Fish, Once the Skin Is Off

A whole fish carries its own evidence. A skinned fillet carries none — and that is the entire problem with the modern fish trade, because filleting is where the identifying features are removed.

The scale is the operative sign

The Torah gives two signs: fins and scales. But they are not equally useful, and the asymmetry is the most practically important fact here.

Anything with a scale has a fin. Some have fins without scales.

Which yields a clean working rule:

• Found a scale? You need not look for a fin.
• Found a fin? That establishes nothing until you know it has a scale.

So every verification question reduces to the scale. Which makes the definition of a scale the thing worth being exact about.

What counts as a scale is narrower than you think

This is where halacha and biology come apart, and the sources say so explicitly: there are fish that have scales by scientific definition and do not have them by halacha.

The halachic test is that the scale peels off, by hand or with a tool. From which three disqualifications follow:

1. It cannot be peeled from the skin — then it is not a scale that permits the fish.
2. Peeling it does not leave the skin intact beneath — then it was part of the skin, not a scale, and the fish is not kosher.
3. The scales sit under the skin, revealed only when the skin is peeled — not kosher.

The sources are candid that this leaves live questions: what exactly "peels off" means, what degree of damage to the skin is tolerable in peeling, and with how much force. There are also varieties with an outer skin layer over the scale, and whether that still counts as a sign of purity is raised as an open question.

These are not academic. They are the questions a processor hits the moment a species is unfamiliar.

Species names prove nothing

The single most useful warning in the chapter, and it is aimed squarely at people working from paperwork:

Fish names change with time and place, and a non-kosher fish and a kosher fish are not infrequently called by the same name.

So a species name on a spec sheet, a landing document or a supplier declaration is not evidence of anything. Two consignments with identical paperwork can be different animals, and the same animal can arrive under different names from different ports.

This is why certification of fish runs on physical evidence rather than documentation — the documentation is describing a name, and the name is not stable.

Fillets, and the position taken on them

Fish sold cleaned, without skin and bones, is addressed directly — and the position recorded is not to permit fish cleaned of skin in any form without signs of kashrus.

The secondary signs do not rescue it. Chazal identified further indicators — a round, wide head and a spinal cord in kosher fish, against a pointed head and no spine in non-kosher ones; the form of the air bladder, one end wide like a jug and one pointed; the shape of the roe, joined together with one end wide and one pointed.

But these are contested. Some Rishonim hold they cannot be relied on, and some kosher fish have bladders pointed at both ends. The ruling is that where a fish is whole and no scales are visible, head and spine are not relied on — and the determining signs remain fin and scale.

So a fillet is not a documentation problem you can solve with better paperwork. It is a fish whose evidence has been discarded.

The one rule that helps with pieces

There is a genuinely useful provision for cut product, and it turns on reassembly.

Where you have pieces of fish and only one of them bears a scale:

• If the pieces fit together — they reassemble cleanly — all of them are permitted, on the presumption that they came from one fish.
• If they do not fit together, only the pieces bearing scales are permitted. The rest are not.

That is a real control, and it has an obvious operational implication: how you cut and how you keep batches together determines whether a single piece of evidence can carry the rest. A line that commingles trim from multiple fish has destroyed the reassembly argument along with the skin.

What this means for a fish processor

1. Retain evidence on the fish, not on the paperwork. Skin-on identification at the point where identity is still visible is worth more than any downstream document.
2. Keep the chain from one fish together where you can, because reassembly is what makes a scale on one piece useful for the others.
3. Treat unfamiliar species as unresolved until the scale question is actually answered — not the naming question.
4. Do not rely on head and spine for a whole fish with no visible scales.

There is also a caution worth knowing for anyone sourcing farmed product: the chapter raises fish presumed kosher that, through intervention, grow without scales — and the reverse. What a species did historically is not automatically what a given stock does now.

For the underlying signs, see which fish are kosher. What your specific product requires is a question for your certifying rav and your coordinator.