Allow me to put some words in front of you.
Do you think the United States should allow public speeches against democracy?
This question dates all the way back to 1941, when it was posed in an issue of Public Opinion Quarterly. A full three-quarters of Americans (75%) responded no, anti-democracy speeches should not be allowed.
But that wasn’t the only question the article posed. In fact, they did what’s called a split-ballot experiment – where one half of a randomly-selected sample received that question, and the other half received this one:
Do you think the United States should forbid public speeches against democracy?
Did you have to read it twice? That’s because at face value, these questions seem almost identical. In fact, they only differ by six letters: one asks whether speeches should be "allowed," and the other asks whether they should be "forbidden." Functionally, they communicate the same thing from two different angles – so it should follow that the results would be analogous.
But in fact, while 75% of Americans would "not allow" speeches against democracy, only 54% would "forbid" them.
It's easy to hypothesize why this could be, but semantics isn't the point. The point is, a six-letter difference completely altered perception of the question - and by extension, the survey results. The experiment became part of survey lore, illustrating that small changes in wording can have massive (and sometimes unintended) consequences.
Why does this matter?
The average American consumes no fewer than 100,000 words each day. (For context, War & Peace is 560,000 words.)
And those words are more than just a means of communication. Research has shown that words don’t just reflect our thoughts... they can actually shape them.