Stabilisers manage water rather than thicken anything

A word that describes the effect badly

The stabilisers used in frozen desserts — locust bean gum, guar, carrageenan, various starches, gelatin in older recipes — get described as thickeners, and that is the least important thing they do. Their function is to bind free water and raise the viscosity of the unfrozen phase, so that water molecules move much more slowly through the mixture. Almost every benefit follows from that restriction on movement.

These are hydrocolloids, long chain molecules that disperse in water and form networks by holding on to it. They are used in small quantities, well under one per cent by weight in most formulations. That is a genuinely tiny addition, and it explains why an ingredient list can look alarming while representing almost nothing of the actual product.

Recrystallisation is the enemy they exist to fight

Ice cream in a freezer is not static. Temperature cycles as the compressor turns on and off and as the door opens, and each small rise melts a fraction of the ice — the smallest crystals first, since they have the most surface relative to their volume and are therefore least stable. When the temperature falls again, that water does not reform small crystals. It deposits onto the larger crystals that survived, which grow.

Repeat that a few dozen times over a fortnight and the crystal population has shifted decisively towards large, and the ice cream is grainy. This is recrystallisation, and it is the main reason ice cream deteriorates in storage rather than any chemical spoilage. A stabiliser slows it by making it physically harder for water to migrate from one crystal to another through the viscous unfrozen syrup between them.

What else they hold together

The same water-binding does secondary work. Free water at the surface of an ice cream sublimates and refreezes on the lid as frost, which is what leaves a dried, cratered surface on a half-eaten tub; less mobile water sublimates less. Stabilisers also thicken the mix before churning, which helps the machine incorporate and retain air, and they reduce the tendency of a mixture to shrink as air escapes over time.

Some of them do more than one thing. Carrageenan is often included at very low levels specifically to stop milk proteins separating out — a job unrelated to ice crystals, and one it does by interacting with casein. Egg yolk in a traditional custard base is doing an overlapping job through its lecithin and protein, which is why old recipes without gums are not necessarily unstabilised.

The case against overuse is real

Too much stabiliser produces a texture that is often described as gummy or chewy, one that resists the spoon and does not melt cleanly in the mouth but instead turns slippery and lingers. It can also mute flavour release, since a viscous matrix holds on to aroma compounds as well as water. Ice cream that seems to coat the mouth without ever quite going away has usually been over-stabilised.

There is also a legitimate stylistic argument. A gelato made to be eaten within a day of churning does not need protection against three weeks of temperature cycling, and the leanest versions rely on being sold fast rather than on additives. Whether a domestic batch benefits is genuinely contested: those who make ice cream weekly and eat it in two days often find gums unnecessary, and those who keep a tub for a month generally find they help a great deal.

The domestic alternatives do the same job less precisely

Before commercial gums, cooks used what was available. Gelatin was common and works, though it sets in a way that can turn slightly elastic. Cornflour is used in some traditional preparations and gives a distinctive stretchy body. Egg yolk, cooked into a custard, contributes both emulsification and thickening. Even milk powder helps, by adding solids that occupy water without freezing.

Each of these has to be hydrated properly to work, which is the step most often skipped. A gum sprinkled onto cold liquid clumps instantly, forming lumps with dry powder trapped inside that will never disperse; gelatin needs to bloom in cold water before it is warmed; starch has to be cooked out or it tastes raw and chalky. Whatever is used, it belongs in the base before churning and generally before chilling, since a cold mixture is much less willing to take it up.

The point worth taking from all of them is that none is a shortcut for a poorly made base. A stabiliser cannot rescue a mixture that froze too slowly and formed large crystals in the first place, because it acts on growth rather than on the crystals already present. It buys time. What it does not do is make a bad ice cream good, and any account that claims otherwise is selling something.