In regards to foods like soda, or the KFC recipe or whatever…
I’ll be surprised to learn that the FDA knows all the ingredients, but doesn’t have all the info about the process. Is it possible for a food made with innocuous ingredients to become toxic in some manner due to a process being performed on it?
The idea of “secret” food recipes is really overrated. You think Pepsi doesn’t know exactly how to make Coke?
The idea of “secret recipes” is really more of a marketing tactic than anything else. Let the customer’s think they’re buying something that competitors are gnashing their teeth wishing they had the recipe for. It’s all a shuck and jive. There aren’t any secrets in the food industry.
I know how to make KFC chicken, by the way, and I’m just an old line cook.
You can spoil any any food in the manufacturing of it. You can certainly find some ingredients that would react badly with other ingredients.
But you can’t take a bunch of ordinary ingredients and make them toxic in the way you suggest.
Every ingredient in every standard recipe is either listed or part of “natural ingredients” or some similar wording. Natural ingredients must, in U.S. law, be from a list of ingredients that are GRAS, Generally Regarded as Safe. It’s the FDA who does the regarding. The exact composition, i.e. the recipe and the process, can be a trade secret, but all the ingredients are mostly harmless, and the mostly is only because not everyone in the world, even among national regulatory agencies, has an exact agreement as to what is GRAS. You also can’t make your recipe consist of all natural ingredients. IIRC, they have to be less than 2%. It’s a total non-issue.
Everybody in the industry knows the ingredients of Coke or KFC’s spice mix. The manufacturing process might be harder to duplicate, but that’s not the point. As DtC stated, a “secret recipe” is a marketing gimmick. Many people can’t tell the difference between Coke and Pepsi in a blind taste test. And if you tell people they were getting Coke and Pepsi you get the vast majority coming down on one side or the other even if you substituted two other colas. If you duplicated Coke you’d get sued out of business and for what? Nobody would buy the product in the first place because nobody would believe it really tasted like Coke. New Coke tasted like Pepsi because a focus group said to make it that way. (OK, it was a touch more complicated but not much more.) It failed because people decided that it wasn’t what they wanted to come out of a Coke can. That’s all a recipe needs to be.
I wonder how patents figure. If you get one, you have to make your process known, but eventually it loses protection. OTOH you could gamble on secrecy and hope the secret remains private longer than a patent protection… or do industrial secrets usually unpatentable things?
Industrial secrets are patentable. Nobody does so because then they aren’t secrets anymore.
Makes me wonder how things like bread and beer were invented I mean if you are a few degrees off or one step differently and you get nothing
It depends on the industry. If your patent allows you to make your product in a specific way that is efficient but prevents your competitor from doing so it can really handcuff those competitors who have to make it in a less cost-effective or efficient manner until they find a better process.
That’s how we get such a wide variety of breads and beers. 
That’s certainly why you patent something, but it’s pretty much the opposite of a secret.
Ah, I didn’t realise you were trying to make a joke.
There are product patents and process patents. You can see your competitor’s product in the market, and if it infringes a patent you have on that product, you can sue him. But if you have a process patent, and your competitor is infringing it, you generally don’t know it. For this reason trade secrecy can be the better choice.
However, if you keep something secret, and somebody else figures it out, they might be able to patent it and prevent you from practicing it anymore.
Wouldn’t the original company be able to contest the patent based on their prior art? Or is a procedure not covered under the prior art concept?
It’s pretty hard to keep as ‘top secret’ anything made as a commercial product. If it’s made in commercial volume, you probably have a factory somewhere to produce it, filled with workers who do the actual production. A competitor who really wants to know the secret can just bribe one of those workers to tell, or even get a spy hired on at the factory.
You can limit this somewhat by keeping part of the operation restricted. For example Kentucky Fried Chicken is flavored by a ‘secret’ blend of 7 herbs & spices. And the minimum wage workers at each KFC don’t know what it is – they just use boxes of premixed flavoring. But somewhere there is a KFC factory that mixes individual herbs & spices to produce those boxes. So a competitor would have to infiltrate a spy into this operation to get the ‘secret’. Even if the ingredients came in unlabeled, there is still a staff of purchasing agents that could be infiltrated.
All these measures reduce the number of people who know the secret, but there are still some. And a fair number, if it’s really produced in commercial volume.
IANAL but as I understand it, prior art that was never made public is irrelevant in the granting of a patent. The point of patents is an exchange - 20 years of court-enforced monopoly is traded for teaching what otherwise would have been a secret, so everybody can use this hard earned information for free at the end of the 20 years. Each patent has teachings, which provide the inventor’s end of the bargain, and claims, which define the court’s end of the bargain.
If you are sitting on valuable teachings rather than letting the world benefit, the government and the people have an interest in making this exchange with somebody else that has the same teachings and offers them in trade.
Big Secrets stole a bag of KFC seasoned flour mix and had it chemically anlyzed. The only seasonings in it were salt, pepper and MSG. Maybe at one time it really had “11 herbs and spices,” but not anymore. The really essential component of KFC chicken is how it’s fried (in pressure cookers), not the seasoning mix.
I read somewhere that the secret formula for Coca Cola isn’t really a secret anymore and hasn’t been for many years.
However, that knowledge is irrelevant since nobody wants to make their soda taste like Coca Cola. Pepsi wants to make their soda taste like Pepsi and Dr. Pepper wants to make their product taste like Dr. Pepper.
And it tastes exactly the same as shop bought KFC?
I can get it pretty close at home, but to to get it exactly, you need the specialized cooker.
When WWII was imminent, dentists used diamond studded drills that were imported from Germany that became unavailable after 1939. My father helped developed a process for making them. They could have patented it, but chose not to, feeling that no one could duplicate the process and perhaps their monopoly could last longer. (I would figure that what one fool could do, another one could too.) So they kept the process secret, rented a small house next door to their main factory and allowed only a few people to enter, etc. It worked too, but around the time the patent would have run out, dentists started using high speed drills that didn’t need diamond tipped drills (in fact would have cut so fast as to be difficult to control) so their monopoly became worthless. Although they did exploit the same technology to make diamond tipped cutting wheels and, eventually, nail files. The basic process was obvious, electroplate nickel in such a way that diamond dust is physically entrapped. But the devil was in the details. Knowing a few of the details, I could probably duplicate the process but only after months of hard work and experimentation. Without those details I know, it would take considerably longer.
Yeah, practically everything is proprietary.
Consider a microprocessor. The manufacturing processes Intel (and everyone else) uses are highly secret. But the detailed layout is also secret, beyond the RTL and gate level design. Theoretically you could reverse engineer it, but in my experience hearing about people looking for defects it is almost impossible starting with the design and layout information to see what is going on at a given metal layer, let alone starting from scratch.
When I was at Bell Labs a design manager for our DSP said that he’d be happy if the gate level description would leak out, since most of the value add was in the custom design, and someone directly implementing the DSP from the gate level netlist would produce a slow piece of junk.