If you like idioms, you’ll be… head over heels… for this research. Allow me to… break the ice… without… beating around the bush. A new study on idioms by Nicholas Griffen and Ira Noveck is… the bee’s knees.
In second-language learning classes, ever notice that idiomatic expressions are covered in the more advanced sections? These figurative expressions seem to be more challenging for language learners to grasp. This may be due to the complex properties that are inherent in these phrases. Sometimes even native speakers get it wrong—perhaps referring to a “doggy dog world,” an “escape goat,” “biting my time,” or “peaking my interest.”
You can’t judge a book by its cover, but you might judge an idiom by its meaning. In fact, many idiom experts say you can judge an idiom based on five properties: meaningfulness, familiarity, literal plausibility, global decomposability, and final word predictability. In the Memory & Cognition paper “What norming reveals about idioms: Making the case for a presuppositional account,” Griffen and Noveck explain a norming study where they evaluated these properties.
Idiom norming studies come out about once every decade. Notable previous studies included Titone & Connine (1994), Libben & Titone (2008), and Blukes & Tanner (2017). Griffen and Noveck are now carrying the torch in 2025.
In their first experiment, five separate groups of participants rated 36 idioms on one of the idiomatic properties (meaningfulness, familiarity, literal plausibility, global decomposability, or final word predictability). Using different groups for each parameter prevented bias from priming or repetition. The selected idioms had the structure: [verb][article][noun] – such as she blew a fuse, he pulled the plug, she hit a wall. The authors also composed nonsense idioms as filler items. These are combinations of two semantically unrelated idioms – such as she smacked a cake, he cramped the air, she chewed the bus. Lastly, they included literal items for more fillers. These are mundane phrases that have the same structure as other items – such as she dropped the bottle, he dusted the ceiling, she got the job.
Participant ratings of idioms were highest for meaningfulness, indicating raters could attribute meaning to the expressions. Familiarity was the second highest-rated dimension, suggesting raters had experience with the items. Literal plausibility and decomposability were more moderately rated, while final word predictability had lower ratings. Each of these dimensions was rated highest for idioms, followed by literal filler items, and lastly nonsense idioms. These results largely replicated findings from the previous norming studies.
For their second experiment, Griffen and Noveck introduced a new dimension for idioms: presuppositional strength. The idea behind this dimension is that idioms can assert one thing while also presupposing something that happened before. For example, “broke the ice” asserts that someone “initiated social contact” while presupposing that “there was pre-existing social tension.” Another example is “fan the flames”, which presupposes there was tension beforehand and asserts that it got worse.
The authors used the same 36 idioms from the first experiment and created simple sentences such as “Joey walked into class and broke the ice.” As a filler condition, they created paraphrased items, which were more in line with dictionary definitions, such as “Joey walked into class and intentionally created a friendly environment.” They also made a nonsense condition, which combined two unrelated phrases, such as “Joey walked into class and drew a ball.”
The authors determined what the presupposed information would be for each idiom, such as “there was tension beforehand,” which is assumed when someone “breaks the ice.” This information is assumed to be implicitly conveyed through the idiomatic expression, but it is less likely to be conveyed through the paraphrased or nonsense alternatives. For the norming study, participants were asked to rate how likely the presupposed information could have been for each item. As predicted, these ratings were significantly higher for the idioms, suggesting that idioms have a unique ability to convey unspoken, presupposed information.
Griffen and Noveck compared the norming data on presuppositional strength to the other normed qualities from the first experiment, and they found that this parameter was highly correlated with familiarity and meaningfulness. This finding implies that the idiom’s meaning triggers the background assumptions.
This study provides novel evidence to support a presupposition-based theory of idiom comprehension, and it brings a new dimension to the figurative language realm. Griffen and Noveck hit the jackpot with this work, which is sure to blow the minds of idiom-lovers everywhere.
Featured Psychonomic Society article
Griffen, N., & Noveck, I. (2025). What norming reveals about idioms: Making the case for a presuppositional account. Memory & Cognition, https://doi.org/10.3758/s13421-025-01719-8
