In 2003 an email went round saying research at Cambridge University showed that only the first and last letters of a word matter. There was no such study at Cambridge. And when researchers did test jumbled words, people read them slower every time.

Try two jumbles

Both keep the first and last letters. The second is much harder, mostly because of “nsepwpaer” (newspaper) and “coowssrrd” (crossword), where letters land far from home. So the rule in the email only works for some jumbles.

Where the email came from

Matt Davis, a researcher who actually works in Cambridge, traced the idea back to a 1976 PhD thesis by Graham Rawlinson at the University of Nottingham. He also points out why the email’s own example felt so easy: short words were left alone, swapped letters stayed close together, no jumble made a new real word, and the sentence was easy to guess (Davis, MRC Cognition and Brain Sciences Unit). In other words, the example was built to be readable.

What the real study found

In 2006, Keith Rayner and colleagues tracked readers’ eyes while they read normal and jumbled sentences. Their paper title says it all: there is a cost.

Letters moved	Reading speed	Slower than normal
None	255 words a minute	none
Inside the word	227 words a minute	about 11%
At the end	189 words a minute	about 26%
At the start	163 words a minute	about 36%

So you can read jumbled words, but you always pay something, and the first letters matter most. A 2015 study in Spanish found the same pattern, and a much bigger cost for people reading braille.

What nobody knows yet

Rayner’s readers saw pairs of letters swapped. Our Jumble script shuffles all the middle letters of a word, which is probably harder, especially in long words. We don’t know of a study that measured exactly that.

What our readers show

Not enough runs yet. Once the Jumble script has at least 20 verified runs from at least 10 different readers, we will add how much slower our readers are with it than with normal text. Our readers choose to come, so these will be nnLabs readers so far, not a scientific sample.

See it for yourself

Try one jumbled passage and see your speed. No sign up, and nothing about you is saved. Try one passage

More on flipped and mirrored text: why upside down text takes about three times as long.

There is also a free Monthly Challenge for adults 18+, with US$100 for the live round winner.

Sources

• Davis, M. The Cmabrigde email. MRC Cognition and Brain Sciences Unit. mrc-cbu.cam.ac.uk
• Rayner, K., White, S.J., Johnson, R.L. and Liversedge, S.P. (2006). Raeding wrods with jubmled lettres: there is a cost. Psychological Science 17(3). doi.org/10.1111/j.1467-9280.2006.01684.x
• Perea, M., Jiménez, M., Martín-Suesta, M. and Gómez, P. (2015). Psychonomic Bulletin and Review 22. doi.org/10.3758/s13423-014-0680-8