Should you play with your kids?
(The answer may surprise you)

If you’re a lonely tween in the year of our lord 2024, you have been algorithm’d into watching a TikTok video where a middle-aged woman insists that not playing pretend with her kids is good, actually. Engagement with these ideas is driven by outrage, which has been well-understood to be social media catnip. The idea of not playing with our kids then gets picked up by outlets like The Atlantic, who appear to make a good business out of commentary that is 10% smarter than TikTok comments. (In fairness to The Atlantic, their take on this subject is perfectly fine). Later, the NYT picks up with an op-ed that uses the unfalsifiable evopsych claim that hunter-gatherers played a certain way, which seems deliberately designed to make me, personally, angry. As it turns out, perhaps the NYT was ahead of this trend, with a similar op-ed in 2021.

In cases where these ideas surface, if you’re a nerd, you might be compelled to wonder: where is the evidence for this? Imagine if you will: a randomized controlled trial (RCT) with two or three groups. Group 1 is the experimental “no adult play” group. Group 2 is the control group, what might be termed the “treatment-as-usual” (TAU) group, where the parents are not given any instruction about play. A potential third group would be where parents are given instructions to play a certain amount with their kids, and perhaps even instructions or coaching on how to play. The results of this study could give evidence towards supporting various claims, depending on the outcomes.

The “parenting influencer” presents no such evidence. But this is not about evidence, per se. This is a (minor) moral panic like many others before it. “Playing with your kids hurts them” is the new “YouTube hurts your kids,” which was the new “Television hurts your kids,” which was the old “Satanists hurt your kids,” which was the new “Books hurt kids.” It is more indicative of a political or philosophical perspective than about evidence. 

What, then, is the evidence? Due to how hard it is to conduct our imaginary RCT on pretend play (e.g., imagine telling parents that they should not engage in pretend play with their children, how ridiculous), there are many observational studies but few controls. The evidence is unsurprisingly mixed. Pretend play may encourage development, or very intelligent kids may be naturally better at pretend play, or perhaps it’s just an individual preference, or even a cultural activity.

Well then, we’ve settled it: should adults engage in imaginary play with their kids? A resounding Maybe. Now that we have that figured out, what is the alternative suggested by “parenting influencers” posting on Instagram?

I checked: she does specifically call for 5 or more hours of independent play from ages “0-7.” Being charitable and speaking of newborns and toddlers, they need approximately 12 hours of sleep per day. Ms. Sannes is suggesting that nearly half of their waking hours are to be spent in “adult-free” play. I understand that she would likely disagree with my interpretation here, but it seems to me that she is suggesting a minimal amount of adult interaction during “adult-free play.” 

This is contradicted by the experts cited in the Atlantic article. This is contradicted by most child development experts. This is also contradicted by the most impactful project to ever arise from behavior analysis: Betty Hart and Todd Risley’s 1995 book, Meaningful Differences.

Meaningful Differences began as an observational study that eventually followed 42 families in 4 socio-economic categories: high, middle, low, and “welfare.” The authors recorded family interactions for 1 hour per month, starting when the children were under 1 year old, continuing until 2 ½. Follow-up measures tracked academic achievement through 3rd grade, though they found that language development at age 3 predicted later academic performance. 

What they found changed the course of childhood development: the children heard more words, and a greater diversity of words, at higher SES levels. This neatly predicted academic performance. Most famously, by age 3, the highest and lowest SES category children had experienced a difference of approximately 30 million spoken words, which became known popularly as “the 30 million word gap.” 

In the preceding years, dozens of initiatives have popped up that encourage speaking to infants and children; California has radio ads that encourage you to “Talk, read, sing.” I referred to Hart & Risley’s work as the most “impactful,” and by that I mean that it had a truly outsized cultural impact that we do not often see from behavior analysis. 

42 families, 1 hour recordings per month (at the time, a researcher would have been physically in the home, increasing possible reactivity), and the fact that this study is now over 30 years old? These are all reasonable criticisms, but this larger 2017 study with automated recordings found similar results. 

