Why can baboons tell written words from gibberish? The eyes — and brain — have it

Charles Darwin would surely have been mesmerized by a paper released last week showing that baboons can recognize written words and distinguish them from gibberish. This was more than a feat of memorization, since the baboons were able to do this even if they'd never seen the words or nonwords before.

In a paper describing their findings, the scientists say perhaps the baboons are able to do some sort of unconscious statistical calculation involving the combinations of letters most likely to form words. "We tend to think that the ability to distinguish what's a real word is fundamentally human," said Duke University neurobiologist Michael Platt, who wrote a commentary accompanying the paper in the journal Science.

Before Darwin went public with his theory of natural selection, he wrote notes and letters pondering the implications of his idea for the minds of humans and other animals. Darwin was particularly interested in philosophical questions about the relative roles of instinct, culture, and reason in shaping our behavior.

"He who understands baboon would do more toward metaphysics than Locke," he wrote in 1838. The quote inspired the title of the 2007 book Baboon Metaphysics by Penn professors Dorothy Cheney and Robert Seyfarth.

If only Darwin could have seen what's been learned. Motivated by treats, six baboons learned to distinguish words such as DONE, LAND, THEM, and VAST from nonwords such as DRAN, LONS, TELK, and VIRT. How they do this without being able to read is a bit weird — like the freakish talents of some savants — but evolutionary neuroscientist Mark Changizi was not at all surprised. "I would be surprised if they couldn't do this," he said.

To Changizi, who works for 2AI Labs in Boise, Idaho, the experiment confirms an idea he described in his 2011 book Harnessed. There, he argued that the human brain didn't evolve to master language, but language — and music — evolved to harness abilities that already existed in the brain. Some of those abilities go back a long way and we share them with other primates, and perhaps nonprimates as well.

"The seemingly unique human powers we have, like speech and reading and music, are a redesign of powers that other primates probably possess," he said.

In some ways this seems obvious — how could we use any other systems of communication except those that rely on our evolved senses? What's interesting, however, is how much of the framework for our language and musical abilities existed long before our ancestors became human. Baboons branched off from our species 30 million years ago.

Reading, Changizi said, has many signature properties of instinctive behavior. People are good at it, and children can learn it without too much effort. But it is really not a reading instinct, but an intertwining of cultural and biological evolution.

It's easy to take vision for granted — we open our eyes and see things — but complicated computation is needed to transform electrical pulses moving down our optic nerves into objects, distances, and motion. Much of our vision goes on unconsciously — our brains are constantly calculating contours and angles, Changizi explains in his book.

We're lucky we don't have to run our own brains.

Something similar seems to be going on in the brains of baboons. Thanks to the abilities of their brains to perform sophisticated statistical calculations, the experimental subjects treated the task as a no-brainer. They signaled their choices by tapping a touch screen, performing without pause or apparent effort.

The work might shed some light on the roots of reading disorders, such as dyslexia, said Duke's Platt. Some reading problems may have to do with pattern recognition rather than language, for example. The insights from this experiment could also rechannel investigations into the question that excited Darwin's curiosity — how do animal minds work and what can they tell us about our own?

When we read, we're tapping into a 30-million-year-old device — one we share with monkeys, said Platt. But why, then, didn't monkeys or apes use this ability to develop language? "What makes us so different is not at this moment clear."

Contact Faye Flam at 215-854-4977, fflam@phillynews.com, or @fayeflam on Twitter. Read her blog at philly.com/evolution.