the brittleness of children and the egos of driven men

January 18, 2019 3:51 PM

In the fall of 1938, Wendell Johnson recruited one of his clinical psychology graduate students, 22-year-old Mary Tudor, who was avid but timorous, to undertake exactly that experiment. She was to study whether telling nonstuttering children that they stuttered would make it so. Could she talk children into a speech defect? The university had an ongoing research relationship with an orphanage in Davenport, Iowa, so Johnson suggested she base her study there. And thus, on Jan. 17, 1939, Mary Tudor drove along the high, swooping bluffs overlooking the Mississippi River to the Soldiers and Sailors Orphans' Home. The study she began that morning became the subject of a multimillion-dollar lawsuit against the State of Iowa and the University of Iowa.
posted by sciatrix (26 comments total)

Yeah academic research was pretty terrible and damaging to many people before the IRB came along.
posted by Young Kullervo at 4:35 PM on January 18, 2019 [12 favorites]

I mean, yeah, I know why we instituted IRBs; I'm busily going through the cases that spurred the creation of IRBs in the first place. That's how I found this piece on this specific study.
posted by sciatrix at 5:35 PM on January 18, 2019 [4 favorites]

I’ve done behavioral research (best described by my username) under a variety of IRBs since 2005. Are they perfect? No. Are they the right thing to have in terms of oversight? Absolutely.

Will I chew large chunks out of people who are cavalier about not consenting subjects? Oh my, yes. Consent, even to be bored for science, isn’t optional. Ever.
posted by Making You Bored For Science at 6:25 PM on January 18, 2019 [24 favorites]

To paraphrase Terry Pratchett: sin is when you treat people like things.

This is gonna be a tough read huh
posted by schadenfrau at 7:11 PM on January 18, 2019 [13 favorites]

If you need an external body to tell you that experimenting on subjects who haven't consented is wrong, then you have a bad moral compass.

This is a horrific thing. Thank you for sharing, sciatrix.
posted by twilightlost at 7:13 PM on January 18, 2019 [2 favorites]

I mean, yeah, I know why we instituted IRBs; I'm busily going through the cases that spurred the creation of IRBs in the first place. That's how I found this piece on this specific study.

Sorry, wasn’t trying to be dismissive of your post with my comment. It is very important to share and discuss these studies. They just hurt my soul to revisit.
posted by Young Kullervo at 7:24 PM on January 18, 2019 [2 favorites]

Jesus. I don't think they've ever brought this up in the alumni magazine...
posted by fifteen schnitzengruben is my limit at 7:25 PM on January 18, 2019 [1 favorite]

Consent, even to be bored for science, isn’t optional. Ever.

It turns out, though, that even to IRBs consent is optional if you're only hurting people a little. Thus, the IRB-approved study where people spammed something like half a million unconsenting facebook accounts with "negative posts," just to see what it looked like when you hurt people a little bit.
posted by GCU Sweet and Full of Grace at 7:31 PM on January 18, 2019 [6 favorites]

Sure but the benefits of the study to the participant or someone else usually have to outweigh the pain if it’s more than minimal. And the researchers have to prove that. And typically, in non-medical studies minimal, that pain is usually nothing more than theorized slight discomfort, nothing on par with making perfectly healthy children mute to the point that it damages their entire lives.

I could be wrong though, it’s been a while since I did human subjects research.
posted by Young Kullervo at 7:36 PM on January 18, 2019 [1 favorite]

Yeah I knew someone who quit Facebook because she was pretty sure she was one of those lucky guinea pigs. Or rather: she quit because it was making her really depressed, like noticeably negative posts all the time, and she only figured it was part of the experiment when the news broke. She was pretty upset about it. Like, they did something to see if it would hurt her, and it did. And then they gaslighted her about it by pretending it was normal.

She said she felt pretty violated.

And, if you think about it, it’s almost worse if she wasn’t part of the experimental group.
posted by schadenfrau at 7:37 PM on January 18, 2019 [18 favorites]

I am curious about this study now.
posted by Young Kullervo at 7:38 PM on January 18, 2019

I'm pretty sure people are talking about this study from 2014. Which was done by FB itself. No outside researchers. No IRB. If there is some other one that was done by a research institution with IRB coverage, I would love to hear about it.
posted by PMdixon at 7:55 PM on January 18, 2019 [8 favorites]

I was irate about the Facebook study on emotional contagion when it came out for precisely this reason.

