Most food safety failures announce themselves. A pathogen shows in a swab, a foreign body shows on a detector, an allergen shows in an ELISA. The system is built on the assumption that a problem leaves a trace you can go and find.
A number of kosher failures do not work that way. The status changes, nothing measurable changes with it, and no downstream test will recover the fact. That difference is worth naming, because it explains why supervision in a kosher programme looks different from the rest of your QA.
A worked example
Poultry cannot be scalded in hot water before it has been salted — heat before kashering cooks the blood in place. There is a working allowance for machine plucking, but only where the water stays below 42°C.
Now consider a scalder that drifts to 50°C for twenty minutes on a night shift and comes back down.
- The birds look identical.
- They weigh the same, cook the same and test the same.
- Nothing in the finished product carries the event.
- Every bird that passed through in those twenty minutes is nonetheless forbidden.
No amount of end-product testing recovers this, because there is nothing to detect. The only control that works is someone watching the reading while the birds are in the water — which is why the requirement attached to the 42°C allowance is continuous supervision rather than a recorded set point. The full rule is at kashering poultry.
Why a log is not the same as a witness
A chart recorder tells you what a sensor reported. That is genuinely useful, and we are not suggesting otherwise. But for this class of parameter it answers a slightly different question than the one being asked.
| A record establishes | The kosher question is |
|---|---|
| What the instrument read | What the product actually experienced |
| That the set point was correct | That the set point was held, at the product, throughout |
| That an excursion was flagged | That the affected product was identified before it merged with the rest |
| That the line was cleaned | That it was cleaned in the manner the status change requires |
The last row is the one that catches people. A validated CIP cycle is excellent hygiene and is not kashering. They are different operations with different requirements, and passing one does not discharge the other.
The parameters that behave this way
Not everything needs this treatment. These are the categories where the failure genuinely leaves nothing behind:
- Temperature thresholds that change status. Yad soledes bo in shared equipment; the 42°C scalder ceiling; the −12°C hold on unkashered frozen meat, which must be reached within 72 hours of shechita. See frozen meat.
- Elapsed time. The 24-hour ben yomo gap between productions, and the three-day salting window. A line run nineteen hours after the previous product looks exactly like one run twenty-five hours after.
- Sequence. Kashering before an allergen-style changeover rather than after it; salting before scalding rather than the reverse. The same operations in the wrong order produce a different result and an identical-looking product.
- Who performed the act. Whether a Jew lit the heat matters for bishul akum, and the same rule that governs a home kitchen is stated to apply in plants. Nothing about the food records it.
- Proportion at the moment of addition. Nullification depends on the ratio when the material entered, not on what the tank holds at the end. Topping up afterwards does not retroactively fix it — see bitul.
What this means for how you run the programme
Three practical consequences, none of which requires new equipment.
Identify which of your steps are in this class. Most are not. On a typical line it is a handful — the heat steps, the changeovers, and anything involving a shared vessel. Those few deserve a different level of attention from the rest of the process, and knowing which they are is most of the work.
Schedule supervision to them, not to convenience. A visit that covers a changeover is worth several that cover steady-state running. If your certified production happens on a night shift, that is when the coverage has to be.
Raise excursions immediately, not at the review. The honest reason is that these situations are frequently recoverable — but the options narrow fast. Meat salted while still frozen inside can often be saved if the salt has not yet been rinsed off, and cannot be once it has. An hour's notice is sometimes the difference between kashering a vessel and writing off a batch. See when a supplier changes something for the same principle applied upstream.
These are general principles rather than a ruling on your plant. What applies to your process, and what any particular excursion requires, is a question for your certifying rav and your coordinator.
Where this comes from
- Shulchan Aruch, Yoreh De'ah 68, 69, 105 and 121
- ספר הכשרות כהלכה, פרק י–יא — frozen meat and poultry
- ספר הכשר כלים, פרק ב and פרק ל — kitchen maintenance and non-Jewish cooking