I’ve been reading a bit about “data-to-concept” models in education. The idea is to let students build models from the ground up rather than just give them the theory first. This approach has its advantages, and I think this line of thought can be applied to diets as well.
For example, there is that tendency to accept a dietary concept without much data to support it. A person might read about Paleo or vegan or whatever, and then decide to implement it based on theoretical grounds. But does it actually work for the person? In other words, does the person find data (how they feel, etc.) to support the concept?
Compare this to building a diet from the ground up. With this method, a person might cycle different foods in and out to see what works for them. Whatever dietary concept they end up with, it was built via data-to-concept.
Of course, there can be some value of going from concept to data. For example, say a person reads about Paleo and then experiments with dropping milk. This would be concept-to-data. But then does dropping milk make them feel better or worse? This would then be data-to-concept (to a possibly new concept).
Here’s one example of when concept-to-data goes awry. This person bought into the vegan diet for 25 years, and was always in terrible health. He swallowed the vegan concept while ignoring the data from his own body that contradicted the theory. And this persisted for 25 years! Eventually, he started eating meat and his problems went away.
I fell into the this concept-to-data trap a few years ago in regards to rice. At some point years ago, I dropped rice just because it wasn’t “Paleo”. And even though I subsequently felt worse without rice, I continued on this path for some time! At some point, common sense took over and I added the rice back in.
So while a mix of data and theory is probably optimal, I think there should be more emphasis on how different diets make people feel, rather than if they fit into a conceptual box.