In terms of consent, the argument made by Facebook was that the user agreement was sufficient, and Cornell's IRB, who was the theoretically responsible party, bought it (see the Editorial Expression of Concern). Quite bluntly, that's bullshit, but that's where it stands. I maintain (and I have this fight with friends in engineering and more applied disciplines regularly) that the Common Rule should apply really very broadly, because it's a matter of respect for persons, to use the language of the Helsinki Declaration.

On top of all of that - and I was pretty disgusted with all the authors involved, because they'd all had graduate training in psychology and had damn well been trained better - the effect they report is miniscule. Here's the relevant excerpt from their results section: "When positive posts were reduced in the News Feed, the percentage of positive words in people’s status updates decreased by B = −0.1% compared with control [t(310,044) = −5.63, P < 0.001, Cohen’s d = 0.02], whereas the percentage of words that were negative increased by B = 0.04% (t = 2.71, P = 0.007, d = 0.001). Conversely, when negative posts were reduced, the percent of words that were negative decreased by B = −0.07% [t(310,541) = −5.51, P < 0.001, d = 0.02] and the percentage of words that were positive, conversely, increased by B = 0.06% (t = 2.19, P < 0.003, d = 0.008)." [I've added the bolding for emphasis].

The key detail? While the effects are significant based on the tests they ran, the effect size is so tiny that there are flatulent mice that are more meaningful.

So, not only did they shamelessly abuse the trust of their users/subjects, they didn't find anything of real meaning.
posted by Making You Bored For Science at 8:01 PM on January 18, 2019 [20 favorites]

Has this study appeared in a tv show or book? I know the study but don't know from where.
posted by any portmanteau in a storm at 8:05 PM on January 18, 2019

That was a great article. Thank you for this post, sciatrix.

Johnson did harm the kids in that study, and it was in form very like studies done under the Nazis in terms of its lack of informed consent or consent of any kind, really. But it was not consciously sadistic, though it does have overtones of cases in which a child who is sexually abused goes on to become an abuser themself. Because Johnson arranged for kids in group IIA to be told exactly what he was told as a child: that they were beginning to stutter.

But by far the greatest harm he did came from suppressing the study, because none of those kids went on to develop a stutter!

And Johnson devoted the rest of his distinguished career to insisting that stuttering was always learned behavior, and was caused by telling children in one way or another that they were stutterers, even though he should have known, and possibly down deep did know, that this was false.

Imagine the anguish of thousands upon thousands of stuttering children who were dragged through years of grim, agonized, and self, parent, or teacher -blaming therapy which was completely futile and hopeless because it was based on premises its architect knew or should have known were false!
posted by jamjam at 8:22 PM on January 18, 2019 [18 favorites]

But it was not consciously sadistic, though it does have overtones of cases in which a child who is sexually abused goes on to become an abuser themself. Because Johnson arranged for kids in group IIA to be told exactly what he was told as a child: that they were beginning to stutter.

Yeah, and that was one of the things that really interested me: the stuttering in and of itself doesn't seem to be the thing that causes all these anxiety and fear and self-esteem things, it's the thing where the stuttering is treated as something the kid is just bad at, an inherent defect. You don't need to be a stutterer to pick up these side-effects that were associated with the stutter itself--as witness, say, the stereotypes that stuttering is in and of itself something that makes people shy or nervous. I've known a number of people with stutters, including adult stutterers who didn't "grow out of it" and people who have tried differing kinds of therapy to deal with the stutter (including one person who spoke to me about just how exhausting the techniques to avoid stuttering are in a social context). And as best I can tell, it's the constant social awkwardness and irritation that wears on people, not the actual stutter itself.

I see strong parallels to experiences of autism and other forms of neurodiversity here, although they're not identical things. Still--the study for me raises the question: how much of the comorbid depressions and anxieties and trauma symptoms that are associated with a number of neurodiverse experiences comes down to that experience of being told "you have this problem, the way you communicate is bad, and you have to work hard to compensate." I definitely grew up with that experience, often very explicitly, and it absolutely will fuck with you.
posted by sciatrix at 8:56 PM on January 18, 2019 [15 favorites]

I stutter occasionally (also slight lisp) and I’m also autistic and I’ve wondered how connected those things are. But honestly my stutter is occasional and I wonder how much of that was because I had really good parents who never focused on it the way other parents do. Thanks for this article Sciatrix, but oof is it painful.
posted by Homo neanderthalensis at 9:00 PM on January 18, 2019 [2 favorites]

Soo... what happened to the lawsuit?

