Most language processing takes places in the left hemisphere of the brain. When we read, carry on a conversation, or listen to speech, most of the action — for right-handers — takes place on the left side of the brain. (For left-handers, the situation is more complex; it’s not simply a mirror image of a right-handed brain. For this reason, most studies involving any sort of brain scan routinely exclude left-handers.)

But there are a few occasions when the right hemisphere gets involved: when we create a narrative, for example, or when we make inferences. Some types of figurative language, such as metaphor and irony, also involve the right hemisphere. But research on perception of metaphor and irony has had difficulty identifying differences in how the two processes work.

Arguably, there isn’t much difference between an ironic statement and a metaphor. Irony involves saying the opposite of what we mean (“George W. Bush is so smart!”), and metaphor involves using language that is not literally true (when we say “Donald Rumsfeld is a lightning rod for criticism” we don’t mean that we place him on top of tall, exposed buildings during thunderstorms).

Some theorists make a distinction between “conventional metaphor” and “novel metaphor.” Conventional metaphors are metaphors that are so common that they often aren’t even perceived as metaphors (“his argument was rock-solid”), and several studies have found that we process conventional metaphors just as quickly as literal statements. But is there really no difference in processing these common metaphors and literal statements? What about common ironic statements? Zohar Eviatar and Marcel Adam Just paid volunteers to submit to an fMRI scan while they read literal, metaphoric, or ironic statements.

In each case, a context was displayed (“Johnny went on a hike with his brother. Suddenly he saw a huge snake next to his foot”). Then the screen went bank for a few seconds. Then the test statement was made (in this case, literal: “He said, ‘I am so scared’”). After another break of a few seconds, the volunteers were asked a quick yes-or-no question to ensure they were paying attention (“was Johnny scared?”).

Here’s an example of a metaphor from the test:

In the morning John came to work early. He started to work right away at a fast pace.

His boss said, “John is a hurricane.”

And here’s an irony example:

George went to Betty’s birthday party. Only two people came.

He said, “It is really crowded here.”

Clearly the point of the study was not to examine our understanding of fine literature — these are the sorts of ironic and metaphoric statements we make every day.

So is there a difference in brain activity while processing irony, metaphor, and literal language? Take a look at this image:

All the images show where brain activity (measured by blood flow) is significantly higher on average while reading the key statement (literal, ironic, or metaphoric) than during rest. The top row shows the left hemisphere, where as you can see, much more is going on compared to right hemisphere. This, for the most part, represents the basic language-processing functions which have been observed by many researchers.

While it may appear that there are fewer differences in left-hemisphere activity compared to the right (in the bottom row), the researchers actually identified four regions in the left hemisphere which had significant differences in activity when different types of statements were being read. They also found two areas in the right hemisphere with similar activity differences.

Not only were these common metaphors and ironic statements processed differently from ordinary language, they were processed differently from each other. So, it seems, our semantic distinction between irony and metaphor actually corresponds to real differences in how the brain processes those statements.

I can already see the objection coming from my seventh-grad English teacher: What about similes? Actually, most psychologists don’t distinguish between similes (comparisons using “like” or “as”) and metaphors (direct, equivalence comparisons). Both were employed in this study.

Eviatar, Z., & Just., M.A. (2006). Brain correlates of discourse processing: An fMRI investigation of irony and conventional metaphor comprehension. Neurpsychologia, 44, 2348-2359.