There is a fish preparation from the coast that I have now attempted in three different cities, following the same instructions from the same person, and it has never once tasted right.
Other things I cook from the same region travel perfectly well. It took a while to work out why this one does not, and the answer turns out to be a useful test for any regional dish.
Some dishes are about technique and some are about ingredients
This is the whole distinction and everything else follows from it.
A dish built on technique — how long you brown something, the order of operations, the heat, the timing — travels. You can take that technique anywhere and apply it to whatever is available, and the result will be recognisably the thing.
A dish built on a specific ingredient does not travel, because the ingredient does not exist elsewhere in the same form. Substituting changes what the dish is, and no amount of skill compensates.
The fish preparation is the second kind. It depends on a fish caught that morning, of a species that does not exist where I live, with a texture and a fat content that nothing available to me shares.
The ingredients that genuinely cannot be replaced
Having thought about this across a lot of failed attempts, they cluster into categories.
Anything that depends on absolute freshness. Fish within hours of the water, curd set that morning, coconut scraped from a nut opened an hour ago. The gap between these and their shop equivalents is not a matter of quality grading, it is a different substance.
Anything fermented locally, where the microbial population is regional. Pickles, certain batters, some dairy. These acquire a character from the place that a recipe cannot carry.
Water, which sounds precious and is real for anything where water is a major component. Hardness affects pulses, tea, and dough noticeably.
And varieties of things that are exported in one form only. The rice available internationally is a narrow selection of what exists. The chillies likewise, and the difference between chilli varieties is enormous — heat, fruitiness, colour and aroma vary independently.
The substitution trap
What I did wrong for a long time was substitute one ingredient at a time and assume the sum would be close enough.
Different fish, but the same recipe. Desiccated coconut instead of fresh, but more of it. Shop yoghurt instead of home-set, hung to thicken.
Each substitution is defensible individually. Four of them together produce something that shares a name with the original and very little else.
What works better is deciding whether the dish is makeable at all, and if it is not, cooking something else from the same tradition that is.
The dishes that travel best
Usefully, there are patterns here too.
Anything built on dried pulses and grains, because those store and ship without much loss.
Anything where the aromatics are dried spices, which travel well if you buy them whole and grind them.
Anything slow-cooked, where long cooking evens out variation in the starting ingredients.
And anything where the technique is the difficult part, because you can carry the technique in your hands.
This is not a coincidence. The dishes that spread furthest from their origins historically are largely these, and the ones that stayed local are the ones dependent on something local.
What I do about the fish now
Two things, and I have made peace with both.
When I visit, I eat it there, made by somebody who has the right fish. That is not a compromise, that is the correct way to have it.
At home, I cook the technique with what I have and call it something else in my head. The same souring agent, the same spice treatment, the same sequence, applied to a fish that actually lives in the sea near me.
It is a good dish. It is not the dish. Being clear about that with myself stopped the frustration entirely, and the food improved once I stopped trying to make it be something it could not be.
Why this matters beyond one recipe
A lot of disappointment with home cooking comes from attempting things that cannot be done in the kitchen attempting them, and then concluding that the cook is the problem.
Sometimes the cook is the problem. Quite often the ingredient is, and there is no version of trying harder that fixes it.
Learning to tell which is which is genuinely useful. It saves effort, it saves money on ingredients that will not work, and it directs attention toward the enormous number of dishes that do travel and can be made properly with what is nearby.
The test I apply now, before starting anything unfamiliar, is to ask which ingredient the dish is really about, and whether I have it or something honestly close to it. If the answer is no, I cook something else, and I have never regretted that choice.