FYI - the speech pathology building at the U of I is named after Wendell Johnson.
posted by Big Al 8000 at 9:16 PM on January 18, 2019

Four years after that article was written, the State of Iowa settled with the affected children for $925,000. The plaintiffs had originally sought $13.5 million.
posted by sciatrix at 9:22 PM on January 18, 2019 [3 favorites]

(I googled IRB and it means
Institutional Review Board, in case I’m not the only person who didn’t know that.)
posted by bendy at 9:36 PM on January 18, 2019 [8 favorites]

It seems highly unlikely that you can make a stutterer.

Some college friends and I took a speech pathology course together, back in the '80s, which did precisely this. My girlfriend had taken the class and warned me about what was coming up, but I couldn't bring myself to fully believe her. But the prof announced that we would next study stuttering, and that he was going to turn us all into Stage I stutterers. We poo-poohed this. By the end of his next lecture, he had proved himself. He wasn't trying to make us stutter, but he had made us all hyper-aware of our speech and intended speech. For three weeks, many of our attempted utterances ended with, "D-d-d-dammit, Ch-Chuck!" in reference to the professor. When we finally got over it, we didn't talk about it much, fearing relapse.

After that experience, I should have been more sensitive to stutterers. But I worked with a man, just a few years later, who stuttered. I once finished a sentence for him. I still cringe in shame over that, and that I was too embarrassed to apologize to him.
posted by bryon at 12:23 AM on January 19, 2019 [5 favorites]

I'm pretty sure people are talking about this study from 2014. Which was done by FB itself. No outside researchers. No IRB.

That's sort of true, but two of the three researchers were actually academics at Cornell, and an IRB was involved in making the determination that IRB oversight was not required. According to the "editorial expression of concern" now published along with the original paper in PNAS, the authors stated to the editors upon submission that “Because this experiment was conducted by Facebook, Inc. for internal purposes, the Cornell University IRB [Institutional Review Board] determined that the project did not fall under Cornell’s Human Research Protection Program.”

To me it's pretty clear that this study should have been overseen by the Cornell IRB, and that the authors were either unclear in describing the nature of the study when reporting to the IRB, or the IRB failed in its duty to exert proper oversight (or both). A significant problem is that while academic research is, and should be, subject to oversight from ethics review boards, after a long and unfortunate history of cases like the one in the FPP, identical research studies undertaken for "business practices" by private companies are not. The Facebook study basically fell into a loophole, in which a study which clearly violated informed consent was deemed outside the purview of ethics oversight because it was done for business purposes rather than science.
posted by biogeo at 7:31 AM on January 19, 2019 [10 favorites]

I'm with biogeo here: Cornell's IRB failed pretty miserably here. As did PNAS (Proceedings of the National Academy of Sciences). I'd argue that the failure to properly consent subjects under the Common Rule, which very much does not recognize clickthrough consent the way Facebook argued, should have been grounds for full retraction of the paper.
posted by Making You Bored For Science at 8:01 AM on January 19, 2019 [7 favorites]

I didn't mean to get hung up on the facebook experiment specifically, just that IRBs have been consistently dropping the ball (IMHO) WRT large-scale field experiments. Other examples include sending unconsenting experimental subjects different kinds of election-related mailers or sending them canvassers with experimental-treatment talking points. In these cases, IRBs seem to typically just say "Any potential harm is negligible" and sign off on using unconsenting subjects.

The problem from my point of view is that the typically large scale of these experiments means that they really need to be worrying, but aren't, about events that are individually very unlikely. Any potential harm would indeed be negligible to a person who is mentally and psychologically healthy, but if you're doing something to 100,000 people you're also very probably going to performing this intervention on someone who is already deeply delusional, someone who has currently untreated paranoid schizophrenia, someone who is deep in a manic episode, someone who is in a state of perisuicidal despair, and so on. If you're doing something to 100,000 people, you need to consider and worry what a literal one in a million negative reaction to the treatment might be, because you have just shy of a ten percent chance of provoking that. But... nope. Whatever seems negligible to the upper-middle-class high-achievers on the IRB is cool.
posted by GCU Sweet and Full of Grace at 8:10 AM on January 19, 2019 [3 favorites]

