Your Metaphor is Misleading Me

When Federal Reserve Chief Ben Bernanke appeared on the TV news show 60 Minutes to persuade us to bail out the banking system, he didn’t bother with charts, figures, or lengthy argument. Instead, he used something far more powerful: Analogy and metaphor.

Imagine, he explained, that you have an irresponsible neighbor who smokes in bed, and sets fire to his house. Should you call the fire department, or should you simply walk away and let him face the consequences of his actions? What if your house—indeed all the houses in the entire neighborhood—are also made of wood? We all agree, he argued, that under those circumstances, we should focus on putting out the fire first. Then we can turn to the issues of assigning blame or punishment, re-writing the fire code, and putting fail-safes in place.

This was a powerful analogy. It communicated the clear and present danger to the economy and the urgency of implementing his proposed solution. Why did we find this type of argument so persuasive? Because, as cognitive scientists Keith Holyoak and Paul Thagard (1989) claim, the “analogical mind is simply the mind of a normal human being.” Or as fellow cognitive scientist and Pulitzer Prize winner Douglas Hofstadter (2009) put it, analogy is “the core of cognition." We can’t help noticing that “this is like that," and arguments that exploit this natural tendency tend to hold sway. This article is not about the bailout and its merits or failures. It is about the reasoning that underlie it.

Lawyers use analogies frequently to draw parallels between an undecided case and a case that has already been decided (a precedent). They will argue that “this case is like that case” and hence the same judgment should be applied. If the cases match up perfectly, the argument is referred to as “on point” (or, in common parlance, a slam-dunk). Some legal scholars, such as Lloyd Weinreb, go so far as to argue that without analogy, a court’s decision is incomplete. It is that important in explaining and justifying legal judgments.

So what is an analogy? It is a relational similarity. Two objects or events are not considered analogs because they refer to the same things but because the structure of the relationships among their elements is the same. Bernanke’s analogy was persuasive not because a burning house has anything physically in common with a failing bank, but because, at a deeper level of analysis, they describe similar relations: A burning house is in danger of physical collapse. A failing bank is in danger of financial collapse. More importantly, both require intervention if the collapse is to be prevented. We could say, then, that the surface features of the situations may differ, but the relationships among the features are the same. Bernanke did a good job of pointing out the similarity between these structures, and hence we found the argument very persuasive. The next step was then simple: If the two problems are so similar, then the solutions should be similar as well: The solution to the burning house scenario is pouring water onto the fire. So, by analogical transfer, the solution to the failing bank scenario is pouring money into the banking industry. The water stops the fire, and the bailout money stops the asset accounts from losing value.

Skeptics skewered this reasoning by making use of analogy and metaphor themselves: While Bernanke argued that “this is like that”, critics argued that “this” was a lot more like something else. For example, Michael Hudson of the Centre for Research on Globalization argued that the analogy was bogus and hence the argument itself was bogus as well. He attacked a crucial relational component that, once eliminated, made this argument collapse like a house of cards. The financiers were not our neighbors, he argued. Instead, they were “the castle on the hill, lording it over the town below.” What impacts one house in a neighborhood typically impacts other houses—and their occupants—as well. The bail out wouldn’t put out the fire in the neighborhood. Instead, it would empower the castle lords to take over the neighborhoods by “taking over houses that have not burned down, throwing out their homeowners and occupants, and turning the property over to the culprits who ‘burned down their own house.’” Yes, he conceded, our economic house was on fire. But now that we realize the bankers don’t live where we live, the real solution “writing off the debts of the economy – the debts that are burning it down."

So how do we evaluate analogical arguments? Here are two crucial steps:

First—and this is usually the most difficult--ask yourself “what else does this remind me of?” It is almost impossible to properly evaluate a proffered analogy-based solution unless you compare it to other relevant solutions. Bernanke’s analogy seemed like a “slam dunk” until other analogies came to light that were better fits.

Second, compare the similarity between the systems of relationships among the cases. The greater the number of relational matches you find, the stronger the analogy and the more likely it is that the solution is the best one for the unsolved case. But if you find a crucial mismatch, the analogy is suspect.

In courts of law, decisions frequently depend on which analog best matches the undecided case. This is your best strategy: Generate alternative analogies, and choose the best match.