I've blabbed before as to how I've often asked nutrition experts "What's so healthy about 'healthy whole grains'?" I've never gotten an actual answer, and as far as I can tell the best one could say is "nothing in particular." And while I have discussed the possible ways that grain consumption could lead to disease, I would have to admit that the evidence that grains have some particular disease-causing properties (outside of those with obvious clinically-detectable problems, like celiac) seems more correlation than causation at this point.
So I've started rethinking this question more as "why does anybody eat anything?" Clearly the need, at some level, to seek out and consume food has to be innate. And animals evolve amazingly complex behaviors around food. I remember giving my dog an egg for the first time, shell and all. As he does with any food, I expected him to swallow it more or less whole, maybe with a couple of crunches for good measure. Instead, he gently picked it up from his bowl, put it on the ground, and ever-so-delicately cracked it open with his front teeth, then licked out the inside and left the shell. I'm pretty sure that wasn't a learned behavior, unless he's been climbing trees and getting into robins' nests behind my back.
But in general, and probably particularly for omnivores, directed behavior associated with food (like "go find some more of those sweet orange spherical thingies") is learned. Babies put everything in their mouths for a reason: they're figuring out which things are worth seeking out and sticking in their mouths again. You may want to check out this fascinating paper on the topic. The short version is this: there seem to be two main areas of the brain associated with taste. The primary taste cortex handles the innate sensing of taste: sweet, salt, bitter, sour, and umami, along with the texture and viscosity of food (to sense fat), temperature, capsaicin, etc. The response of the primary taste cortex is NOT attentuated by satiety. Something sweet tastes just as sweet whether you're hungry or full. But the primary taste cortex doesn't assign value to a particular taste, i.e. it does not decide whether something tastes "good" or "bad". That's the job of the secondary taste cortex. It is the secondary taste cortex that "decides" sweet things taste good when you're hungry, but no so much after eating a whole box of candy. Secondary taste cortex neurons learn what's good and what isn't, and are further tuned to specific foods. For instance, you can be fed to satiety with fat, and certain neurons will decrease their response to further fat. But the response of those same neurons to the taste of glucose does not decrease, regardless of whether or not you're full of butter. In other words, "there's always room for dessert".
Anyway, let me get to the punch-line from the closing paragraph:
The outputs of the orbitofrontal cortex reach brain regions such as the striatum, cingulate cortex, and dorsolateral prefrontal cortex where behavioural responses to food may be elicited because these structures produce behaviour which makes the orbitofrontal cortex reward neurons fire, as they represent a goal for behaviour. At the same time, outputs from the orbitofrontal cortex and amygdala, in part via the hypothalamus, may provide for appropriate autonomic and endocrine responses to food to be produced, including the release of hormones such as insulin.
In other words, the external response to food (behavior) is a learned response driven by the secondary taste cortex, while the internal response (e.g. hormonal) is innate, originating in the primary taste cortex. That means that you learn what things taste "good" by the secondary taste cortex integrating feedback (positive and negative) from the rest of the body (primary taste cortex, glucose sensors, etc.), reinforcing or weakening the association of that taste with the behavior that led to those stimuli. So the fact that you "like" potato chips is intimately tied up with the impulse to get off the coach at midnight and stumble into the kitchen to finish off the bag. And the only reason you "like" any food is because your brain learned to, associating the flavor with some feedback signals which it interprets as being associated with a net positive outcome.
One other point which is probably obvious, but important: the smaller the time between the flavor stimulus and relevant physiological response, the stronger the change in association with the behavior. Thus, getting cancer 10 years after eating a poisonous plant is not very helpful in weakening that behavior. It is certainly possible to crave something that produces a strong short-term reward, but has a net negative outcome. The brain (both consciously and unconsciously) is notably short-sighted in its assessment of value.
Which brings me back to the original question: why do people eat grains? And I don't mean that as implying there's some moral judgment to made - food morality is just another religion. And there's obviously a spectrum of answers depending on the temporal proximity of the act of eating to a specific endpoint. On end is "prepared properly, they taste good" (I like sourdough toast dripping in butter as much as the next guy, though I eat it rarely). On the other end is the evolutionary argument so brilliantly put forth by Kurt Harris, basically that the net effect of domesticating grains was an advantage in reproductive fitness over hunter-gatherers, regardless of the relative "health" of those doing the reproducing. Evolution cares about making babies, and doesn't care if you have bad teeth and a bum ticker, as long as you contribute genes to more babies than the guy still killing perfectly serviceable beasts of burden with a rock on a stick.
No, I'm interested in the middle area (logarithmically speaking), which is why we learned to like grains. And why do we like them so much that we're willing to go to some amount of trouble to eat them? Why do I so love sourdough toast and butter, even though it doinks my blood sugar and gives me acne?
(Maybe it's the butter - New Zealand makes REALLY good butter.)
I have nothing but vague guesses, and am hoping to get some interesting discussion in the comments.