a bake with no salt at all is missing a variable, and the effect of restoring it shows up in flavours that have nothing to do with salt.