In 1932, Frederic Bartlett developed the concept of "schema." Take the classic school-game "telephone" (which I used to illustrate the introduction of error in information processing in Chapter 9 of "A Theory of Power") , but now, rather than passing a verbal message, consider passing along a picture. The first person in a chain is given a very unique drawing of a human face--say something that Pablo Picasso may have drawn--and told to quickly copy the drawing and pass their copy on to the next person. Each person down the line is only given the sketch of produced by the prior person, not the original drawing. What happens is that the "uniqueness" of the original drawing is lost. Unlike the relatives "flat" linguistic topography of a verbal message, the errors in information processing introduced in this exercise trend the developing image toward a norm for what humans consider to be a generic face. This generic face is an example of Bartlett's "schema." It is an example of when the errors introduced into information processing converge on an ontogenically established norm (because of a "valley" topography of our neural circuitry) rather than diverging randomly.
Another example of schema is the categorization of enemy/friend. This schema has been defined by our ontogeny and hardwired to a large degree into our neural circuitry by observed similarities between "us" an "them" (a theme that, on a side note, was borrowed from Robert Anton Wilson by the writers of the series "Lost"). This creates an interesting epiphenomenon in our modern connectivity society: talking about any group that is different than "us," especially talking about those differences, leads to the schematic categorization of "enemy." This is even true to the point that talking *about* any group, rather than *with* that group is autonomically interpreted by human brains as instruction to categorize that group as "enemy." Of course, the context of "enemy" varies, from social competitors to military opponents, but the basic categorization preempts rationality.
So, if--hypothetically--you want to polarize the American people against Iran by identifying them as "our" enemy, all that is necessary is to leverage the neural circuitry underlying this schema by starting a discussion about them. It really doesn't matter what we say about "them," as long as the entire discussion is framed in the context of "us" and "them." Long hard-wired neural circuitry that we all share, that preempts rationality, and that *did* work quite well for us on the African savannas assures that "they" become the enemy.