Let’s Put Infinite Monkeys To The Test

Every so often, the world of science will exceed my wildest dreams of ridiculosity.  About a decade ago, some extremely lucky and/or ballsy researchers set out to test the Infinite Monkey concept experimentally — by giving some actual live monkeys an actual working computer to type on.  I shit you not.

Back in 2003, folks from the University of Plymouth installed a computer terminal in a monkey enclosure at the Paignton Zoo in Britain with a group of six macaque monkeys, giving the monkeys a good solid month or so to bang out some spec scripts.  The entire setup was monitored so researchers could watch the Infinite Monkey idea come to life before them.  Check out the dramatic documentary of the project, courtesy of Geoff Cox (who spearheaded this effort):

Dramatic, eh? If for some reason this embedded link to the youtube video disappears, it’s also over at Vimeo.  This is science at its best!

"I can't believe these humans got a grant for this. Well, time to pee on the keyboard..."

So what happened?  What did the monkeys produce?  They produced my single favorite scientific result of all time, bar none — they mostly took big steaming dumps on the computer.  Fantastic!  From the press reports at the time:

After a month, the Sulawesi crested macaques had only succeeded in partially destroying the machine, using it as a lavatory, and mostly typing the letter “s”.

Gentlemen, start preparing next year’s grant application!  We’re going to induce a herd of emus to vomit on a Skype terminal!

If you’re interested, Chris over on blogspot has posted the entire text on his blog, where you can indeed see that the monkeys mostly just hit one key over and over again, when they weren’t depositing turdage on the keyboard.  They produced what looks like about 5 pages of text, about 4 of which are the letter “S” repeated.  Nevertheless, a quick scan of the text does reveal a couple English words — I see the words “mass”, “sass”, and the girl’s name “Ava”.  The completists out there will want to purchase the book that was produced, although it appears to be out of print for some reason.  The book is misleadingly named “Notes Toward the Complete Works of Shakespeare“, so here’s hoping some Shakespeare aficionado buys this book only to discover 4 or 5 pages of random gibberish.

And yes, I know there are various in silico attempts at this, where someone has launched a computer program to generate random text.   Jesse Anderson’s Million Monkeys blog documents the progress of his simulator, perhaps the most famous random-text generating computer program to set sights on Shakespeare.  The famous scientist Richard Dawkins dabbled in this area, albeit with a practical application in mind (exploring theories of evolution).  And over at pixelmonkeys.org, the idea has moved beyond text into the realm of images — at this site you can ask their simulator to produce random images, which is a cool idea until you’ve rendered the 32nd image that looks like static.  But to my mind, simulations are too easy — I’m far more impressed with the real-world implementation, especially one that may involve one part of the experiment shitting on another part.

A highlight of this story, for me, is the reaction of the Zoo’s scientific officer, Dr. Amy Plowman, who said (and I quote)  “The work was interesting but had little scientific value, except to show that the ‘infinite monkey’ theory is flawed.”  Theory?  Really?  It’s the theory that is flawed?  Of course, the simple rebuttal to this is that you didn’t let them type for a long enough time — after all, this “theory” does have the word “Infinite” in the title.  Expecting the monkeys to crap out Shakespeare (pun delightfully intended) after just a month is wildly optimistic.  In fact a real random-text generation spitting out random characters far faster than a pack of monkeys could manage, would still need far longer than a month to get anywhere near random generation of Shakespeare.  I also enjoy that she calls it a “theory” (probably confusing it with the terminology Infinite Monkey Theorem) — as if this were a behavioral ecology theory worked up by a field researcher in the Amazon rainforest.  “If only there were some way to test my theory, that if monkeys were presented with a working computer terminal, they would crap on it continually for a month…”

Geoff Cox, head of the project, is more realistic:

“It was a hopeless failure in terms of science but that’s not really the point,” said Geoff Cox ,of Plymouth University’s MediaLab, who designed the test. So what were the academics trying to achieve? “It wasn’t actually an experiment as such, it was more like a little performance,” said Mr Cox.  “The   monkeys aren’t reducible to a random process. They get bored and they shit on the keyboard rather than type.”

Why didn’t I get into this line of work?


© 2011 TimeBlimp Thith ith a pithy statement. Suffusion theme by Sayontan Sinha