In the 1950s, calling the tiny grey crustacean that curls into a ball a roly-poly was strictly a southern thing.
Forty-five years later, the word had taken over the country. No broadcaster pushed it, no textbook taught it.
What turned a pocket regionalism into a near-universal term has puzzled linguists for decades. A statistical physicist recently cracked it open using the mathematics of magnets.
Borrowing from physics
James Burridge has been chasing that question for years. A professor of probability and statistical physics at the University of Portsmouth, he usually studies atoms and bubbles rather than adverbs.
His new paper argues that human speech follows the same broad patterns at scale. Each speaker tilts slightly toward whatever variant the people around them use.
Multiply that small nudge across millions of people over decades, and language change stops looking random.
Patterns emerge, and they match something physicists have been describing in magnets for a century.
Words mapped over time
Survey data sit at the center. Burridge tested the approach against the Cambridge Online Survey of World Englishes, a large collection of American dialects assembled by linguist Bert Vaux.
Participants reported what they actually call hundreds of everyday things. The model learns its parameters straight from those answers, not from any pre-baked theory.
Plot those answers on a map and you get clean geographic blocks.
Soda dominates the Northeast and California, pop owns the Midwest, and Coke covers much of the South.
Those lines aren’t random squiggles. They obey the same kind of mathematics that governs the boundaries between regions in magnetic materials and droplets of liquid.
The roly-poly takeover
The clearest demonstration involves a creature most kids have flipped over on a sidewalk: the woodlouse.
In 1950, calling one a roly-poly was a southern thing – confined to a small pocket of the country.
By 1995, the term had eaten the map. Roly-poly spread across most of the United States in roughly two generations.
Burridge’s model can reproduce that takeover from realistic starting conditions, without anyone hand-tuning the equations.
When the local word holds its ground
Not every word loses its turf. Burridge documented the opposite case in an earlier paper on England, tracking splinter – the word for a sliver of wood lodged in skin.
Splinter pushed north from southern England and quietly took over almost the entire country. Almost. One stubborn region in the northeast refused to give up its old term.
“Splinter is used across almost all of England, except around Newcastle, where people still say spelk,” said Burridge.
“Although Newcastle itself is densely populated, it is surrounded by more sparsely populated areas, which helps the local form hold its ground and prevents splinter from taking over.”
Bubbles and dialect borders
Boundaries between regional words are far from being passive.
The model gives them surface tension – the same property that holds a soap bubble together, smoothing curves and pulling smaller pockets of speech inward.
Population density appears to warp those edges as well. The model’s inferred parameters suggest dense urban areas tug the lines toward them, while sparsely populated countryside slows the arrival of incoming words.
Earlier work had floated the idea – Burridge laid the groundwork in a 2017 study.
Until this paper, no one had tied the surface-tension picture directly to large modern survey data, with parameters learned from real dialect history rather than guessed.
A hidden bias
The novel piece arrives next. Burridge’s model contains what he calls a bias field – a hidden push that nudges certain words to grow or shrink in particular places.
That push does not last forever. It fades, with a measurable half-life. Once enough time passes, the bias decays, and predictions about which word will dominate dissolve into guesswork.
That decay places a limit on how far ahead scientists can predict language change.
Much like long-range weather forecasts, small uncertainties build over time until the original signal becomes lost in the noise.
The forces that shape our speech
For dialectologists, the model offers a way to spot when a regional change is being driven by ordinary conformity or by something genuinely unusual – a cultural force the equations cannot explain on their own.
Voice and translation technologies that tune to local speech now have a principled limit on how far ahead those adjustments can reach.
“My research suggests that language may be much more law-like than it first appears,” said Burridge
“Beneath the creativity and messiness of human speech, there may be hidden statistical forces shaping how we all end up talking.”
The study is published in the journal Physical Review E.
—–
Like what you read? Subscribe to our newsletter for engaging articles, exclusive content, and the latest updates.
Check us out on EarthSnap, a free app brought to you by Eric Ralls and Earth.com.
—–