Well, that’s it, then! The science is more or less settled: talk to kids a lot, it’s good. Please close your browser window now. Do not look at social media or –

How could there possibly be controversy around “talk to your kids?” The most obvious criticism is that there is blame, classism, or even racism at the heart of the findings. Critics claim that the authors say: Poor people have dumb kids, which is their own fault. I believe this to be the diametrical opposite of the conclusion. Betty Hart and Todd Risley were shaped by the behaviorism of the 60s; these were optimistic times when behavior analysts had no conception of the limits of the science. They thought they could change the world. The message of this study, to a behaviorist? Race or class doesn’t determine achievement – the environment does. Hart & Risley aren’t blaming the parents, and in fact they’re suggesting that every parent can have better outcomes for their children by changing their environment.

Other criticisms are aimed at cultural considerations. A later research team suggested that Hart & Risley didn’t understand that, in some cultures, adults don’t talk as frequently to children, but talk just as often around children. This criticism doesn’t stand up to scrutiny for me. Consider that spanking is a cultural practice, and that it may harm children. In this case, it is worth considering the value of the practice weighed against the harms it may produce. Is it worthwhile to continue speaking less frequently to children, if it causes harm?

Then there is a somewhat incoherent line of thinking that is quite popular on the internet. It goes something like this: why are the standards of rich people something to aspire to? This criticism would make more sense if Hart & Risley were encouraging kids to be future capitalists. The whole idea was to understand school performance and language acquisition. I’ll set aside the subject of school for the moment, because it’s true: some people think school is for producing employees for the rich. But having an extensive vocal repertoire isn’t just helpful for good little capitalists; if anything, if you want lil baby Karl to transform America into a Marxist utopia, he’s going to have to be an above-average communicator. Use your words, Karl!

Where might this moral panic come from, if the evidence is so clear? It’s been frequently reported that in 2024 adult Americans have fewer friendships than ever. Real wages have been stagnant since the 70s, and childcare is expensive; people are juggling more work and less free time. The remaining free time spent with kids is more isolated – adults have fewer friends, and after several pandemic years, so do children. This panic is a backlash against being your child’s only playmate, against the pressure of being solely responsible for play. 

No parent should be the only playmate of a child in virtually any case. Adults need to take care of themselves, mentally and physically, to be good parents. Children should play with other children. But adults can, and in fact should, play constructively with children at least some of the time.

What good could come from pretend play, specifically? Again, we don’t have the evidence yet. But there is theoretical overlap with ACT. One of the foundational skills of ACT is psychological flexibility. Psychological flexibility is, more or less, an ability to perceive and react to actual contingencies; psychological inflexibility is behaving according to verbal rules, and with insensitivity to actual contingencies. To illustrate: Flexible Felix looks around the room, but sees no phone; he picks up a toy banana and speaks into it, “Hello?” Inflexible Ian replies, “No! That is a banana! Banana is not phone!”

Flexible Felix and Inflexible Ian don’t sound remarkable if you know children. Most kids play pretend at some point, but some kids don’t like it. This scene plays out in every preschool in America almost every day. As far as we know currently, kids playing pretend doesn’t have any bearing on their adult life.

But in adults, the evidence in favor of flexibility is relatively robust. Flexibility is associated with some good outcomes, and inflexibility is associated with several bad outcomes. One goal of ACT is to increase psychological flexibility. An illustration, from Steven Hayes’ excellent book Get Out of Your Mind and Into Your Life: you need to loosen a screw, but you don’t have a screwdriver. You do have a plastic toothbrush and a lighter. Think of how you might solve this problem.

An inflexible person might think: I don’t have a screwdriver, and that’s what I need, I can’t do this. This is stupid. I’m angry at you for this entire exercise.

A flexible person might think: use the lighter to soften the plastic toothbrush, push the toothbrush against the screw, wait for the toothbrush to solidify, and use it to turn the screw.

By at least one definition of inflexibility, animals are not inflexible. Pigeons, for example, are generally sensitive to ratios of reinforcement. Pecking red circle A pays out at 1 pellet per 10 pecks (10:1) and green circle B pays out at 1 pellet per 40 pecks (40:1). The pigeon will generally peck the red circle approximately four times as often as the green. When the ratio is suddenly switched without warning, the pigeon will eventually change responding accordingly. Humans may be insensitive to actual contingencies because of verbal rules, and thus may engage in “irrational” responding to the more costly contingency (e.g., “This one was working before, maybe it will change back” or “Green is good”).

Could it be that pretend play is, ultimately, an ability to be flexible – one which has implications for better outcomes? I’m not the first person to see the overlap between ABA and improv, but for a less literal take, what about improvising communicative responses? People have good reasons, I’m sure, for not playing pretend with their children. But maybe they should try being a little bit more flexible.