The “horse” of the title of The Horse That Won’t Go Away is the famous (well, famous in some circles anyway) “Clever Hans,” and the sense in which he “won’t go away” is that the fallacy that has come to be associated with him (the “Clever Hans phenomenon”) continues to lead people into unjustified beliefs.
The Horse That Won’t Go Away is a critical thinking book. It is just over a hundred pages, not to mention it’s of smaller than average dimensions with bigger than average print. So it reads like an extended article in Skeptical Inquirer—a magazine I’ve subscribed to and enjoyed for decades now.
So what is the Clever Hans phenomenon, and what does it have to do with the horse Clever Hans?
The Clever Hans phenomenon occurs when information is transmitted through subtle, unintentional signaling or cuing, and then people who aren’t aware it was transmitted in this way (including potentially the signaler and/or signalee) mistakenly infer some other explanation or interpretation of what is happening.
In Germany in the early 1900s, Wilhelm von Osten set out to prove that horses were intelligent animals and could be educated not unlike people. After working diligently for a long period of time with his primary equine student Clever Hans, he announced that he had achieved a breakthrough.
To the delight and astonishment of local audiences—of regular folks, and increasingly of academics, scientists, and various dignitaries—von Osten would pose questions, generally arithmetical, to Clever Hans, and Clever Hans would consistently respond with the correct answer.
Now of course Clever Hans couldn’t speak his responses—this was decades before Mr. Ed came on the scene—so instead he was taught to tap them out with his hoof. He would be asked what 15 minus 11 is, and he would methodically tap his hoof one, two, three, four times, and then stop.
Over time Clever Hans was able to solve more and more difficult math problems. Certainly there were people who suspected some kind of trickery, but their number diminished the longer no such trickery could be identified. Evidence in favor of the notion that this was something legitimate and not some kind of ruse is that von Osten manifested zero interest in making money from his wonderful horse, for instance by charging people to attend his exhibitions.
It took a while to figure out what was really going on, and von Osten resisted the truth as long as he plausibly could (then reluctantly agreed, but subsequently reversed himself and went back to believing in Clever Hans). Clever Hans, it turns out, wasn’t responding to what was being said to him at all. Instead he was merely watching the people present, mostly von Osten himself, and responding to how they reacted to his hoof tapping.
When he would make his last tap that would constitute the correct answer—e.g., when he would make his fourth tap in response to the 15 minus 11 question—von Osten and the people in the audience would typically stop staring at his hoof and look up. He learned that if he stopped when they looked up like that, he would be praised and given some sort of treat.
The reason it was so hard to find the deception is that there wasn’t any. Neither von Osten nor anyone else present was intentionally altering their body language in order to cue Clever Hans and trick people into thinking the horse could do math. Everyone was quite sincere. The cuing was completely unintentional and involuntary.
So Clever Hans was indeed pretty clever, but not in a way that people weren’t used to animals being clever (understanding human language and being able to solve arithmetic problems), but rather in a way that I think we’re mostly already aware animals can be clever (responding to body language cues and such to give people what they want in order to be rewarded).
By the way, this is very much the kind of skill you teach your kids when you “help” them with their homework and think you’re teaching them the actual material itself. The way you—intentionally or unintentionally—cue them with little hints, changes in facial expression, changes in tone of voice, leading questions like a sing-song “Are you suuuuuuuure?,” etc. tells them if they’re right or wrong, hot or cold, so they then can adjust their guessing accordingly. You may think you’re teaching them the alphabet, or fractions, or whatever, but what you’re really teaching them is to read your body language and understand that when you lean in a certain way it’s safe for them to stop guessing and to let their last effort count as their final answer.
I also think this nicely illustrates how often what we are convinced we perceived or directly experienced is really an interpretation. People insist they’ve seen a ghost, experienced being cured by something from alternative medicine, felt God’s love through some mystical experience or whatever, and in their mind these beliefs can’t possibly require some sort of argument or justification because they experienced them directly. If someone questions them they can only interpret that as calling them a liar (or perhaps calling them crazy). At most they might concede that other people, who haven’t had the relevant experiences themselves, might need other evidence and arguments to come to a sound conclusion, but surely they and the people like them who have seen something with their own eyes or experienced something directly already know the truth.
It’s not that they think they’ve made a justified inference; they’re convinced there was no inference at all, that they perceived or experienced something directly.
So say someone’s arthritis is particular painful one day, they hang a magic magnet on the part of their body that’s most painful, and they wake up the next day and it feels noticeably better. Often they won’t think of this as “I put the magnet on and later felt better. From this I inferred that the magnet was the most likely cause of my feeling better. So I conclude that magnets can probably help people with arthritis.” They think of it as “I know that magnets help people with arthritis because I myself have experienced my arthritis pain being reduced by wearing a magnet.” In their mind, there’s no inference, there’s no causal hypothesis, they’re not taking an educated guess based on the evidence of what they experienced—they directly experienced the magnet’s effect on their arthritis pain. So there’s nothing to doubt, nothing to question; they know what happened because it happened to them.
