The smart way to learn about the world is this: Start with no conclusions. Draw in data. Form a tentative conclusion from those. Revise conclusion as new data require.
The common "culturology" way to do it: Start with a conclusion. Welcome all data that fit. Label data that don't fit as "exceptions" or "a paradox".
The mistake is called confirmation bias. Here's a really trivial example – meaningless by any measure, but worth pointing out because, hey, nice pizza ad!
For those not hep to the lingo, that's a Fall 2007-model Pizza Hut Gourmet pie, mixing Sausage Roll crust, three-cheese Cheesy Roll crust, little hamburger patties, and other goodies, into one circle of awesomeness. (With "fruity catchup" for the sausage crust, and honey maple syrup for the cheesy crust!) The pic is from a post at food blog Slashfood, which reads as such:
For as much as Japanese cuisine is about delicate, precious presentation and things as simple and stunning as sushi (not to mention healthy), it often makes me laugh to see some of the absolutely ridiculous things that come out of their culinary creativity.
No need to take the post too seriously. But for the heck of it, it's a chance to make this observation:
Japanese cuisine is not "about delicate, precious presentation and things as simple and stunning as sushi". It's about whatever people in Japan eat. That includes beautifully-presented sushi and greasy pizzas alike. Artistic kaiseki ryouri and four-patty Mega Mac burgers. And everything else we could point at. Those are the data; those are what define the cuisine. No data point is an exception or a paradox or anything of the sort!
That's about all there is to say here. Again, the above is a trivial example that I assume wasn't meant to be a serious cultural comment anyway (no disrespect toward the fine Slashfood people!). But the data-before-conclusion principle is important, and I promise will be revisited often.