Off course it's trial and error, what else can it be? I suspect that you may be underestimating the terror of constant life in the shadow of a famine. This may give an idea:

A woman doctor wrote to a friend in June 1933 that she had not yet become a cannibal, but was "not sure that I shall not be one by the time my letter reaches you." The good people died first. Those who refused to steal or to prostitute themselves died. Those who gave food to others died. Those who refused to eat corpses died. Those who refused to kill their fellow man died. Parents who resisted cannibalism died before their children did.

https://en.m.wikipedia.org/wiki/Holodomor

With that kind of reality all sorts of experimenting will be happening.

"If honor requires that I starve, then I starve."

Honor has been collectively adaptive often enough to have arisen in many places, despite local failings.

> The case with the cake is even more mesmerizing, it slowly blocked overtime their ability to process B1. How could people tell that it was from that food??

Keep the information about ill effects as part of the cultural package alongside the prep rituals? That way you can kinda sorta notice that the ill effects still occur and cooks can try to add new preparation methods.

Alternatively, simple evolutionary process: cooks semi-randomly add new steps (whether because luck or opportunity), some tribes push back deadly effects further back and thus have better survival rates, leading to their preparation methods spreading. Nardoo occurs throughout Australia, so native populations would have tried using it for a very long time (tens of thousands of years).

What's even more puzzling than people figuring out complex processes with multiple non obvious steps to successfully remove increasing amounts of (slow acting) poison from plants in relatively short amounts of time is the contrast to medicine.

It took people literally millenia to figure out pretty obvious things. That blood letting probably wasn't such a hot idea for most ailments. Or that the awesome cadaverous smell of your hands that you took such professional pride in basically meant that your obstetric services massively decreased both mother and child's survival chances compared to no outside help at all.

[Edit: I wonder if the difference can be purely explained by people being much better at figuring out what increases their and their kin's survival chances compared to those of others]

> How could people tell that it was from that food? And how do they go, yeah, we just forgot to do X

There's no need for people to understand what is going on. People have a tendency to try random stuff without any reason. The folks who accidentally do the right thing survive, while the others die.

And it's also important to note that probably not a single person came up with the whole procedure. People might use a recipe that worked for a similar plant, and adapt it. A lot of people probably died or became ill. They likely never realized their mistake. It's just that over the years the people who did it right had an evolutionary advantage and survived, and now we wonder how they figured it out, when it was all just an accident.

Not only trying random stuff but being subject to random events -- you have to flee camp, you come back a couple of days later to find your porridge has 'grown', you leave it by the fire, get distracted, come back later and you've got some leaven 'bread'.

You carry milk in a gourd, the end of the journey and it's separated, you try to make the same again but go too far end up with butter ...

> There's no need for people to understand what is going on. People have a tendency to try random stuff without any reason.

You're not giving people very much credit. The things they try may seem random to an outside observer, but they made sense to the person trying them. They had some mental model of how it works, and applied that model. Sometimes even a very inaccurate model is good enough.

Yeah, but it's an important point that taking risks without a model can lead to solutions that would not be found otherwise. Suppose there are ten options for survival and 100 people who are under threat. If they try to develop a theory and all fail, and do not act without hope of success, then they all die. If they develop a theory and follow it, and it is wrong, then they all die. But if at least one option is correct, and they simply try them at random, disagreeing with each other, then about ten people will survive and pass on their genes and/or knowledge.

People are not 100% rational, and being 100% rational is not evolutionarily optimal.

I think you are giving people too much credit. I feel like most of the time the mental model only comes after the fact, in an attemp to rationalize whatever ridiculous things people do.

Look at people who flock to homeopathy, or those who buy "water vitalisation" devices, or other ridiculous gimmicks. There's no rational reason why people would come up with ridiculous things like that. Unless you look at it from an evolutionary perspective. These ridiculous ideas are just people randomly exploring the search space, and natural selection picks the ones who got something right.

> Look at people who flock to homeopathy, or those who buy "water vitalisation" devices, or other ridiculous gimmicks. There's no rational reason why people would come up with ridiculous things like that.

What are you talking about? Those things absolutely have a model behind them. It's an incredibly bad model that doesn't stand up to any serious scrutiny, but it is a model none the less.

> These ridiculous ideas are just people randomly exploring the search space, and natural selection picks the ones who got something right.

I think you might be giving the word "model" too much weight. It doesn't have to be a mathematical proof or anything. "God won't let me die because I'm too pretty" is a model.

you could say its something more but thats a question of belief. most of these cultures will tell you eventually they have learnt it either from ancestors or from spirits / god. that being nothing which can be proven by the scientific method, for science, it's nothing more, and if thats what you believe, cultural knowledge and trial and error is all that remains as a possibility.

lots of cultures carry knowledge which is 'strange' for them to have or carry. how they came about it is often a mystery at best.

In simpler times, with fewer distractions and less variety in food, it's possible there were fewer variables to account for.

If a poisonous plant is abundant then the first tribe that can eat the poisonous plant will out compete everyone else.

Someone didn't like the cake, and they were fine, maybe?