Because Johnson arranged for kids in group IIA to be told exactly what he was told as a child

not exactly the same. not if he was told the truth about his speech, however cruelly and with whatever abusive judgments and bad treatments that went along with it. these children got that same abuse, but in addition, they were also told lies. they were successfully taught to disregard what they could feel and hear, and substitute what they were told by authorities to believe. I mean: they weren't just made to feel defective for possessing certain qualities, they were made to lose touch with objective reality. that is two types of abuse, not just one.

for that reason it is really frustrating that there wasn't follow-up, even in the article, with the other experimental group, IA, the stuttering children who were lied to and told they "now spoke fine." especially the ones who got worse instead of getting better or staying the same. because they didn't get the stigma (from the study; I'm sure they did at other times from other people) but they did get the lies. so did they, also, experience the denial of reality by authority figures as psychological abuse? did they know they were being lied to at the time, or figure it out later? did they hope it was really true, did they believe it completely, or did they just play along?
posted by queenofbithynia at 8:48 AM on January 19, 2019 [5 favorites]

The Alliance for Human Research Protection has a page on the study, which includes an article from the San Jose Mercury News (sounds like the article that prompted the lawsuit, and which seems based on the idea that the experiment did prove Johnson's theory). The link to the Mercury News no longer works, but part of it is quoted as:

Theory improved treatment and understanding of stuttering
BY JIM DYER Mercury News
[...] According to Mary Tudor’s records, the last time anyone contacted the orphans in connection with the experiment was August 1940, when Tudor made her last visit to the orphanage. Six decades later, a search that began with her records identified 20 of the 22 orphans, of whom at least 13 are still alive.

They had never heard about the experiment. When the Mercury News told them, most became angered by the experiment. Some responded stoically. Others cried.

“Oh, dear God,” said Donna Hughes Collings of Des Moines, who had been a normal speaker in a control group and therefore suffered no damage. Her husband held her while he lambasted the people responsible, comparing them to the Nazis saying, “The end never justifies the means.”

Others just stared, incredulous. “Why? Why would they do that to us?” asked Ralph Fry of Nora Springs, Iowa. He had been a stutterer, was placed in a control group and given positive therapy, and his speech improved.

Some were not surprised.

“We knew they were experimenting on us,” said Hazel Potter Dornbush, 77, of Fulton, Ill. At 15 years of age, Hazel, a normal speaker induced to stutter, had been one of the oldest subjects. “Every week somebody else from the university would come and start testing us for God knows what.”

Others, like Robert Hamer, 73, of Waterloo, Iowa, listened long and hard before commenting. He was a stutterer included in a control group and retained somewhat halting speech. “They might not have known the negative effects beforehand,” he said. “If they knew, then it was wrong.”

Jane Ann Pugh Fleming, a normal speaker who had been in a control group and now lives in Milwaukee, at first refused to listen to anything about the experiment. “I don’t even want to know,” she said, shaking her hand in front of her face and ushering the reporter out of her home.

For her younger sister, Norma Jean Pugh, learning about the experiment came as an epiphany. At 6, she had spoken fluently, but was induced to stutter in the experiment. She suffered for years, wondering why she found it so hard t o be with people.

“At least I know it’s not me,” she said. Now going by the name Kathryn Meacham, she moved from foster family to foster family during her childhood, repeatedly rejected because they considered her a misfit and slow in school. She remembers how classmates teased her about her stuttering.

“My speech bothered me as a young child,” she said. “The kids were cruel.”

Now 68, she lives as a recluse in her tiny town of Linden, Iowa, the one who never attends high school reunions, who never leaves her home, who rarely talks to anyone except her children.

When Mary Korlaske Nixon was told about the experiment, she was stunned. She stared, her smoky blue eyes fixed on every word.

Of the six normal speakers induced to stutter, she retained the most noticeable speech repetitions. She has difficulty with words that begin with “s” and sometimes repeats words in her sentences. When she is nervous, her words jumble and she struggles to get them out.

“It’s affected me right now,” she said. “I don’t like to read out loud because I’m afraid of making a mistake. I don’t like talking to people because of saying the wrong word.”

The orphanage “was a cruel place, but I didn’t realize they was pulling that on me.”

She remembered many of the children in the experiment, including her friend Dorothy, who died last year, and Marian Higdon, who she learned had been her control group counterpart.

“I don’t like what they did to me, but I’m glad it was me,” she said. “Marian might not have been able to handle it. I’m a fighter. I’ll make them pay for what they did to us.”

posted by lazuli at 11:29 AM on January 19, 2019 [4 favorites]