From Humanoid Faults to Manipulated Consensus, 35

Advertising and propaganda are an unlimited source of examples: when they want to sell you a car, they pretend to be selling you a whole beautifully desirable world surrounding it, which is totally unreal; when they want to discredit an opponent, they put him or her close to something politically incorrect, that may be just as unreal, or actually totally extraneous anyway.
Indeed, metaphors, parables and other forms of association are seemingly used solely to increase the impingement of something by associating it with something else more familiar to the target audience, but even in rhetoric there’s more than meets the eye as to our being associative: “white” becomes “snow white”, “cold” “ice cold”, “hot” “hot like hell”, “that thing” “that hell of a thing”, the “lost” becomes “lost sheep” and the “rescuer” “good shepherd”, and so on.
Well, what we’re not much aware of is that these are not just mere intensifiers: some of what they mean to us seeps into what they’re associated with. Hence, to us, that white will taste of snow, that cold will taste of ice, that hot will taste of hell, that thing will be somewhat hellish, rescue will imply submission, and so on.

We specialise in merging the book and its cover, the datum with its presentation, and then in judging the book by the cover, persuaded that the presentation of the datum is the datum. You say it’s too unconcealed to work? Remember we’re quantitative: let’s take a second look at it after they worked at it enough. So it’s not that what’s wrong with being associative is overdoing it: there is not an acceptable level of being associative, it’s a pitfall in itself. It’s playing into the hands of our manipulators.