Time flies like an arrow; fruit flies like a banana.
For psychologists who study humor, this statement is a classic. It embodies the ambiguity of language that much humor exploits. In this case, the words “flies” and “like” have different meanings that come into conflict in the reader’s mind. The way our cognitive processes resolve this conflict lies at the heart of the nature of humor, say theorists.
Humor showcases the speed and flexibility of human cognition at its most impressive. Clearly, the ability to reproduce this behavior would be hugely useful in machines that could appreciate humor and generate laughs.
So psychologists and computer scientists would dearly love to understand and reproduce the cognitive processes behind humor. Sadly, progress in this area has been slow, not least because it is hard to properly model this cognitive conflict.
Today, that changes, at least in part, thanks to the work of Liane Gabora at the University of British Columbia in Canada and Kirsty Kitto at the Queensland University of Technology in Australia. These guys have created a new model of humor based on the mathematical formalism of quantum theory. They then apply it to verbal puns and cartoons.
The basic problem with modeling humor is to find a way to represent a joke at the moment it is understood. That’s tricky because it requires the ability to able to handle two or more conflicting interpretations at the same time.
In the joke above, the brain first assimilates the set-up statement “time flies like an arrow,” in which flies is verb meaning “to travel through the air.” It then assimilates the punch line statement “fruit flies like a banana,” in which flies is a noun describing flying insects.
By themselves, these phrases are not particularly amusing. The humor arises when the meaning from the set-up phrase clashes with the meaning in the punch line. This clash requires the brain to hold both meanings at the same time.
Gabora and Kitto say the process of holding two ideas simultaneously in our brains is analogous to the process of quantum superposition. This is the bizarre quantum phenomenon in which a single object can exist in two places at the same time. The object’s position only becomes localized when it is measured and the superposition collapses.
Similarly, the brain holds two meanings in mind at the same time and the process of getting a joke resolves this conflict as the brain settles on one meaning or the other. Gabora and Kitto’s idea is that the mathematics behind quantum superposition can also model this kind of double-think.
They are not saying that the brain relies on quantum processes, only that quantum formalism can be used to model it. “The quantum approach enables us to naturally represent the process of ‘getting a joke,’ ” they say.
A crucial part of this is the context of the joke. Humor is famously context-dependent—the same joke can be funny or not depending on factors such as the way it is told, the personal circumstances of the listener, and so on. In theory, the quantum formalism allows all this to be taken into account so the probability that the listener finds the joke funny can depend on the context.
So what evidence is there that the quantum approach works? Gabora and Kitto point to classical probability theory, which predicts that the average funniness of a joke should be the sum of the funniness of each of its possible interpretations. Their contention is that any deviation from this prediction could suggest that quantum thinking might be a better approach.
To find out, Gabora and Kitto gather evidence by measuring the way people evaluate the humor in jokes, in variations of the same jokes and in the set-up lines and punch lines alone. They do this by asking 85 undergraduates to fill in a survey in which they have to rate the funniness of statements on a scale of 1 to 5 (where 5 is hilarious).
It turns out that the total funniness does not equal the sum of the funniness of all the interpretations. But why not? One possibility is that humor does not follow a classical model but another is that there is some problem with the experiment itself.
Gabora and Kitto prefer the former explanation and go even further. They say, hopefully, that it provides preliminary evidence of their own theory. “We have preliminary evidence that humor should perhaps be treated using a quantum inspired model,” they say.
Others might not be so generous. The fact that their data does not match the predictions of classical probability theory is not evidence that quantum theory would do better.
More likely is that the result is a reflection of shortcomings in the experimental method itself. Indeed, Gabora and Kitto happily acknowledge that their experiment does not really measure the funniness of all possible interpretations in every context. So it’s not really surprising that the sums do not add up.
The problem, of course, is that there is no known way of measuring the funniness of all interpretations. And therein lies at least part of the challenge.
For the moment, it looks as if the foundations of humor are set to elude researchers for some time to come.
But while quantum theory can’t model humor, at least it can still provide the odd smirk.
Q: Why won't P and X live in the suburbs?
A: Because they don't commute.
Ref: arxiv.org/abs/1703.04647 : Towards a Quantum Theory of Humour