Sorbet is an exercise in freezing point depression

Sugar in a sorbet is doing physics before it does flavour

A sorbet has no fat, no egg and usually no dairy, which strips the problem down to two variables: water and dissolved solids. Its texture is decided almost entirely by what proportion of the water is frozen at serving temperature, and that proportion is set by the concentration of sugar. Sweetness is a side effect that happens to be pleasant.

The underlying principle is freezing point depression. Dissolved molecules interfere with water molecules organising themselves into an ice lattice, so a solution has to be cooled below zero before it will freeze, and the more concentrated the solution the lower that point falls. Crucially, freezing is not an all-or-nothing event in a sugar solution. As ice forms, the remaining liquid becomes more concentrated, which depresses its freezing point further, so the mixture freezes gradually across a range of temperatures rather than at one.

Which is why the window is narrow at both ends

Too little sugar and most of the water freezes at freezer temperature. There is no residual syrup separating the crystals, they fuse, and what you have is a block of flavoured ice that has to be chipped rather than scooped. Too much sugar and so much water stays liquid that the sorbet never sets — it stays slushy, and it also tastes cloying, since there is nothing to balance the sweetness.

The practical range is therefore tighter than most people assume, and it moves with the fruit. A fruit purée already contains its own sugars and its own water, so a sorbet made from a ripe mango and one made from a sharp blackcurrant need different additions to arrive at the same solids concentration. This is what a refractometer is for in a professional kitchen: it measures dissolved solids directly, and it takes the guesswork out of a variable that is genuinely hard to judge by taste.

The old tests were measuring the same thing

Before instruments, sorbet makers judged the syrup by floating something in it — a whole egg was the classic — and looking at how much of it broke the surface. That is not folklore. It is a density measurement, and density tracks dissolved solids closely enough to be useful. The same principle underlies the hydrometer, which is the instrument the egg was standing in for.

Modern kitchens tend to work by weight instead, keeping the ratio of purée to syrup within a known band and adjusting by taste and by how the finished sorbet behaves. Both approaches accept the same truth: the sugar level cannot simply be reduced because a sorbet tastes too sweet. Cutting it changes the texture as surely as changing the churn does, and the way to reduce perceived sweetness is usually acid rather than less sugar.

Without fat, everything else has to work harder

Fat in ice cream hides a multitude of textural sins, and a sorbet has none of it. The crystals are exposed, so they must be smaller, and the mixture has no fat network to give it body. This is why sorbets so often include a small quantity of glucose syrup or invert sugar: those sugars depress the freezing point more effectively per unit of sweetness than sucrose does, because the effect depends on the number of dissolved molecules rather than their sweetness.

Some recipes add a modest amount of egg white or a hydrocolloid for the same structural reason. Both are contested. Egg white contributes protein that stabilises air and gives a slightly smoother, paler result, and purists object that it dulls the fruit, which it can. A small quantity of pectin or another stabiliser thickens the unfrozen phase and slows crystal growth in storage. Neither is required, and a very fresh sorbet eaten within an hour of churning needs neither.

Fruit brings its own complications

Fruit is not just flavour here; it is water, sugar, acid and pectin arriving together in proportions that vary with variety, ripeness and season. A very fibrous purée gives body that a clear juice cannot. An acidic fruit tastes less sweet at the same sugar concentration, which tempts you into adding more sugar and softening the texture past where you wanted it. Ripeness alone can shift the sugar content of the same fruit substantially between one week and the next.

That variability is the honest reason sorbet recipes are less reliable than ice cream recipes. The base of an ice cream is largely dairy and sugar, both consistent; the base of a sorbet is mostly fruit, which is not. The workable approach is to treat any written ratio as a starting point, taste the mixture cold rather than warm — cold blunts sweetness noticeably — and expect to adjust rather than to follow.