“I tell myself I shouldn’t be feeling the way that I’m feeling.”
That is an item on a questionnaire, rated from never or very rarely true to very often or always true. Psychologists have been handing it to people for years to measure whether a feeling is allowed to be a feeling, or whether it arrives and is immediately overruled.
A team of four psychologists, Brett Ford, Phoebe Lam, Oliver John and Iris Mauss, put versions of that question to more than a thousand people and then went looking for what the answers went with. Their paper appeared in the Journal of Personality and Social Psychology in 2018. Most of what a reader would reasonably want to know is answerable from the paper itself. Why any of it should work at all is not.
Isn’t that just giving up?
The researchers seem to have expected the objection, because they built a test for it.
Alongside the questions about accepting your own emotions, two of their samples answered a second, quite different set about accepting the situation — items like “I’ve been accepting the reality of the fact that it has happened.” The two turned out to have essentially nothing to do with each other. Someone who makes peace with an event is not thereby someone who lets themselves feel what they feel about it.
Accepting the situation also went almost nowhere. In the undergraduate sample it was associated with none of the psychological health measures. In the community sample it was associated with one of five, and even there, accepting your own emotions predicted well-being just as strongly with situational acceptance held constant. Ford and colleagues go further and point out that resigning yourself to a stressful situation could plausibly make things worse over time, if the situation is one you could have changed.
So the distinction they are drawing is between the event and your commentary on your own reaction to it, and it is doing real work.
Nobody involved in writing this has clinical training. What follows describes patterns across groups of strangers, and none of these studies screened anyone for a diagnosis or could tell a bad month apart from a depression that needs treating.
Does it hold when something actually goes wrong?
For the second study, 156 women from the California Bay Area came into a lab. Nearly all of them had been through a stressful life event in the previous six months. They watched a neutral film clip and rated how they felt. Then they were given two minutes to prepare a three-minute speech on why they were qualified for a job, told it would be filmed, and told that four trained judges would later take notes on the manner, content and quality of what they said and document their body language. A camera sat in front of them. If they went quiet for more than twenty seconds, an experimenter told them to keep going.
It worked as a stressor: negative emotion rose sharply, a large effect by the study’s own measure. Women higher in acceptance felt somewhat less of it during the speech, and the association held when their baseline mood was controlled for. It was a small association, and it had no counterpart on the positive side. Acceptance did nothing to how much excitement or calm they reported. Whatever it touches, it touches only the negative end.
The design is doing something specific here. Every participant got the identical stressor, which closes off the easiest alternative reading, that people who accept their feelings simply have gentler lives to accept.
It does not close off the harder one. Nobody in any of these studies was assigned to accept anything. Acceptance was a trait people already had when they walked in, measured by asking them, and people who do not berate themselves for feeling bad may differ from people who do in a dozen unmeasured ways. The causal arrow is an argument, not a finding. The authors say so themselves and call for intervention studies to settle it.
Does anything still show six months later?
The third study followed 222 adults around Denver, from 340 who enrolled, all of whom had been through a stressful event in the previous three months. Every night for fourteen consecutive nights they were asked to fill in a paper diary, naming the worst thing that had happened in the past twenty-four hours and rating twelve negative emotions they had felt during it. Roughly six months after the diaries ended, they answered a battery of psychological health questions.
People higher in acceptance had reported less negative emotion during their daily stressors. They had not reported more positive emotion, or less. And that accumulated daily negative emotion statistically accounted for the link between acceptance and how they were doing half a year on: completely, in the case of life satisfaction and depressive symptoms, and partly for well-being and the two anxiety measures.
Why it should work that way is the part the paper does not settle. The authors offer three candidate routes and label all of them untested: acceptance may head off rumination, or head off attempts to suppress the thought, or head off the second feeling about the first one, the guilt about the anger, which is what keeps the anger in the room. Any of those would produce the same numbers.
Who these people were shapes how far any of it travels. The first study ran on Berkeley undergraduates, the second was entirely women, the third was three quarters European American, and almost everything in all three is self-report, including how bad the day was and how bad it felt. The lab study came in below the statistical power the authors had set as their own target.
An earlier study says it depends on how bad things are
A prospective study of 55 women in Denver, published in 2010 by Amanda Shallcross, Allison Troy, Matthew Boland and Iris Mauss, found that acceptance predicted lower depressive symptoms after high life stress but not after low stress. In other words, it mattered when things were bad and was beside the point when they were not.
Ford and colleagues looked for exactly that interaction, in both an undergraduate and a community sample, and did not find it. They suggest the difference may come down to the two teams using different questionnaires; the earlier one measured how hard people work to avoid feeling bad, which is a related but not identical thing. That is the likelier reading, and the larger study with the fuller set of controls is the better bet — though the earlier finding is a live enough disagreement that anyone quoting either number should say the other exists.
What survives all of it is an average difference visible across crowds of people. It describes a tendency, and no study here tested it as something a person could take up and apply. Depression and anxiety are treatable, and treatment is what a therapist is for.
The authors close by putting their result inside an argument much older than any questionnaire. The Stoics treated emotions as dangerous and urged people to get them under control. Hume and the sentimentalists thought feelings should be permitted to run their natural course. Ford and colleagues write that their results “caution against the Stoic approach” — a very restrained way of saying that a great deal of advice about mastering your feelings may have been aimed at the wrong target.
