Ketchup Duck

What are switcheroo contrarian quotes?

Edit: Just found out these are actually called antimetaboles.

There is a category of quotes that I’m calling the “switcheroo contrarian quotes.” The most famous example is probably from John F. Kennedy’s speech:

Ask not what your country can do for you; ask what you can do for your country.

The general structure of switcheroo contrarian quotes is:

Blah₁ A blah₂ B blah₃, [word indicating disagreement, optional] blah₁ B blah₂ A blah₃.

This structure shows up in Kennedy’s quote:

Ask not what your country can do for you; ask what you can do for your country.

I call these sentences “switcheroo” sentences because the first part of the sentence says A and B, and the second part says the same thing but switches A and B. These sentences are contrarian because the author indicates disagreement with the first part of the sentence, and supports the second part.¹ They are quotes because a switcheroo contrarian sentence lands with rhetorical flair, which makes the sentence impactful, and thus repeatable.

Let’s look at the Kennedy quote again. The first part goes “Ask not what your country can do for you.” But why not? Why shouldn’t Americans ask what the country can do for them? After all, the “government of the people, by the people, for the people” should serve the people. Politicians get elected by promising all sorts of things the country will do for the people. If Kennedy had gotten up on stage that day and said, “Ask what your country can do for you, for I will deliver,” no one would have batted an eye. The first part of the quote is immensely reasonable. That’s the secret of the switcheroo contrarian quote; it leads you in with a proposition (like asking what your country can do for you) that seems reasonable.

The fact that the first part seems reasonable makes the eventual contrarianism seem more intelligent. Anyone can object to an objectionable idea, but smart people object to reasonable-sounding ideas. Gandhi objected to the reasonable-sounding idea that “you have to physically fight your oppressors to defeat them.” Einstein objected to the reasonable-sounding idea that “spacetime does not change based on your frame of reference.” Judith Rich Harris objected to the reasonable-sounding idea that “parenting molds children.” There is a correlation between intelligence and willingness to reject prevailing ideas. This has the corollary that objecting to prevailing ideas is a good way to signal intelligence, regardless of whether you are actually intelligent or not. This, the contrarianism in switcharoo contrarian quotes sounds intelligent, precisely because the contrarianism is directed towards a reasonable proposition.

Finally, the second part of the quote switches A and B to indicate complete disagreement with the first part of the quote. The speaker is not indicating mild disagreement; he is turning the first part of the quote on its head. On top of that, the speaker repurposes the same words to prove an opposing position; the first part of the sentence gets hoisted by its own petard. The second part sounds much more convincing because of what came before it than it would on its own. After all, if JFK had just said “you should ask what you can do for your country,” we wouldn’t still remember his words.

Blah₁ A blah₂ B blah₃, [word indicating disagreement, optional] blah₁ B blah₂ A blah₃.

All this combines to make the above switcheroo contrarian formula deliciously effective. In fact, it’s so effective that it can be used to say anything at all.

JFK encouraged patriotism and public-mindedness with Ask not what your country can do for you; ask what you can do for your country.

A more libertarian politician could easily repurpose the quote to say Ask not what you can do for your country; ask what your country can do for you.

Both quotes have the same amount of persuasiveness. If you had never been exposed to the original, you would probably find both quotes equally convincing. The switcheroo contrarian formula makes ideas sound convincing, regardless of what the idea is. If you are a speaker, it is a useful tool to add to your arsenal. If you are a listener, you would be well served if you divorce the substance of switcheroo contrarian quotes from their presentation, and judge the former on its merits.

…

…

…

So far, so good. I’ve laid out a theory, and in its support I’ve provided some convincing examples and reasoning. It’s a common blogging genre. But I have not proven anything. Let me try to do so now.

Trust, but verify

Here is my hypothesis: Switcharoo contrarian quotes can be just as persuasive if they are repurposed to say the opposite of what they originally said. In other words, their persuasiveness stems mostly from their structure, not their content.

I created a survey in which people were asked to rate five switcheroo contrarian quotes on a scale from 1 (strongly disagree) to 7 (strongly agree). All quotes were presented without their source.² The quotes were:

1. (V1) I fear not the man who has practiced 10,000 kicks once, but I fear the man who has practiced one kick 10,000 times.
2. (V2) A wise man can play the part of a clown, but a clown can’t play the part of a wise man.
3. (V1) A fool thinks himself to be wise, but a wise man knows himself to be a fool.
4. (V2) People are more impressed by the power of our example rather than the example of our power.
5. (V1) We love people not so much for the good they've done us, as for the good we've done them.

There were two versions of the survey. In version 1, the switcheroo contrarian quotes marked V1 had A and B reversed, so that each quote said the opposite of what it originally said. In version 2, the switcheroo contrarian quotes marked V2 had A and B reversed. After I offered some money to the MTurk Gods, I had my ratings.

Results

To see whether switcheroo contrarian quotes are convincing even when doctored to say the opposite of what they originally said, I compared the ratings given to the original quote to the ratings given to the doctored quote.

shows the ratings given to the original and doctored versions of “I fear not the man who has practiced 10,000 kicks once, but I fear the man who has practiced one kick 10,000 times.” The left side of the graph represents the “disagree” half of the scale and the right side of the graph represents the “agree” half of the scale. The top of the graph indicates 50%. Here, for the original quote, 5% of raters neither agreed or disagreed, 33% somewhat agreed, 38% agreed, and 17% strongly agreed.

A brief glance at the graph is enough to judge whether the two sets of ratings are similar or dissimilar. However, to be more rigorous about measuring the similarity between the two sets of ratings, I performed the Mann-Whitney test³ on my data. This test helped me measure how similar the two sets of ratings are by giving me the chance that both sets of ratings are drawn from the same distribution. For the present quote, p = 0.01, so there is only a 1% chance that both sets of ratings are drawn from the same distribution. Thus, the ratings are pretty dissimilar. The higher the p-value, the more similar the two sets of ratings.

The results for all quotes are as follows:

1. I fear not the man who has practiced 10,000 kicks once, but I fear the man who has practiced one kick 10,000 times.
   , p = 0.01.
2. A wise man can play the part of a clown, but a clown can’t play the part of a wise man.
   , p = 0.71.
3. A fool thinks himself to be wise, but a wise man knows himself to be a fool.
   , p = 0.24.
4. People are more impressed by the power of our example rather than the example of our power.
   , p = 0.18.
5. We love people not so much for the good they've done us, as for the good we've done them.
   , p = 0.36.

Most science papers look for p < 0.05 to judge an effect significant, and by that measure, only the first quote had original ratings that were significantly different from doctored ratings. I suppose this sort of proves my point that switcheroo contrarian quotes are convincing even when doctored to say the opposite of what they originally said, although I expected the effect to be stronger than it is, and the effect varies quite a lot between different quotes. I am also a poor predictor of which quotes would be most convincing when switched, because I thought it would be quotes #1 and #4, and those are actually the quotes that were least convincing when switched.

I’m not an expert in experiment design, so I made several mistakes. If I did this again, I would use a five-point rating scale instead of a seven-point rating scale. I think a seven-point rating scale splits hairs, and indicates precision where none exists. I might also mix non-switcheroo quotes into the survey, to avoid people getting fatigued from looking at the same quote structure over and over. Survey respondents spent about 15 seconds on each question, and they might have spent more time if they had more varied quotes to look at. If I had more money to spend, I would have had more people fill the survey to get more robust results. You need to survey a hundred people to detect that men are heavier than women, and I only had 35 people fill out my survey.