A pre-K teacher in Oklahoma is making news this week after forcing a left-handed 4-year-old boy to write with his right hand.

The boy was sent home from school with an article discussing left- and right-handedness. The article mentions historic attitudes toward left-handedness that associate it with evil and the devil. It's written carelessly enough that it isn't clear whether the writer believes left-handedness is still seen as evil or whether that was only the case in the past.

Regardless, it's surprising that today, in 2015, a teacher would try to force a child to write with his non-dominant hand. Roughly 10 percent of people are left-handed, according Chris McManus, a University College of London researcher who wrote a book chapter on the history and geography of left-handedness.

Scientists generally agree that there is a strong genetic component that determines whether a person is right- or left-handed. But good data on the prevalence of left-handedness has been difficult to find, McManus writes.

One of the best available data sets on left-handedness comes from a scratch-and-sniff survey of olfactory ability mailed out to millions of National Geographic subscribers in the 1980s.

Go ahead and read that sentence again — it doesn't get any less weird the second time around.

In 1986, National Geographic published a special issue on smell. As McManus recounts it, the issue "was accompanied by a 'scratch and sniff' card, which readers were encouraged to scratch, report what, if anything, they could smell, and then, after completing a brief demographic questionnaire, return the card."

The researchers who created the survey, from the Monell Chemical Senses Center in Philadelphia, thought that smell and handedness might be linked. So they included questions about the dominant hand used for writing and throwing things in their demographic section.

So, 1.4 million people scratched, sniffed and mailed back their cards. As it turned out there was no link between smell and left-handedness, but the resulting 1.4 million datapoints created the largest dataset of hand preference ever constructed.

As an opt-in survey of magazine readers, it has obvious weaknesses. Ninety-seven percent of respondents were white, and National Geographic readers in the 1980s were almost certainly economically and demographically distinct from the general population in other ways, too.

Still, the sheer size of the database gave researchers a glimpse into the geography and history of left-handedness that they'd never had before. For starters, they show that rates of left-handedness fell during the late Victorian era, reaching their nadir in the early 1900s. They then rebounded  steadily until about the 1950s or so, when they flattened out.


 

McManus theorizes that the late-Victorian dive in left-handedness reflects stigma against southpaws, of the kind that might have been on display in that elementary school in Oklahoma. He attributes a lot of that to the Industrial Revolution. "Left-handers may also have appeared less capable and more clumsy, as left-handed adults worked on machines that were almost certainly designed with right-handers in mind, and left-handed children were taught to write with steel dip pens that needed to be dragged across the paper from left to right by right-handers, and were not capable of being pushed across by the left hand without digging into the paper and making blots and stains."

The National Geographic data also show significant variation in rates of left-handedness by U.S. state, with higher rates in the Northeast and lower rates in the middle of the country.


Again, these differences may largely be explained by genetics. If certain families in certain areas are more likely to pass on left-handedness traits than others, you'd expect to see higher rates of left-handedness in those areas.

Other than the occasional joke or stupid comment, you might think that Victorian-era attitudes toward left-handedness are a relic of the past. But the Oklahoma case makes clear that these attitudes survive, at least on some level.