The people at Clever Hans’s demonstrations walked away convinced that they had seen Clever Hans solve arithmetic problems. But they had seen nothing of the kind. That he was solving arithmetic problems was merely one possible interpretation of what they had seen, perhaps even a justified one relative to the evidence available to them. What they had actually seen (and heard), though, was von Osten saying something like “Now, Hans, what is fifteen minus eleven?,” Clever Hans tapping four times with his hoof and stopping, and von Osten smiling and reaching up to pat his horse and feed him a treat, while the audience oohed and aahed in appreciation.
One possible explanation for what they saw is that Clever Hans understood what was said to him, quickly did the math in his head and came up with the answer 4, and understood how he was supposed to communicate this answer and so tapped his hoof four times. But another possible explanation—and the one that ultimately proved more consistent with all the available evidence—is that Clever Hans had no clue what the words said to him meant, but having learned in the past that he was consistently rewarded in situations like this if he started tapping his hoof and then stopped when he saw von Osten raise his head slightly, did that and thus tapped four times. The audience didn’t directly observe either of these explanations; the potential explanations had to be assessed and evaluated by discussing the evidence, altering and rerunning the experiment in different, relevant ways, questioning von Osten and the people involved, considering other cases it might or might not be analogous to, and on and on.
The moral of the story is, “I don’t need to justify what I believe to you or anybody else, because it actually happened to me [or “I saw it with my own eyes”] so I know it’s true!” is rarely as definitive as people routinely take it to be.
Anyway, the Clever Hans phenomenon, and the related errors in belief acquisition that people commit, is not just something from the past that has since been corrected, not just some interesting story about a horse and human psychology to read about. The rest of The Horse That Won’t Go Away illustrates this with more recent examples of people still falling prey to the Clever Hans phenomenon.
The example that the authors go into to the greatest extent by far is “facilitated communication (FC).”
FC was very big for a few years in the disability community. People claimed they had discovered a revolutionary way to enable people with certain disabilities—most often severe autism—to communicate.
The theory was that such folks have just as sound minds as anyone, and the same desire to communicate as anyone, but they are trapped in a body that does not allow them to translate those desires into practice. They want to be able to say, “I’m cold,” “I love you, Mommy,” “Gimme a ham sandwich,” or “15 minus 11 is 4,” but they are unable to work their vocal chords so as to speak, unable to work their hands so as to use sign language, hold a pencil, type on a keyboard, etc., so they live a horribly frustrating life of non-communication.
But with FC, a “facilitator” holds the hand of such a person over a keyboard, together they touch certain keys, and sometimes coherent messages result. The proponents of this methodology were sure that while the human touch of the facilitator was somehow helping or enabling things, the autistic person was actually doing the typing; it was his or her words, his or her thoughts, being typed onto the page.
Frankly the whole thing looked bogus from the beginning—far, far less convincing than Clever Hans—but the power of wishful thinking kicked in and hopeful parents and others convinced themselves that the autistics’ power to communicate had miraculously been unlocked. But, I mean, come on. It should have been a clue that typically while the typing was occurring the facilitator was looking at the keyboard and the autistic person was gazing off into space.
Just as in the case of Clever Hans, people were less on guard for some kind of trickery because the people involved seemed completely sincere. And indeed most of them were. Eventually some people figured out how to make plenty of money selling FC as a service, training facilitators, etc., but the vast majority of folks involved weren’t trying to fool anyone. They were just trying to help people in need, and convinced that they were doing so.
But as was pretty obvious all along, even the facilitators who were totally sincere were actually unconsciously cuing the autistic people. They’d move their hand in some subtle way, squeeze just a bit without meaning to, whatever, and the autistic person—again probably not even consciously—would realize this was the signal to press down on the key that was directly below their typing finger at that time, and eventually some sort of message would result.
For a time FC seemed like a positive thing. Its believers thought this was the greatest advance in autistic treatment in history. Autistic children could now be placed in classrooms, and could take notes and answer questions on tests and such through the use of FC. Even most of those who had developed some doubts that FC really constituted the autistic people themselves communicating figured “What harm can it do? Look how happy it’s making their parents and everyone to believe they’re finally communicating with these people they love!”
That was when the messages ranged from neutral and innocuous to positive and heart-warming. That was when the messages were the kind of thing parents had always desperately longed to hear. That was when there were a lot of “I love you Mommy and Daddy. I know how incredibly difficult this has been for you, and I really appreciate all the sacrifices you’ve made for me”-type messages.
But then things took a much darker turn. All of a sudden, along with those kinds of messages, increasingly often there were messages like “For the last few years, Daddy has been coming into my room at night and sticking his finger in my anus.” The tears of joy now sometimes became tears of outrage. There were accusations, impassioned pleas of innocence, fights, divorces, forcible removal of children from homes deemed unsafe, and criminal charges.
Eventually sanity more or less prevailed, proper experiments debunked FC, and most people realized there had never been anything to it.
But it never went away entirely. The number of its proponents in the disabled community decreased, but, like the “immunization causes autism” folks, there has always remained a sizable bloc of them. Indeed, the authors report that, if anything, FC has been making a modest comeback lately. That horse just won’t go away.
This is a modest book, but it does a very nice job. It is well-written, logically organized, clear and understandable, entertaining, and educates people about a phenomenon that continues to adversely affect our thinking and to sometimes have tragic real world consequences. If you enjoy books on critical thinking and the debunking of pseudoscience and such, The Horse That Won’t Go Away is a worthwhile read.