Kevin Drum recently laid out a long-standing unsolved problem, one that has preoccupied such luminaries as Paul Krugman, James Fallows, and Glenn Beck ("Saving the Frogs", Mother Jones, 9/23/2009). The problem is that there's no good substitute for the over-used and untrue story about how a frog, if placed in a pot of gradually heated water, will eventually allow itself to be boiled without jumping out. And since this is a rhetorical problem, Drum describes the failure as a linguistic one:
So here's what I'm interested in. The boiling frog cliche is untrue. But it stays alive because, as Krugman says, it's a useful metaphor. So why aren't there any good substitutes?
This is very strange. Most useful adages and metaphors not only have substitutes, they have multiple substitutes. "Look before you leap" and "Curiosity killed the cat." "Fast as lightning" and "Faster than a speeding bullet." Etc. Usually you have lots of choices.
But in this case we don't seem to have a single one aside from the boiling frog. Why? Is it because it's not really all that useful a metaphor after all? Because the frog has ruthlessly killed off every competitor? Because it's not actually true in any circumstance, let alone with frogs in pots of water? What accounts for this linguistic failure?
Yesterday, Jonathan Lundell sent me a link to Drum's article, with the comment "Sounds like a job for Language Log". That was almost enough to make me move on immediately: when Geoff Pullum and I started Language Log, I promised myself that if it ever got to feel like a job, I'd quit.
But this morning, after half a cup of coffee, I realized that Jonathan's remark was just an instance of the conventionalized phrasal template "sounds like a job for ___". And this one usually refers to the super-activities of superheros, which are by definition superfluous to their day jobs.
Thus this Non Sequitur strip from 8/23/2009:
So let me start by noting that the frog-boiling business is a very different kind of cliche from the other ones that Drum cites. "Look before you leap", "curiosity killed the cat", etc., are fixed phrases, involving not only a conventionalized metaphor but also a specific string of words. The frog story has no standard linguistic form — it's a conventionalized metaphorical narrative, not a conventionalized metaphorical phrase.
In that respect, it's like the original snowclone, which involves explaining that since the Eskimos have some large number of words for snow, so the members of some other group must have even more words for some substance, activity or concept believed to be typical of them. You can explain that in any words that you like, and it still works as a rhetorical gesture, as long as your audience doesn't object to the fact that its premise is untrue.
And as far as I know, there isn't really any suitable overall substitute for this linguistic abuse of Eskimos. The 18th-century version about Arabs and lions is extinct, and my suggestion about Somalis and camels has never caught on. Similarly, I can't think of any substitute for the false story about the Chinese characters for "crisis".
So there you are: perhaps it's a rhetorical generalization that conventionalized metaphorical narratives are both false and unsubstitutable. This would follow from a couple of facts: people like to embellish stories to improve their fit to particular rhetorical circumstances, and rhetorical value is uncorrelated with truth (or perhaps negatively correlated). Based on those premises, you can show that Really Useful Stories will almost always be false, and also that Really Useful Stories will be the end point of a process of invention and memetic selection that's not easy to equal by mere intelligent design.
So what about substitutes for the frog-boiling narrative? Am I going to undermine my point by offering some?
Yes, sort of. There's the Niemöller "first they came" passage; but this is specific to the gradual spread of tyranny, and yet is unlikely to appeal to Glenn Beck, who appears to be the only pundit who has actually boiled a frog on television. There's the introduction of wide-band noise in Tinnitus retraining therapy, which must be gradual and carefully calibrated so as to avoid triggering aversive limbic responses. This is (I think) a valid instance of the false "frogs won't get upset if increases in water temperature are gradual" concept; but it's too complicated, and the result of gradual stimulus increase is good rather than bad, and anyhow curing people of annoying imaginary sounds doesn't have the emotional impact of boiling frogs. See?